THE	
  PROCESS	
  OF	
  ADAPTATION	
  
	
  
By	
  	
  
	
  
Samantha	
  Katrina	
  Baard	
  Perry	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
A	
  DISSERTATION	
  
	
  
Submitted	
  to	
  
Michigan	
  State	
  University	
  
in	
  partial	
  fulfillment	
  of	
  the	
  requirements	
  
for	
  the	
  degree	
  of	
  	
  
	
  
Psychology	
  –	
  Doctor	
  of	
  Philosophy	
  
	
  
2015	
  
	
  
	
  

	
  
ABSTRACT	
  

	
  
	
  

	
  

THE	
  PROCESS	
  OF	
  ADAPTATION	
  
	
  
By	
  
	
  
Samantha	
  Katrina	
  Baard	
  Perry	
  
It	
  is	
  no	
  surprise	
  why	
  there	
  is	
  resurgence	
  in	
  research	
  on	
  adaptation	
  –	
  we	
  engage	
  in	
  

work	
  environments	
  characterized	
  by	
  change.	
  However,	
  do	
  we,	
  as	
  researchers	
  and	
  
practitioners	
  in	
  organizational	
  science,	
  truly	
  understand	
  how	
  individuals	
  deal	
  with	
  these	
  
changes?	
  In	
  other	
  words,	
  do	
  we	
  understand	
  how	
  individuals	
  adapt?	
  For	
  the	
  last	
  several	
  
decades,	
  it	
  appears	
  the	
  answer	
  would	
  be:	
  not	
  sufficiently.	
  This	
  research	
  endeavor	
  was	
  
developed	
  to	
  push	
  the	
  boundaries	
  of	
  the	
  adaptation	
  literature	
  by	
  theorizing	
  a	
  dynamic	
  
process	
  of	
  adaptation,	
  examining	
  the	
  embedded	
  cognitive	
  and	
  motivational	
  self-­‐regulatory	
  
mechanisms,	
  and	
  specifying	
  the	
  first	
  and	
  second	
  order	
  dynamics	
  involved	
  (i.e.,	
  trajectory	
  
and	
  relationship	
  changes).	
  Through	
  empirically	
  examining	
  both	
  the	
  post-­‐change	
  adaptation	
  
process	
  and	
  subsequent	
  routine	
  performance	
  process,	
  the	
  findings	
  reveal	
  that	
  these	
  two	
  
processes	
  are	
  distinct.	
  Two	
  types	
  of	
  adaptive	
  changes	
  were	
  investigated,	
  and	
  the	
  results	
  
show	
  similar	
  trajectories	
  and	
  relationships	
  across	
  these	
  change	
  types,	
  providing	
  initial	
  
evidence	
  of	
  the	
  generalizability	
  of	
  the	
  adaptation	
  process.	
  Furthermore,	
  the	
  patterns	
  of	
  the	
  
cognitive	
  and	
  motivational	
  cycles	
  presented	
  an	
  interesting	
  picture	
  of	
  the	
  inner	
  workings	
  of	
  
the	
  adaptation	
  process	
  in	
  correcting	
  performance	
  decreases	
  after	
  a	
  change.	
  Through	
  using	
  
discontinuous	
  growth	
  curve	
  analyses,	
  and	
  a	
  partially	
  crossed,	
  partially	
  time-­‐varying,	
  cross-­‐
lag	
  panel	
  regression	
  model,	
  a	
  clear	
  distinction	
  was	
  evident	
  in	
  the	
  utility	
  of	
  these	
  dynamic	
  
techniques.	
  Although	
  the	
  more	
  standard	
  trajectory	
  analyses	
  offered	
  a	
  straightforward	
  
picture	
  of	
  how	
  variables	
  increase	
  or	
  decrease	
  over	
  time,	
  the	
  more	
  sophisticated	
  cross-­‐lag	
  
	
  

	
  
analysis	
  provided	
  a	
  unique	
  perspective,	
  as	
  that	
  model	
  accounts	
  for	
  many	
  sources	
  of	
  
variance	
  in	
  the	
  estimation	
  of	
  relationships.	
  This	
  research	
  provides	
  a	
  first	
  step	
  into	
  what	
  
could	
  be	
  a	
  very	
  fruitful	
  path	
  investigating	
  the	
  dynamics	
  of	
  the	
  adaptation	
  process.	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

	
  

This	
  work	
  is	
  dedicated	
  to	
  my	
  wonderful	
  husband,	
  William	
  C.	
  Perry,	
  
and	
  to	
  my	
  loving	
  parents,	
  Paul	
  and	
  Veronica	
  Baard.

iv	
  

	
  
ACKNOWLEDGEMENTS	
  
	
  
I	
  would	
  like	
  to	
  sincerely	
  thank	
  my	
  advisor,	
  Dr.	
  Steve	
  Kozlowski,	
  who	
  provided	
  me	
  
with	
  invaluable	
  guidance	
  and	
  direction	
  in	
  my	
  graduate	
  career	
  and	
  superb	
  intellectual	
  
training	
  so	
  I	
  could	
  effectively	
  reach	
  my	
  goals.	
  He	
  not	
  only	
  assisted	
  me	
  in	
  the	
  expansion	
  of	
  
my	
  knowledge	
  of	
  conducting	
  behavioral	
  research,	
  but	
  also	
  was	
  instrumental	
  in	
  my	
  growth	
  
in	
  theory	
  development	
  and	
  writing	
  prowess.	
  
A	
  special	
  thank	
  you	
  is	
  also	
  due	
  to	
  my	
  committee	
  members,	
  Dr.	
  Georgia	
  Chao,	
  Dr.	
  
Rick	
  DeShon,	
  and	
  Dr.	
  Kevin	
  Ford	
  for	
  their	
  excellent	
  advice	
  and	
  generous	
  dedication	
  of	
  time	
  
and	
  effort	
  on	
  my	
  behalf.	
  It	
  has	
  been	
  an	
  honor	
  to	
  work	
  with	
  them	
  and	
  see	
  how	
  their	
  insights	
  
have	
  not	
  only	
  assisted	
  in	
  the	
  betterment	
  of	
  this	
  dissertation,	
  but	
  also	
  in	
  my	
  educational	
  
development	
  and	
  clarity	
  of	
  my	
  thinking.	
  
I	
  also	
  owe	
  immense	
  gratitude	
  to	
  my	
  ever-­‐supportive	
  husband,	
  William	
  Perry,	
  who	
  
has	
  been	
  a	
  consistently	
  positive	
  presence,	
  keeping	
  my	
  perspective	
  in	
  check	
  through	
  each	
  
step	
  of	
  this	
  process.	
  I	
  would	
  also	
  like	
  to	
  share	
  my	
  deepest	
  appreciation	
  to	
  my	
  parents,	
  Paul	
  
and	
  Veronica	
  Baard,	
  who	
  have	
  provided	
  incredible	
  support	
  and	
  encouragement	
  for	
  me	
  to	
  
not	
  only	
  complete	
  this	
  dissertation	
  but	
  also	
  to	
  pursue	
  education	
  throughout	
  my	
  entire	
  life,	
  
and	
  to	
  my	
  God,	
  Jesus	
  Christ,	
  by	
  whose	
  strength	
  I	
  could	
  accomplish	
  this	
  and	
  to	
  whom	
  I	
  will	
  
always	
  give	
  thanks.	
  
	
  

	
  

v	
  

	
  

	
  

TABLE	
  OF	
  CONTENTS	
  

LIST	
  OF	
  TABLES	
  ......................................................................................................................................	
  x	
  
	
  
LIST	
  OF	
  FIGURES	
  ..................................................................................................................................	
  xi	
  
	
  
INTRODUCTION	
  .....................................................................................................................................1	
  
	
  
ADAPTATION	
  –	
  THE	
  PROCESS	
  OF	
  RESPONDING	
  TO	
  CHANGES	
  	
  ..............................................3	
  
What	
  is	
  Change?	
  .................................................................................................................................3	
  
Overview	
  of	
  the	
  Adaptation	
  Literature	
  .....................................................................................6	
  
The	
  Process	
  of	
  Adaptation	
  .............................................................................................................9	
  
Part	
  1:	
  The	
  Dynamic	
  Nature	
  of	
  the	
  Process	
  ............................................................................................	
  10	
  
The	
  Changing	
  Task	
  Environments	
  ..............................................................................................	
  10	
  
Mapping	
  Complexity	
  Change	
  Onto	
  the	
  Larger	
  Literature	
  on	
  Change	
  .....................	
  11	
  
Adaptation,	
  Learning,	
  and	
  Self-­‐Regulation	
  as	
  Dynamic	
  Processes	
  ..........................	
  13	
  
Part	
  2:	
  The	
  Sub-­‐Cycles	
  ......................................................................................................................................	
  18	
  
The	
  Cognitive	
  Cycle	
  .............................................................................................................................	
  20	
  
The	
  Motivational	
  Cycle	
  ......................................................................................................................	
  23	
  
Behavioral	
  Mechanisms....................................................................................................................	
  25	
  
Two	
  Orders	
  of	
  Change	
  ......................................................................................................................................	
  26	
  
	
  
FIRST	
  ORDER	
  CHANGES:	
  TRAJECTORIES	
  ....................................................................................	
  27	
  
The	
  Trajectory	
  of	
  Performance	
  .................................................................................................	
  28	
  
The	
  Driver	
  of	
  Performance	
  Trajectory	
  Changes	
  ...................................................................................	
  32	
  
	
  The	
  Trajectories	
  of	
  the	
  Process	
  Mechanisms	
  ......................................................................	
  33	
  
The	
  Cognitive	
  Cycle	
  ...........................................................................................................................................	
  34	
  
Adaptive	
  Environment	
  ......................................................................................................................	
  34	
  
Performance	
  Environment	
  .............................................................................................................	
  38	
  
The	
  Motivational	
  Cycle	
  ....................................................................................................................................	
  40	
  
Adaptive	
  Environment	
  ......................................................................................................................	
  40	
  
Performance	
  Environment	
  .............................................................................................................	
  44	
  
	
  
SECOND	
  ORDER	
  CHANGES:	
  RELATIONSHIPS	
  ............................................................................	
  47	
  
The	
  Cognitive	
  Cycle	
  .......................................................................................................................	
  48	
  
Performance	
  and	
  Learning-­‐Oriented	
  Effort	
  ...........................................................................................	
  50	
  
Learning-­‐Oriented	
  Effort	
  and	
  Metacognition	
  .......................................................................................	
  51	
  
Metacognition	
  and	
  Performance	
  ................................................................................................................	
  53	
  
Performance	
  and	
  Evaluation	
  ........................................................................................................................	
  55	
  
Evaluation	
  and	
  Learning-­‐Oriented	
  Effort	
  ...............................................................................................	
  57	
  
Evaluation	
  and	
  Metacognition	
  .....................................................................................................................	
  58	
  
The	
  Motivational	
  Cycle	
  .................................................................................................................	
  60	
  
Performance	
  and	
  Goals	
  ...................................................................................................................................	
  61	
  
Goals	
  and	
  Outcome-­‐Oriented	
  Effort	
  ...........................................................................................................	
  63	
  
Outcome-­‐Oriented	
  Effort	
  and	
  Performance	
  ...........................................................................................	
  65	
  
Performance	
  and	
  Self-­‐efficacy	
  ......................................................................................................................	
  66	
  
Self-­‐efficacy	
  and	
  Goals	
  .....................................................................................................................................	
  68	
  
Self-­‐efficacy	
  and	
  Outcome-­‐Oriented	
  Effort	
  .............................................................................................	
  70	
  

	
  

vi	
  

	
  
METHOD	
  ................................................................................................................................................	
  72	
  
Participants	
  .....................................................................................................................................	
  72	
  
Task	
  ....................................................................................................................................................	
  73	
  
Design	
  ................................................................................................................................................	
  74	
  
Day	
  1:	
  Training	
  ...................................................................................................................................................	
  74	
  
Familiarization	
  Phase	
  ........................................................................................................................	
  75	
  
Training	
  Phase	
  .......................................................................................................................................	
  75	
  
Day	
  2:	
  Performance	
  ...........................................................................................................................................	
  76	
  
Measures	
  ...........................................................................................................................................	
  78	
  
Statistical	
  Models	
  ...........................................................................................................................	
  81	
  
	
  
RESULTS	
  ................................................................................................................................................	
  85	
  
Pre-­‐Hypothesis	
  Testing	
  ...............................................................................................................	
  85	
  
Variable	
  Information	
  and	
  Data	
  Cleaning	
  ................................................................................................	
  85	
  
Tests	
  of	
  Training	
  Effectiveness	
  .....................................................................................................................	
  89	
  
Hypothesis	
  Testing	
  ........................................................................................................................	
  91	
  
Identifying	
  the	
  Transition	
  Point	
  ..................................................................................................................	
  91	
  
Examining	
  Trajectory	
  Changes	
  ....................................................................................................................	
  96	
  
Understanding	
  Relationship	
  Changes	
  ....................................................................................................	
  103	
  
Results	
  of	
  the	
  Cognitive	
  Cycle	
  ......................................................................................................	
  107	
  
Results	
  of	
  the	
  Motivational	
  Cycle	
  ..............................................................................................	
  111	
  
	
  
DISCUSSION	
  .......................................................................................................................................	
  114	
  
Discussion	
  of	
  the	
  Results	
  ..........................................................................................................	
  115	
  
1st	
  Order	
  Changes:	
  Transition	
  and	
  Trajectories	
  ................................................................................	
  115	
  
2nd	
  Order	
  Changes:	
  Relationships	
  ............................................................................................................	
  118	
  
Cognitive	
  Cycle	
  Observations	
  ......................................................................................................	
  118	
  
Motivational	
  Cycle	
  Observations	
  ...............................................................................................	
  120	
  
Overall	
  Observations	
  of	
  the	
  Results	
  .........................................................................................................	
  122	
  
Practical	
  Implications	
  ...............................................................................................................	
  125	
  
Limitations	
  and	
  Future	
  Research	
  ..........................................................................................	
  126	
  
Conclusion	
  .....................................................................................................................................	
  133	
  
	
  
APPENDICES	
  ......................................................................................................................................	
  134	
  
Appendix	
  A:	
  Longitudinal	
  Cross-­‐Lag	
  Simulations	
  ............................................................	
  135	
  
Appendix	
  B:	
  Overall	
  Flow	
  of	
  the	
  Two-­‐Day	
  Experiment	
  ..................................................	
  156	
  
Appendix	
  C:	
  Training	
  Materials	
  .............................................................................................	
  158	
  
Appendix	
  D:	
  Measures	
  ...............................................................................................................	
  162	
  
Appendix	
  E:	
  IRB	
  Documentation	
  ...........................................................................................	
  173	
  
	
  
REFERENCES	
  .....................................................................................................................................	
  176	
  
	
  
	
  

	
  

vii	
  

	
  
LIST	
  OF	
  TABLES	
  
	
  
	
  
	
  
Table	
  1	
  Types	
  of	
  Change	
  in	
  the	
  Organizational	
  Literature	
  .....................................................4	
  
	
  
Table	
  2	
  Specific	
  Adaptive	
  Manipulations	
  for	
  Each	
  Experimental	
  Condition	
  .................	
  83	
  
	
  
Table	
  3	
  Descrptives	
  and	
  Correlations:	
  Adaptive	
  Environment	
  ..........................................	
  86	
  
	
  
Table	
  4	
  Descriptives	
  and	
  Correlations:	
  Performance	
  Environment	
  ................................	
  87	
  
	
  
Table	
  5	
  Discontinuous	
  Growth	
  Curve	
  Analysis	
  of	
  Training	
  and	
  Performance	
  ..............	
  90	
  
	
  
Table	
  6	
  Results	
  of	
  Discontinuous	
  Growth	
  Curve	
  Analyses	
  for	
  Performance	
  .................	
  94	
  
	
  
Table	
  7	
  Results	
  of	
  Discontinuous	
  Growth	
  Curve	
  Analyses	
  for	
  Type	
  of	
  Effort	
  ................	
  95	
  
	
  
Table	
  8	
  Summary	
  of	
  Trajectory	
  Hypotheses,	
  Analyses,	
  and	
  Results	
  ................................	
  98	
  
	
  
Table	
  9	
  Latent	
  Growth	
  Curve	
  Analyses	
  of	
  Cognitively	
  Focused	
  Variables	
  ...................	
  101	
  
	
  
Table	
  10	
  Latent	
  Growth	
  Curve	
  Analyses	
  of	
  Motivationally	
  Focused	
  Variables	
  ..........	
  102	
  
	
  
Table	
  11	
  Autoregressive	
  Results	
  from	
  the	
  Cross-­‐Lag	
  Model	
  ............................................	
  105	
  
	
  
Table	
  12	
  Summary	
  of	
  Hypotheses	
  and	
  Results	
  from	
  the	
  Cross-­‐Lag	
  Model	
  ..................	
  105	
  
	
  
Table	
  13	
  Fully	
  Crossed	
  Time	
  Invariant	
  Longitudinal	
  Cross-­‐Lag	
  Simulation	
  With	
  Two	
  	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Variables	
  .........................................................................................................................	
  141	
  
	
  
Table	
  14	
  Partially	
  Crossed	
  Partially	
  Time	
  Varying	
  Longitudinal	
  Cross-­‐Lag	
  	
  	
  	
  	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Simulation	
  With	
  Seven	
  Variables	
  ............................................................................	
  149	
  
	
  
Table	
  15	
  Day	
  1:	
  Training	
  ..............................................................................................................	
  156	
  
	
  
Table	
  16	
  Day	
  2:	
  Performance	
  ......................................................................................................	
  157	
  

	
  

x	
  

	
  
LIST	
  OF	
  FIGURES	
  
	
  
	
  
	
  
Figure	
  1	
  Heuristic	
  of	
  a	
  Continuum	
  of	
  Change	
  Types	
  ..............................................................	
  13	
  
	
  
Figure	
  2	
  Heuristic	
  of	
  the	
  Adaptation	
  Process	
  ...........................................................................	
  17	
  
	
  
Figure	
  3	
  Example	
  Trajectories	
  of	
  Key	
  Mechanisms	
  Across	
  the	
  Adaptation	
  and	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Performance	
  Environments	
  ..........................................................................................	
  31	
  
	
  
Figure	
  4	
  Heuristic	
  Representations	
  of	
  the	
  Relationships	
  in	
  the	
  Adaptation	
  and	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Performance	
  Processes	
  ...................................................................................................	
  49	
  
	
  
Figure	
  5	
  Representation	
  of	
  the	
  TANDEM	
  Task	
  Environment	
  ..............................................	
  74	
  
	
  
Figure	
  6	
  Example	
  of	
  a	
  Fully	
  Crossed	
  Longitudinal	
  Cross-­‐Lag	
  Panel	
  Model	
  With	
  	
  	
  	
  	
  	
  	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Two	
  Variables	
  .....................................................................................................................	
  83	
  
	
  
Figure	
  7	
  Example	
  of	
  the	
  Hypothesized	
  Relationships	
  in	
  the	
  Longitudinal	
  Cross-­‐Lag	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Panel	
  Regression	
  Model	
  ..................................................................................................	
  84	
  
	
  
Figure	
  8	
  Discontinuous	
  Growth	
  Curves	
  for	
  Performance	
  ....................................................	
  94	
  
	
  
Figure	
  9	
  Discontinuous	
  Growth	
  Curves	
  for	
  Type	
  of	
  Effort	
  ...................................................	
  95	
  
	
  
Figure	
  10	
  Trajectories	
  of	
  Cognitively	
  Focused	
  Variables	
  ..................................................	
  101	
  
	
  
Figure	
  11	
  Trajectories	
  of	
  Motivationally	
  Focused	
  Variables	
  ...........................................	
  102	
  
	
  
Figure	
  12	
  Comparison	
  of	
  the	
  Theoretical	
  Model	
  and	
  Empirical	
  Findings	
  ...................	
  106	
  
	
  
Figure	
  13	
  Individual	
  Variability	
  in	
  Performance	
  Trajectories	
  ........................................	
  131	
  
	
  
Figure	
  14	
  Individual	
  Variability	
  in	
  Self-­‐Regulatory	
  Variables	
  ........................................	
  132	
  
	
  
Figure	
  15	
  IRB	
  Informed	
  Consent	
  and	
  Approval	
  Documents	
  .............................................	
  173	
  

	
  

xi	
  

	
  
INTRODUCTION	
  
Changing,	
  dynamic,	
  unstable,	
  and	
  unpredictable	
  are	
  words	
  typically	
  used	
  to	
  
characterize	
  the	
  present	
  workplace.	
  It	
  is	
  less	
  than	
  surprising	
  then	
  that	
  adaptation	
  has	
  
become	
  a	
  frequently	
  used	
  term.	
  Although	
  companies	
  consistently	
  say	
  they	
  need	
  individuals	
  
who	
  can	
  quickly	
  adapt	
  to	
  changes,	
  little	
  is	
  truly	
  understood	
  about	
  the	
  process	
  of	
  adaptation.	
  
However,	
  the	
  process	
  –	
  or	
  a	
  reoccurring	
  set	
  of	
  behaviors	
  –	
  is	
  what	
  leads	
  to	
  the	
  desirable	
  
outcomes	
  (i.e.,	
  high	
  performance)	
  organizations	
  seek	
  when	
  new	
  situations	
  arise	
  (Burke,	
  
Stagl,	
  Salas,	
  Pierce	
  &	
  Kendall,	
  2006;	
  Kozlowski,	
  Gully,	
  Salas	
  &	
  Cannon-­‐Bowers,	
  1996).	
  
One	
  may	
  ask:	
  why	
  is	
  it	
  necessary	
  to	
  understand	
  how	
  individuals	
  change	
  as	
  long	
  as	
  
they	
  do	
  so	
  effectively?	
  I	
  offer	
  two	
  reasons	
  why	
  researchers	
  should	
  be	
  interested	
  in	
  the	
  
means	
  as	
  well	
  as	
  the	
  ends:	
  1)	
  through	
  identifying	
  and	
  understanding	
  the	
  process,	
  we	
  can	
  
learn	
  to	
  manipulate	
  it	
  through	
  training	
  in	
  order	
  to	
  make	
  individuals	
  more	
  effective	
  at	
  
adapting	
  to	
  changes,	
  and	
  2)	
  since	
  adaptation	
  is	
  an	
  inherently	
  longitudinal	
  and	
  dynamic	
  
phenomenon,	
  a	
  process	
  framework	
  allows	
  the	
  opportunity	
  to	
  discuss	
  how	
  the	
  mechanisms	
  
dynamically	
  adjust	
  to	
  the	
  demands	
  presented	
  by	
  different	
  environments	
  over	
  time.	
  	
  
While	
  the	
  term	
  adaptability	
  or	
  adaptation	
  has	
  been	
  in	
  the	
  organizational	
  literature	
  
since	
  the	
  early-­‐to-­‐mid-­‐1900s	
  (e.g.,	
  Ghiselli,	
  1951;	
  Terreberry,	
  1968;	
  Tiffin	
  &	
  Lawshe,	
  1942;	
  
Trites,	
  Kubala,	
  &	
  Cobb,	
  1959),	
  the	
  majority	
  of	
  thorough,	
  scientifically-­‐driven	
  approaches	
  to	
  
the	
  study	
  of	
  adaptation	
  has	
  only	
  emerged	
  over	
  the	
  past	
  few	
  decades,	
  as	
  researchers	
  have	
  
responded	
  to	
  the	
  need	
  to	
  investigate	
  the	
  increasingly	
  dynamic	
  organizational	
  
environments.	
  Particularly	
  in	
  reference	
  to	
  the	
  process	
  approach	
  to	
  the	
  study	
  of	
  adaptation,	
  
research	
  has	
  fallen	
  short	
  of	
  addressing	
  the	
  two	
  key	
  reasons	
  of	
  why	
  investigating	
  dynamic	
  
processes	
  is	
  so	
  vital	
  in	
  three	
  ways.	
  First,	
  investigations	
  need	
  to	
  be	
  conducted	
  at	
  the	
  

	
  

1	
  

	
  
individual	
  level	
  of	
  analysis	
  since	
  individuals	
  are	
  the	
  ones	
  who	
  adapt	
  to	
  changes	
  within	
  
teams	
  and	
  organizations;	
  however,	
  the	
  vast	
  majority	
  of	
  the	
  theoretical	
  frameworks	
  of	
  the	
  
adaptation	
  process	
  are	
  at	
  the	
  team	
  level	
  (see	
  Baard,	
  Rench	
  &	
  Kozlowski,	
  2014	
  for	
  a	
  review).	
  
Second,	
  longitudinal	
  and	
  empirical	
  studies	
  are	
  also	
  needed	
  to	
  investigate	
  a	
  process;	
  
however,	
  the	
  current	
  set	
  of	
  articles	
  on	
  the	
  adaptation	
  process	
  is	
  exclusively	
  theoretical	
  
(Baard	
  et	
  al.,	
  2014).	
  Third,	
  and	
  of	
  critical	
  importance	
  for	
  this	
  research	
  endeavor,	
  new	
  
dynamic	
  analyses	
  are	
  needed	
  to	
  capture	
  the	
  cyclical	
  nature	
  of	
  a	
  process	
  (i.e.,	
  through	
  cross-­‐
lag	
  and	
  autoregressive	
  analyses	
  as	
  opposed	
  to	
  latent	
  growth	
  curve	
  analyses);	
  however,	
  not	
  
only	
  is	
  this	
  difficult	
  because	
  of	
  the	
  complexity	
  of	
  the	
  data	
  analysis	
  skills	
  needed,	
  but	
  the	
  
programs	
  for	
  such	
  analyses	
  are	
  only	
  beginning	
  to	
  be	
  established	
  in	
  the	
  literature.	
  
This	
  research	
  endeavor	
  will	
  address	
  all	
  three	
  of	
  these	
  gaps.	
  I	
  will	
  first	
  provide	
  an	
  
overview	
  of	
  several	
  types	
  of	
  change,	
  describe	
  the	
  main	
  perspectives	
  in	
  the	
  adaptation	
  
literature,	
  and	
  identify	
  where	
  this	
  research	
  endeavor	
  fits	
  as	
  well	
  as	
  where	
  it	
  will	
  advance	
  
the	
  field.	
  Then	
  I	
  will	
  describe	
  how	
  self-­‐regulation,	
  learning	
  and	
  control	
  theories	
  assist	
  in	
  
understanding	
  how	
  individuals	
  engage	
  in	
  the	
  adaptation	
  process	
  when	
  a	
  change	
  is	
  
introduced.	
  Following	
  this	
  overall	
  description	
  of	
  the	
  adaptation	
  process,	
  I	
  will	
  specify	
  the	
  
dynamics	
  of	
  the	
  mechanisms	
  involved	
  by	
  describing	
  two	
  orders	
  of	
  change	
  (Carver	
  &	
  
Scheier,	
  1982;	
  Powers,	
  1973):	
  the	
  trajectories	
  of	
  the	
  variables	
  and	
  the	
  relationships	
  
between	
  them.	
  “Trajectory”	
  changes	
  describe	
  how	
  the	
  variables	
  adjust	
  in	
  level	
  over	
  time	
  
while	
  “Relationship”	
  changes	
  present	
  the	
  dynamics	
  of	
  the	
  direction	
  and	
  strength	
  between	
  
variables	
  as	
  individuals	
  engage	
  in	
  the	
  adaptation	
  process.	
  Finally,	
  I	
  will	
  explicate	
  the	
  way	
  in	
  
which	
  this	
  theory	
  will	
  be	
  empirically	
  investigated	
  through	
  the	
  application	
  of	
  a	
  newer	
  
analysis	
  that	
  can	
  capture	
  the	
  reciprocal	
  relationships	
  inherent	
  in	
  the	
  adaptation	
  process.	
  

	
  

2	
  

	
  
ADAPTATION	
  –	
  THE	
  PROCESS	
  OF	
  RESPONDING	
  TO	
  CHANGES	
  
What	
  is	
  Change?	
  
	
  

Before	
  beginning	
  the	
  discussion	
  of	
  how	
  adaptation	
  is	
  defined	
  in	
  the	
  literature	
  and	
  

how	
  I	
  will	
  specifically	
  investigate	
  it,	
  it	
  is	
  necessary	
  to	
  take	
  a	
  step	
  back	
  and	
  consider	
  what	
  
triggers	
  the	
  need	
  for	
  adaptation	
  –	
  namely,	
  change.	
  In	
  the	
  organizational	
  literature,	
  many	
  
types	
  of	
  changes	
  have	
  been	
  identified,	
  though	
  these	
  are	
  conceptualized	
  at	
  the	
  system	
  level	
  
as	
  opposed	
  to	
  the	
  reactions	
  of	
  individuals	
  when	
  facing	
  a	
  change.	
  It	
  appears	
  that	
  one	
  trend	
  
in	
  the	
  change	
  literature	
  is	
  to	
  bifurcate	
  change	
  into	
  two	
  extremes	
  and	
  compare	
  them.	
  Using	
  
a	
  systems	
  oriented	
  perspective	
  on	
  change,	
  Nasim	
  and	
  Sushil	
  (2011)	
  discuss	
  four	
  sets	
  of	
  
change	
  types.	
  Table	
  1	
  summarizes	
  these	
  perspectives	
  from	
  their	
  review.	
  	
  	
  
Of	
  particular	
  interest	
  in	
  this	
  research	
  endeavor	
  is	
  the	
  difference	
  between	
  
incremental	
  and	
  revolutionary	
  changes.	
  They	
  present	
  these	
  two	
  types	
  as	
  having	
  roots	
  in	
  
many	
  historical	
  debates.	
  To	
  be	
  more	
  specific	
  about	
  these	
  two	
  types,	
  incremental	
  change	
  (or	
  
first-­‐order	
  change,	
  as	
  some	
  call	
  it)	
  is	
  described	
  as	
  engaging	
  in	
  small	
  variations	
  in	
  behaviors	
  
or	
  strategies	
  to	
  continue	
  on	
  the	
  same	
  desired	
  trajectory	
  (Nadler	
  &	
  Tushman,	
  1995),	
  but	
  it	
  
“lacks	
  creativity	
  to	
  discover	
  new	
  strategic	
  ideas”	
  (Atlas,	
  2007;	
  p.	
  258).	
  The	
  other	
  end	
  of	
  the	
  
spectrum	
  is	
  known	
  by	
  many	
  names,	
  including	
  revolutionary,	
  metamorphic,	
  discontinuous,	
  
and	
  transformational.	
  They	
  are	
  grouped	
  together	
  here	
  for	
  the	
  sake	
  of	
  showing	
  how	
  these	
  
types	
  are	
  often	
  discussed	
  –	
  as	
  the	
  antithesis	
  of	
  incremental	
  change	
  –	
  but	
  there	
  are	
  
differences.	
  Both	
  metamorphic	
  (Tushman	
  &	
  Romanelli,	
  1985)	
  and	
  transformative	
  (Dibella,	
  
2007)	
  change	
  types	
  are	
  described	
  as	
  large	
  recreations	
  or	
  reorientations,	
  whereby	
  
organizations	
  shift	
  their	
  core	
  values	
  and	
  beliefs.	
  However,	
  Nadler	
  and	
  Tushman	
  (1995)	
  
take	
  a	
  slightly	
  less	
  extreme	
  perspective	
  with	
  their	
  definition	
  of	
  discontinuous	
  change,	
  

	
  

3	
  

	
  
which	
  is	
  described	
  as	
  something	
  that	
  occurs	
  at	
  multiple	
  periods	
  of	
  time	
  where	
  a	
  new	
  
configuration	
  (e.g.,	
  a	
  new	
  strategy	
  or	
  organizational	
  structure)	
  is	
  needed.	
  They	
  describe	
  
this	
  period	
  as	
  being	
  very	
  turbulent,	
  but	
  that	
  it	
  results	
  in	
  stabilization.	
  This	
  definition	
  is	
  
more	
  in	
  line	
  with	
  an	
  adaptive	
  event	
  that	
  individuals	
  or	
  teams	
  may	
  deal	
  with	
  given	
  that	
  it	
  
would	
  be	
  highly	
  unlikely	
  for	
  individuals	
  to	
  be	
  asked	
  to	
  change	
  their	
  core	
  values	
  or	
  beliefs	
  in	
  
order	
  to	
  effectively	
  adapt	
  to	
  a	
  change.	
  For	
  example,	
  individuals	
  in	
  an	
  organization	
  will	
  not	
  
be	
  required	
  to	
  shift	
  from	
  being	
  an	
  accountant	
  to	
  a	
  human	
  resources	
  manager,	
  but	
  the	
  way	
  
in	
  which	
  some	
  authors	
  describe	
  this	
  metamorphic	
  change	
  suggests	
  that	
  organizations	
  may	
  
need	
  to	
  “re-­‐create”	
  themselves	
  if	
  faced	
  with	
  a	
  serious	
  enough	
  shift	
  in	
  their	
  environment.	
  
Some	
  authors	
  break	
  from	
  the	
  dichotomous	
  conversation	
  of	
  change	
  types	
  to	
  discuss	
  a	
  
mid-­‐level	
  type	
  of	
  change.	
  Atlas	
  (2007)	
  specified	
  that	
  a	
  second-­‐order	
  change	
  is	
  a	
  request	
  for	
  
Table	
  1	
  
Types	
  of	
  System	
  Level	
  Change	
  in	
  the	
  Organizational	
  Literature	
  
Type	
  of	
  change	
  
Planned	
  
vs.	
  
Emergent	
  
Static	
  (or	
  episodic)	
  
vs.	
  
Dynamic	
  

Piecemeal	
  	
  
vs.	
  
Holistic	
  

Incremental	
  
vs.	
  
Revolutionary	
  

	
  

Definition	
  
Planned	
  -­‐	
  change	
  is	
  due	
  to	
  a	
  series	
  of	
  pre-­‐planned	
  steps	
  
organizations	
  create	
  and	
  follow	
  
Emergent	
  -­‐	
  nonlinear	
  movement	
  between	
  stages	
  based	
  
on	
  the	
  environment’s	
  uncertainty	
  and	
  complexity	
  
Static	
  -­‐	
  similar	
  to	
  the	
  planned	
  change;	
  change	
  is	
  linear	
  
movement	
  from	
  one	
  state	
  to	
  another	
  
Dynamic	
  -­‐	
  based	
  off	
  of	
  the	
  emergent	
  approach;	
  focuses	
  
on	
  the	
  discontinuous	
  nature	
  of	
  organizations	
  and	
  their	
  
work	
  environments	
  
Piecemeal	
  -­‐	
  dealing	
  with	
  change	
  is	
  focusing	
  on	
  one	
  
aspect	
  of	
  the	
  change	
  at	
  a	
  time	
  (e.g.,	
  a	
  process,	
  context,	
  or	
  
outcome)	
  
Holistic	
  –	
  dealing	
  with	
  change	
  is	
  simultaneously	
  
considering	
  all	
  the	
  factors	
  that	
  contribute	
  to	
  effective	
  
change	
  
Incremental	
  -­‐	
  a	
  series	
  of	
  more	
  routine	
  modifications	
  
that	
  are	
  made	
  to	
  maintain	
  the	
  pursuit	
  of	
  a	
  particular	
  
goal	
  or	
  outcome	
  
Revolutionary	
  -­‐	
  large	
  changes	
  that	
  are	
  needed	
  to	
  
overcome	
  the	
  inertia	
  that	
  the	
  organization	
  has	
  in	
  order	
  
to	
  remain	
  the	
  same	
  as	
  it	
  had	
  been.	
  

4	
  

Example	
  
Lewin’s	
  (1958)	
  unfreezing,	
  
change,	
  refreezing	
  model	
  

Senge’s	
  (1990)	
  complex	
  
dynamic	
  systems	
  of	
  learning	
  

Beer,	
  Eisenstat	
  &	
  Spector	
  	
  
(1990)	
  six	
  steps	
  to	
  
organizational	
  change	
  
Quinn’s	
  (1978)	
  theory	
  of	
  
logical	
  incrementalism	
  
	
  
Tushman	
  &	
  Romanelli’s	
  
(1995)	
  punctuated	
  
equilibrium	
  	
  

	
  
innovation.	
  The	
  author	
  suggests	
  there	
  are	
  two	
  key	
  aspects	
  to	
  be	
  dealt	
  with:	
  
transformational	
  (or	
  the	
  strategy,	
  mission,	
  and	
  culture	
  of	
  the	
  organization)	
  and	
  
transactional	
  (or	
  the	
  “psychological	
  and	
  organizational	
  variables	
  that	
  predict	
  and	
  control	
  
the	
  motivational	
  and	
  performance	
  outcomes”,	
  p.	
  258).	
  Both	
  of	
  these	
  types	
  will	
  be	
  of	
  
interest	
  in	
  the	
  discussion	
  of	
  adaptation,	
  and	
  the	
  concept	
  of	
  a	
  mid-­‐range	
  type	
  of	
  change	
  will	
  
be	
  a	
  focus	
  of	
  attention.	
  Similarly,	
  Reger	
  (1994)	
  described	
  “tectonic”	
  change	
  as	
  something	
  
larger	
  than	
  incremental	
  change,	
  but	
  not	
  as	
  substantial	
  as	
  transformational	
  change.	
  Instead,	
  
tectonic	
  changes	
  require	
  significant	
  action	
  on	
  the	
  part	
  of	
  the	
  organization,	
  but	
  it	
  builds	
  
upon	
  existing	
  aspects	
  of	
  the	
  identity	
  of	
  the	
  organization,	
  rather	
  than	
  requiring	
  a	
  re-­‐creation	
  
of	
  values	
  or	
  beliefs.	
  This	
  concept	
  of	
  a	
  change	
  being	
  larger	
  than	
  incremental	
  but	
  smaller	
  than	
  
revolutionary	
  will	
  be	
  key	
  to	
  how	
  the	
  present	
  study	
  will	
  be	
  investigating	
  adaptation.	
  
In	
  addition	
  to	
  the	
  consideration	
  of	
  mid-­‐range	
  changes,	
  some	
  authors	
  discuss	
  change	
  
types	
  using	
  a	
  continuum	
  framework.	
  Nadler	
  and	
  colleagues	
  expanded	
  on	
  the	
  ideas	
  
proposed	
  by	
  Tushman	
  and	
  Romanelli	
  (1985)	
  and	
  developed	
  a	
  taxonomy	
  describing	
  a	
  
gradient	
  of	
  change	
  types	
  (Nadler,	
  1988;	
  Nadler	
  &	
  Rushman,	
  1995).	
  These	
  authors	
  proposed	
  
that,	
  in	
  addition	
  to	
  incremental	
  and	
  discontinuous	
  changes,	
  there	
  are	
  anticipatory	
  and	
  
reactive	
  changes.	
  With	
  anticipatory	
  changes	
  organizations	
  pre-­‐plan	
  their	
  change,	
  which	
  can	
  
either	
  be	
  a	
  small	
  step	
  in	
  the	
  pursuit	
  of	
  a	
  goal,	
  or	
  a	
  dramatic	
  one.	
  On	
  the	
  other	
  hand,	
  reactive	
  
changes	
  are	
  not	
  expected	
  by	
  the	
  organization	
  and	
  require	
  immediate	
  action.	
  Nadler	
  and	
  
Tushman	
  (1995)	
  present	
  a	
  2x2	
  framework	
  of	
  incremental	
  and	
  discontinuous	
  change	
  paired	
  
with	
  either	
  anticipatory	
  or	
  reactive	
  change	
  to	
  make	
  a	
  gradient	
  of	
  four	
  paired	
  types	
  of	
  
change	
  (e.g.,	
  incremental	
  and	
  anticipatory).	
  With	
  these	
  four	
  factors	
  the	
  authors	
  suggest	
  that	
  
discontinuous	
  changes	
  that	
  are	
  also	
  reactive	
  have	
  the	
  highest	
  severity	
  of	
  impact	
  on	
  an	
  

	
  

5	
  

	
  
organization,	
  followed	
  by	
  discontinuous	
  changes	
  that	
  are	
  anticipatory.	
  Following	
  these	
  in	
  
severity	
  of	
  impact	
  are	
  incremental	
  changes,	
  with	
  those	
  that	
  are	
  reactive	
  having	
  a	
  stronger	
  
impact	
  than	
  those	
  that	
  are	
  anticipatory.	
  This	
  taxonomy	
  presents	
  an	
  initial	
  framework	
  that	
  
attempts	
  to	
  determine	
  which	
  types	
  of	
  changes	
  are	
  more	
  difficult	
  to	
  deal	
  with	
  than	
  others,	
  
but	
  these	
  authors,	
  like	
  many	
  others	
  in	
  the	
  organizational	
  change	
  literature,	
  have	
  chosen	
  to	
  
remain	
  with	
  the	
  mentality	
  that	
  there	
  are	
  two	
  key	
  types	
  of	
  change	
  in	
  any	
  given	
  framework,	
  
and	
  have	
  not	
  begun	
  the	
  process	
  of	
  unpacking	
  what	
  elements	
  of	
  change	
  many	
  make	
  the	
  
scaling	
  more	
  granular.	
  This	
  concept	
  will	
  be	
  key	
  for	
  understanding	
  how	
  this	
  study	
  
investigates	
  adaptation	
  in	
  the	
  context	
  of	
  the	
  change	
  literature.	
  However,	
  before	
  going	
  into	
  
much	
  depth	
  on	
  that	
  issue,	
  it	
  is	
  necessary	
  to	
  first	
  delve	
  into	
  understanding	
  the	
  breadth	
  of	
  
the	
  adaptation	
  literature	
  and	
  how	
  it	
  will	
  be	
  discussed	
  in	
  this	
  study.	
  
	
  
Overview	
  of	
  the	
  Adaptation	
  Literature	
  	
  
For	
  many	
  decades,	
  researchers	
  have	
  added	
  to	
  our	
  understanding	
  of	
  the	
  
phenomenon	
  broadly	
  labeled	
  “adaptation”	
  through	
  investigating	
  how	
  individuals,	
  teams	
  
and	
  organizations	
  adjust	
  behaviors	
  in	
  light	
  of	
  a	
  new	
  or	
  changed	
  environment	
  (e.g.,	
  Caldwell	
  
&	
  O’Reilly,	
  1982;	
  Kozlowski,	
  Gully,	
  Nason	
  &	
  Smith,	
  1999;	
  Pulakos,	
  Arad,	
  Donovan	
  &	
  
Plamondon,	
  2000;	
  Rosen,	
  Bedwell,	
  Wildman,	
  Fritzsche,	
  Salas,	
  &	
  Burke,	
  2011).	
  This	
  concept	
  
of	
  adaptation,	
  though	
  focused	
  on	
  performance	
  in	
  this	
  research,	
  has	
  been	
  discussed	
  broadly	
  
in	
  different	
  literatures;	
  asking	
  questions	
  about	
  socialization,	
  expatriate	
  adjustment,	
  stress	
  
and	
  coping.	
  Given	
  the	
  great	
  interest	
  in	
  this	
  phenomenon,	
  there	
  has	
  been	
  divergence	
  in	
  the	
  
conceptualization	
  and	
  study	
  of	
  performance	
  adaptation.	
  Therefore,	
  prior	
  to	
  describing	
  the	
  
importance	
  and	
  dynamics	
  of	
  the	
  self-­‐regulatory	
  mechanisms	
  involved	
  in	
  the	
  adaptation	
  

	
  

6	
  

	
  
process	
  as	
  it	
  is	
  presented	
  in	
  this	
  research	
  endeavor,	
  it	
  is	
  critical	
  to	
  determine	
  which	
  parts	
  of	
  
the	
  adaptation	
  literature	
  are	
  relevant	
  to	
  this	
  discussion.	
  	
  
In	
  a	
  recent	
  effort	
  to	
  organize,	
  synthesize,	
  and	
  direct	
  future	
  efforts	
  at	
  understanding	
  
adaptation,	
  Baard	
  and	
  colleagues	
  (2014)	
  created	
  a	
  taxonomy	
  to	
  organize	
  the	
  diverse	
  
definitions	
  that	
  have	
  been	
  proposed	
  over	
  the	
  years,	
  using	
  a	
  bottom-­‐up	
  approach	
  to	
  review	
  
the	
  extant	
  literature.	
  They	
  bifurcated	
  the	
  literature	
  into	
  domain	
  general	
  (a	
  situation-­‐
spanning	
  approach	
  where	
  the	
  adaptive	
  capability	
  of	
  individuals	
  and	
  the	
  instantiation	
  of	
  
adaptation	
  in	
  environments	
  are	
  generalizable	
  across	
  settings)	
  and	
  domain	
  specific	
  
conceptualizations	
  (a	
  situation-­‐specific	
  perspective	
  where	
  adaptation	
  requires	
  knowledge,	
  
skills	
  and	
  abilities	
  inherent	
  to	
  that	
  environment).	
  Of	
  the	
  domain	
  general	
  perspectives,	
  two	
  
streams	
  of	
  research	
  were	
  identified:	
  adaptation	
  as	
  a	
  performance	
  construct	
  (e.g.,	
  the	
  eight-­‐
dimension	
  framework	
  presented	
  by	
  Pulakos,	
  et	
  al.,	
  2000)	
  and	
  adaptation	
  as	
  an	
  individual	
  
difference	
  construct	
  (e.g.,	
  the	
  I-­‐ADAPT	
  measure	
  presented	
  by	
  Ployhart	
  &	
  Bliese,	
  2006).	
  As	
  
both	
  of	
  these	
  perspectives	
  view	
  adaptation	
  as	
  a	
  construct	
  versus	
  a	
  process,	
  as	
  this	
  
framework	
  does,	
  the	
  research	
  from	
  these	
  two	
  approaches	
  will	
  be	
  drawn	
  upon	
  rarely	
  in	
  the	
  
description	
  of	
  the	
  adaptation	
  process.	
  	
  
Domain	
  specific	
  research	
  efforts	
  (i.e.,	
  the	
  performance	
  change	
  and	
  process	
  
approaches),	
  however,	
  oftentimes	
  discuss	
  the	
  role	
  of	
  self-­‐regulatory	
  mechanisms	
  in	
  
adaptive	
  performance.	
  The	
  performance	
  change	
  perspective	
  is	
  primarily	
  concerned	
  with	
  
the	
  change	
  in	
  performance	
  that	
  is	
  evident	
  when	
  individuals	
  shift	
  from	
  a	
  routine	
  task	
  
environment	
  to	
  an	
  adaptive	
  one.	
  Research	
  in	
  this	
  line	
  of	
  adaptation	
  research	
  is	
  split	
  into	
  
three	
  streams:	
  a	
  simple	
  input-­‐output	
  change	
  (e.g.,	
  Dormann	
  &	
  Frese,	
  1994;	
  Frese,	
  
Broadbeck,	
  Heinbokel,	
  Mooser,	
  Schleiffenbaum,	
  &	
  Thiemann,	
  1991),	
  a	
  change	
  influenced	
  by	
  

	
  

7	
  

	
  
the	
  self-­‐regulatory	
  learning	
  process	
  (e.g.,	
  Bell	
  &	
  Kozlowski,	
  2002b,	
  2008;	
  Chen,	
  Thomas,	
  &	
  
Wallace,	
  2005),	
  and	
  a	
  difference	
  in	
  the	
  trajectory	
  of	
  performance	
  between	
  routine	
  and	
  
adaptive	
  scenarios	
  (e.g.,	
  LePine,	
  2003,	
  2005).	
  Research	
  investigating	
  the	
  learning	
  process	
  
involved	
  prior	
  to	
  adaptation	
  will	
  be	
  scrutinized	
  in	
  depth	
  as	
  these	
  works	
  provide	
  critical	
  
insights	
  into	
  the	
  self-­‐regulatory	
  mechanisms	
  involved,	
  with	
  some	
  of	
  these	
  researchers	
  
suggesting	
  that	
  the	
  pre-­‐adaptation	
  learning	
  process	
  mirrors	
  the	
  adaptation	
  process	
  
(Kozlowski,	
  Toney,	
  Mullins,	
  Weissbein,	
  Brown	
  &	
  Bell,	
  2001).	
  Given	
  the	
  importance	
  of	
  
understanding	
  the	
  dynamics	
  involved	
  in	
  adaptation,	
  the	
  literature	
  examining	
  the	
  
performance	
  trajectories	
  in	
  adaptation	
  (e.g.,	
  Lang	
  &	
  Bliese,	
  2009;	
  LePine,	
  2003,	
  2005)	
  will	
  
also	
  be	
  impactful	
  as	
  the	
  adaptation	
  process	
  is	
  examined.	
  The	
  second	
  domain-­‐specific	
  
approach,	
  the	
  process	
  perspective,	
  will	
  also	
  be	
  examined	
  in	
  depth.	
  This	
  stream	
  in	
  the	
  
adaptation	
  literature	
  typically	
  uses	
  self-­‐regulatory	
  mechanisms	
  to	
  specifically	
  describe	
  the	
  
adaptation	
  process	
  (e.g.,	
  Burke,	
  et	
  al.,	
  2006;	
  Kozlowski,	
  Watola,	
  Jensen,	
  Kim	
  &	
  Botero,	
  2009;	
  
Rosen,	
  et	
  al.,	
  2011).	
  These	
  theories	
  come	
  as	
  close	
  to	
  a	
  dynamic	
  understanding	
  of	
  the	
  self-­‐
regulatory	
  process	
  of	
  adaptation	
  as	
  is	
  currently	
  available	
  in	
  the	
  literature.	
  	
  
A	
  key	
  goal	
  of	
  this	
  research	
  is	
  to	
  present	
  a	
  model	
  that	
  combines	
  the	
  efforts	
  of	
  these	
  
theory	
  and	
  research	
  streams	
  in	
  the	
  adaptation	
  literature.	
  Similar	
  to	
  Chen	
  and	
  colleagues	
  
(2005),	
  I	
  extend	
  the	
  work	
  of	
  Kozlowski	
  and	
  colleagues	
  (Bell	
  &	
  Kozlowski,	
  2008;	
  Kozlowski,	
  
Gully,	
  Brown,	
  Salas,	
  Smith	
  &	
  Nason,	
  2001)	
  to	
  empirically	
  examine	
  the	
  self-­‐regulatory	
  
mechanisms	
  that	
  occur	
  after	
  training.	
  However,	
  I	
  go	
  beyond	
  Chen	
  et	
  al.	
  (2005)	
  in	
  that	
  the	
  
self-­‐regulatory	
  mechanisms	
  will	
  be	
  investigated	
  not	
  only	
  after	
  training,	
  but	
  after	
  an	
  
adaptive	
  event	
  is	
  initiated.	
  I	
  also	
  extend	
  the	
  work	
  of	
  the	
  researchers	
  investigating	
  
performance	
  change	
  after	
  an	
  adaptive	
  event	
  (e.g.,	
  Lang	
  &	
  Bliese,	
  2009;	
  LePine,	
  2003,	
  2005),	
  

	
  

8	
  

	
  
as	
  I	
  will	
  be	
  examining	
  the	
  changes	
  of	
  the	
  adaptive	
  process	
  mechanisms	
  over	
  time	
  as	
  well.	
  
Finally,	
  I	
  will	
  be	
  extending	
  the	
  process	
  approach	
  to	
  the	
  study	
  of	
  adaptation	
  (which	
  has	
  
focused	
  on	
  theoretical	
  frameworks	
  at	
  the	
  team	
  level)	
  in	
  two	
  ways:	
  through	
  presenting	
  a	
  
theory	
  of	
  individual	
  level	
  adaptation	
  and	
  by	
  examining	
  the	
  process	
  empirically	
  over	
  time.	
  
Therefore,	
  as	
  my	
  primary	
  focus	
  is	
  on	
  the	
  longitudinal	
  and	
  dynamic	
  process	
  of	
  adaptation,	
  I	
  
use	
  the	
  research	
  efforts	
  of	
  the	
  performance	
  change	
  approach	
  to	
  extend	
  our	
  understanding	
  
of	
  the	
  adaptation	
  as	
  I	
  attempt	
  to	
  fill	
  several	
  large	
  gaps	
  in	
  the	
  performance	
  adaptation	
  
literature	
  (see	
  Baard	
  et	
  al.,	
  2014).	
  
	
  
The	
  Process	
  of	
  Adaptation	
  
Adaptation	
  is	
  defined	
  in	
  this	
  research	
  as	
  the	
  set	
  of	
  “cognitive,	
  affective,	
  motivational,	
  
and	
  behavioral	
  modifications	
  made	
  in	
  response	
  to	
  the	
  demands	
  of	
  a	
  new	
  or	
  changing	
  
environment,	
  or	
  situational	
  demands”	
  (Baard	
  et	
  al.,	
  2014;	
  p.	
  46).	
  This	
  definition	
  identifies	
  
two	
  essential	
  elements	
  of	
  adaptation:	
  1)	
  it	
  is	
  a	
  dynamic	
  process	
  of	
  responding	
  to	
  changes	
  in	
  
the	
  environment,	
  and	
  2)	
  there	
  are	
  several	
  components,	
  or	
  sub-­‐cycles	
  (i.e.,	
  cognitions,	
  
motivations	
  and	
  behaviors),	
  that	
  require	
  modifications	
  when	
  such	
  changes	
  are	
  introduced.	
  
In	
  order	
  to	
  address	
  these	
  two	
  elements	
  of	
  the	
  adaptation	
  process,	
  I	
  will	
  first	
  describe	
  the	
  
demands	
  of	
  the	
  different	
  environments	
  in	
  which	
  individuals	
  work	
  as	
  well	
  as	
  present	
  the	
  
relevance	
  of	
  other	
  dynamic	
  theories	
  (i.e.,	
  learning	
  and	
  self-­‐regulation)	
  in	
  the	
  
conceptualization	
  of	
  the	
  process	
  individuals	
  engage	
  in	
  while	
  performing	
  tasks	
  in	
  these	
  
environments.	
  Then	
  I	
  will	
  address	
  the	
  second	
  element	
  of	
  the	
  definition	
  above	
  by	
  using	
  
research	
  from	
  the	
  adaptation,	
  learning,	
  and	
  self-­‐regulation	
  literatures	
  in	
  discussing	
  the	
  
unique	
  but	
  interrelated	
  components	
  of	
  the	
  adaptation	
  process.	
  

	
  

9	
  

	
  
Part	
  1:	
  The	
  Dynamic	
  Nature	
  of	
  the	
  Process	
  
The	
  Changing	
  Task	
  Environments.	
  Processes	
  are	
  dynamic	
  in	
  nature	
  and	
  
performance	
  adaptation	
  is	
  no	
  exception,	
  as	
  it	
  is	
  described	
  as	
  the	
  dynamic	
  process	
  of	
  
responding	
  to	
  environmental	
  demands.	
  Individuals	
  typically	
  perform	
  tasks	
  in	
  routine	
  
performance	
  environments	
  where	
  there	
  is	
  a	
  clear	
  understanding	
  of	
  the	
  impact	
  of	
  certain	
  
actions	
  on	
  outcomes	
  of	
  interest.	
  However,	
  when	
  a	
  change	
  occurs,	
  there	
  is	
  a	
  shift	
  in	
  
environments	
  –	
  now	
  an	
  adaptive	
  environment	
  is	
  present	
  with	
  a	
  new	
  set	
  of	
  demands.	
  
Individuals	
  must	
  now	
  engage	
  in	
  the	
  adaptation	
  process	
  in	
  order	
  to	
  adjust	
  their	
  cognitions	
  
and	
  behaviors,	
  given	
  the	
  new	
  requirements	
  of	
  the	
  environment.	
  Once	
  these	
  modifications	
  
have	
  been	
  made,	
  individuals	
  re-­‐enter	
  a	
  performance	
  environment	
  where	
  there	
  is	
  again	
  a	
  
clear	
  understanding	
  of	
  the	
  relationship	
  between	
  behaviors	
  and	
  outcomes	
  (although	
  these	
  
relationships	
  might	
  be	
  different	
  than	
  they	
  were	
  in	
  the	
  performance	
  environment	
  prior	
  to	
  
the	
  change).	
  Changes	
  activate	
  the	
  adaptation	
  process	
  as	
  individuals	
  must	
  detect	
  that	
  a	
  
change	
  occurred,	
  diagnose	
  the	
  source,	
  determine	
  what	
  an	
  appropriate	
  response	
  is,	
  act	
  
based	
  on	
  that	
  strategic	
  understanding,	
  and	
  regulate	
  motivations	
  to	
  devote	
  effort	
  to	
  address	
  
the	
  change	
  (Burke	
  et	
  al.,	
  2006;	
  Jundt,	
  2009;	
  Kozlowski,	
  et	
  al.,	
  1999;	
  Rosen	
  et	
  al.,	
  2011).	
  
Although	
  this	
  may	
  seem	
  to	
  be	
  a	
  sequential	
  set	
  of	
  actions,	
  these	
  cognitive	
  and	
  motivational	
  
behaviors	
  inform	
  each	
  other,	
  forming	
  a	
  cycle	
  where	
  the	
  testing	
  of	
  one’s	
  strategies	
  will	
  
inform	
  where	
  effort	
  is	
  placed	
  and	
  the	
  outcomes	
  of	
  behavior	
  inform	
  the	
  extent	
  of	
  the	
  need	
  
for	
  more	
  strategic	
  actions.	
  	
  
The	
  speed	
  with	
  which	
  individuals	
  shift	
  from	
  an	
  adaptive	
  environment	
  to	
  a	
  
routinized	
  environment	
  will	
  likely	
  depend	
  on	
  the	
  type	
  of	
  change	
  they	
  encounter.	
  Wood	
  
(1986)	
  suggests	
  that	
  there	
  are	
  three	
  types	
  of	
  complexity:	
  component,	
  coordinative	
  and	
  

	
  

10	
  

	
  
dynamic.	
  Component	
  complexity	
  is	
  a	
  change	
  in	
  the	
  number	
  of	
  distinct	
  actions	
  that	
  must	
  be	
  
executed	
  or	
  the	
  number	
  of	
  distinct	
  pieces	
  of	
  information	
  that	
  must	
  be	
  processed.	
  This	
  is	
  the	
  
simplest	
  form	
  of	
  complexity	
  change	
  and	
  typically	
  requires	
  individuals	
  to	
  increase	
  their	
  
effort	
  as	
  they	
  perform	
  more	
  acts	
  within	
  a	
  similar	
  period	
  of	
  time	
  (e.g.,	
  taking	
  on	
  your	
  
colleague’s	
  job	
  if	
  he/she	
  suddenly	
  becomes	
  incapacitated).	
  Coordinative	
  complexity	
  is	
  a	
  
change	
  in	
  the	
  relationship	
  between	
  an	
  input	
  and	
  an	
  outcome	
  through	
  a	
  shift	
  in	
  the	
  strength,	
  
structure,	
  or	
  sequence	
  of	
  information,	
  actions	
  or	
  outcomes.	
  This	
  is	
  a	
  more	
  challenging	
  form	
  
of	
  complexity	
  change	
  and	
  is	
  typically	
  evident	
  in	
  a	
  change	
  in	
  the	
  timing	
  of	
  the	
  task,	
  the	
  
frequency	
  of	
  information	
  or	
  communication	
  that	
  is	
  required,	
  or	
  the	
  way	
  in	
  which	
  the	
  task	
  is	
  
conducted	
  through	
  underlying	
  rule	
  changes	
  (e.g.,	
  adjusting	
  to	
  the	
  priorities	
  of	
  a	
  new	
  boss).	
  
The	
  third	
  and	
  most	
  complex	
  type	
  of	
  change	
  is	
  called	
  dynamic,	
  where	
  the	
  change	
  is	
  non-­‐
stationary	
  over	
  time.	
  Wood	
  (1986)	
  describes	
  that	
  this	
  type	
  can	
  occur	
  from	
  a	
  sudden	
  
intense	
  change	
  in	
  component	
  or	
  coordinative	
  complexity	
  that	
  has	
  implications	
  for	
  some	
  
time	
  (e.g.,	
  an	
  organizational	
  merger	
  is	
  a	
  one-­‐time	
  change	
  that	
  has	
  implications	
  over	
  the	
  
course	
  of	
  several	
  months	
  or	
  years),	
  or	
  as	
  a	
  continuous	
  shift	
  (e.g.,	
  having	
  an	
  information	
  
technology	
  job	
  that	
  requires	
  a	
  consistent	
  monitoring	
  and	
  learning	
  of	
  new	
  products	
  on	
  the	
  
market),	
  which	
  can	
  be	
  controllable	
  or	
  non-­‐controllable.	
  
Mapping	
  Complexity	
  Change	
  Onto	
  the	
  Larger	
  Literature	
  on	
  Change.	
  As	
  discussed	
  
earlier,	
  there	
  are	
  many	
  types	
  of	
  change.	
  Most	
  popularly	
  discussed	
  is	
  the	
  dichotomy	
  
between	
  incremental	
  and	
  metamorphic	
  change	
  (Tushman	
  &	
  Romanelli,	
  1985).	
  However,	
  
these	
  two	
  ends	
  of	
  the	
  spectrum	
  do	
  not	
  seem	
  to	
  capture	
  the	
  kind	
  of	
  complexity	
  change	
  to	
  
which	
  Wood	
  (1986)	
  refers.	
  Incremental	
  change	
  appears	
  too	
  basic,	
  since	
  it	
  is	
  described	
  as	
  a	
  
way	
  of	
  engaging	
  with	
  the	
  environment	
  to	
  make	
  minor	
  changes	
  that	
  result	
  in	
  the	
  

	
  

11	
  

	
  
maintenance	
  of	
  a	
  particular	
  trajectory.	
  As	
  an	
  aside,	
  there	
  is	
  a	
  distinction	
  between	
  a	
  routine	
  
performance	
  environment	
  and	
  incremental	
  change.	
  The	
  former	
  is	
  a	
  situation	
  where	
  
individuals	
  have	
  a	
  stable	
  equilibrium	
  that	
  they	
  are	
  maintaining,	
  with	
  no	
  desire	
  to	
  increase	
  
performance	
  or	
  goals;	
  whereas,	
  incremental	
  change	
  is	
  where	
  individuals	
  are	
  more	
  likely	
  to	
  
be	
  pursuing	
  a	
  higher	
  goal,	
  but	
  this	
  goal	
  pursuit	
  does	
  not	
  require	
  any	
  change	
  in	
  strategy	
  but	
  
rather	
  the	
  continuation	
  of	
  engagement	
  in	
  self-­‐regulatory	
  practices.	
  	
  
The	
  typical	
  incremental	
  and	
  metamorphic	
  change	
  types	
  do	
  not	
  appear	
  to	
  be	
  
appropriate	
  categories	
  for	
  the	
  current	
  perspective	
  of	
  adaptation.	
  Adaptive	
  change	
  differs	
  
from	
  incremental	
  change	
  in	
  that	
  the	
  latter	
  does	
  not	
  require	
  urgency	
  of	
  action,	
  and	
  it	
  is	
  not	
  
necessarily	
  based	
  on	
  external	
  pressure,	
  although	
  it	
  could	
  be	
  (e.g.,	
  an	
  organization	
  increases	
  
performance	
  goals	
  for	
  next	
  year	
  which	
  requires	
  more	
  effort,	
  but	
  is	
  not	
  strategy	
  altering).	
  
However,	
  adaptive	
  changes	
  are	
  usually	
  triggered	
  by	
  an	
  external	
  event	
  that	
  forces	
  
individuals	
  to	
  alter	
  their	
  strategy	
  and	
  do	
  so	
  immediately.	
  Individuals	
  do	
  not	
  typically	
  know	
  
what	
  specifically	
  changed	
  or	
  how,	
  or	
  whether,	
  they	
  will	
  be	
  able	
  to	
  re-­‐calibrate	
  their	
  
knowledge,	
  skills,	
  and	
  abilities	
  to	
  effectively	
  deal	
  with	
  the	
  change.	
  On	
  the	
  other	
  hand,	
  the	
  
idea	
  of	
  metamorphic	
  change	
  appears	
  to	
  be	
  too	
  extreme,	
  in	
  that	
  this	
  type	
  of	
  change	
  is	
  
described	
  as	
  a	
  shift	
  in	
  values	
  or	
  beliefs,	
  but	
  it	
  is	
  highly	
  unlikely	
  that	
  an	
  individual	
  would	
  be	
  
required	
  to	
  do	
  such	
  a	
  thing	
  in	
  a	
  typical	
  work	
  situation.	
  	
  
Therefore,	
  it	
  is	
  anticipated	
  that	
  the	
  types	
  of	
  change	
  that	
  are	
  associated	
  with	
  this	
  
investigation	
  of	
  the	
  adaptation	
  process	
  fall	
  in	
  the	
  middle	
  of	
  a	
  continuum	
  between	
  
incremental	
  and	
  metamorphic	
  change	
  types.	
  Figure	
  1	
  below	
  presents	
  a	
  general	
  framework	
  
for	
  the	
  distinction	
  between	
  the	
  two	
  types	
  of	
  adaptive	
  changes	
  that	
  will	
  be	
  examined,	
  as	
  well	
  
as	
  presenting	
  a	
  small	
  definition	
  of	
  what	
  these	
  changes	
  incorporate.	
  Given	
  that	
  Wood	
  (1986)	
  

	
  

12	
  

	
  
specified	
  that	
  component	
  complexity	
  change	
  is	
  a	
  less	
  extreme	
  form	
  of	
  change,	
  this	
  type	
  of	
  
adaptive	
  change	
  is	
  located	
  closer	
  to	
  incremental	
  change	
  than	
  coordinative	
  complexity.	
  
However,	
  there	
  is	
  a	
  large	
  gap	
  between	
  the	
  two	
  adaptation	
  change	
  types	
  and	
  metamorphic	
  
change	
  to	
  represent	
  the	
  difference	
  in	
  the	
  definitions,	
  specifically	
  with	
  metamorphic	
  change	
  
being	
  consistently	
  described	
  as	
  a	
  value-­‐	
  or	
  belief-­‐altering	
  change,	
  which	
  is	
  not	
  the	
  focus	
  of	
  
this	
  research	
  effort.	
  
	
  

M

et
a
Ch mo
a n rp h
ge ic
	
  
	
  

	
  

Ra
d
in ica
	
  st l	
  c
b e ra t h a
ha eg ng
vi y/ e	
  
or 	
  
	
  

Q
Co
in uick
cr
Ad mp
ea 	
  ne
ap o n
se ed
ta en
	
  ef 	
  to
tio t	
  
fo 	
  
n	
  
rt
Qu
	
  
C
m ic
oo
od k	
  
A d rd
ify ne
ap i n a
	
  st ed
ta ti v
ra 	
  to
ti o e
te 	
  
n	
   	
  
gy
	
  

T
se ypic
lf-­‐ a
re l	
  u
gu se
la 	
  of
tio 	
  
n	
  

eq Sta
ui b l
lib e	
  
riu
m
	
  

In

Ro
Ch uti
a n ne
ge 	
  
	
  

cr
Ch eme
a n nt
ge al	
  
	
  

Figure	
  1	
  
	
  
Heuristic	
  of	
  a	
  Continuum	
  of	
  Change	
  Types	
  
	
  

	
  

	
  
	
  
	
  
	
  

	
  
	
  
Adaptation,	
  Learning,	
  and	
  Self-­‐Regulation	
  as	
  Dynamic	
  Processes.	
  The	
  concept	
  of	
  
the	
  adaptation	
  process	
  as	
  being	
  a	
  series	
  of	
  re-­‐occurring	
  phases	
  or	
  cycles	
  has	
  roots	
  in	
  some	
  
of	
  the	
  early	
  conceptualizations	
  of	
  team	
  adaptation.	
  Kozlowski	
  and	
  colleagues	
  (1996)	
  
proposed	
  that	
  teams	
  go	
  through	
  several	
  phases	
  as	
  they	
  mature,	
  developing	
  adaptive	
  
capabilities,	
  which	
  then	
  result	
  in	
  the	
  team’s	
  ability	
  to	
  self-­‐regulate	
  in	
  response	
  to	
  
incremental	
  and	
  novel	
  changes.	
  Burke	
  and	
  associates	
  (2006)	
  also	
  use	
  a	
  phase	
  framework	
  to	
  
describe	
  team	
  adaptation	
  with	
  the	
  first	
  phase	
  being	
  situation	
  assessment	
  when	
  a	
  change	
  is	
  
recognized	
  through	
  the	
  constant	
  monitoring	
  of	
  the	
  environment.	
  The	
  second	
  phase	
  
	
  

13	
  

	
  
involves	
  setting	
  a	
  strategy	
  (i.e.,	
  creating	
  goals,	
  assigning	
  roles)	
  to	
  adapt	
  to	
  the	
  change,	
  while	
  
the	
  third	
  phase	
  comprises	
  of	
  executing	
  that	
  plan	
  and	
  monitoring	
  the	
  performance	
  
outcomes	
  based	
  on	
  those	
  initial	
  goals.	
  Finally,	
  the	
  last	
  phase	
  is	
  the	
  reflection	
  on	
  the	
  
adaptive	
  response	
  made	
  in	
  order	
  to	
  understand	
  why	
  the	
  change	
  occurred	
  and	
  to	
  determine	
  
the	
  effectiveness	
  of	
  the	
  team’s	
  response.	
  
These	
  theoretical	
  frameworks	
  of	
  the	
  team	
  adaptation	
  process	
  and	
  the	
  empirical	
  
works	
  of	
  investigating	
  adaptive	
  performance	
  change	
  indicate	
  that	
  the	
  conceptual	
  roots	
  of	
  
the	
  adaptation	
  process	
  are	
  evident	
  in	
  the	
  learning	
  and	
  self-­‐regulation	
  literatures.	
  
Theoretical	
  and	
  empirical	
  evidence	
  from	
  these	
  fields	
  have	
  specified	
  that	
  the	
  cognitive,	
  
motivational,	
  affective	
  and	
  behavioral	
  process	
  mechanisms	
  impact	
  how	
  individuals	
  set	
  
goals,	
  monitor	
  their	
  progress	
  toward	
  them,	
  and	
  adjust	
  their	
  behaviors	
  in	
  light	
  of	
  any	
  
discrepancies	
  (Bandura,	
  1986,	
  1991;	
  Blau,	
  1993;	
  Flavell,	
  1979;	
  Klein,	
  1989;	
  Latham	
  &	
  
Locke,	
  1991).	
  Bell	
  and	
  Kozlowski	
  (2008)	
  provided	
  a	
  link	
  between	
  self-­‐regulation,	
  
adaptation	
  and	
  learning	
  through	
  their	
  use	
  of	
  a	
  self-­‐regulatory	
  perspective	
  when	
  they	
  
describe	
  the	
  learning	
  process	
  that	
  occurred	
  prior	
  to	
  adaptation.	
  	
  
Learning	
  and	
  self-­‐regulation	
  processes	
  are,	
  in	
  and	
  of	
  themselves,	
  dynamic	
  
phenomena.	
  Zimmerman	
  (1989)	
  presented	
  a	
  triadic	
  model	
  of	
  self-­‐regulation	
  where	
  self-­‐
efficacy,	
  goals	
  and	
  cognitions	
  drove	
  the	
  changes	
  in	
  individuals’	
  internal	
  motivation	
  and	
  
external	
  behaviors	
  as	
  they	
  pursued	
  various	
  outcomes	
  of	
  interest.	
  Zimmerman	
  suggests	
  that	
  
the	
  person,	
  behavior,	
  and	
  environment	
  are	
  dynamically	
  and	
  reciprocally	
  related	
  to	
  each	
  
other.	
  The	
  regulatory	
  actions	
  are	
  dependent	
  on	
  the	
  internal	
  state	
  of	
  the	
  individuals,	
  which	
  
impact	
  their	
  behaviors	
  and	
  influence	
  how	
  they	
  operate	
  in	
  the	
  environment,	
  and	
  the	
  
environment	
  subsequently	
  impacts	
  those	
  internal	
  processes	
  and	
  external	
  behaviors.	
  The	
  

	
  

14	
  

	
  
Adaptive	
  Character	
  of	
  Thought	
  (ACT-­‐R)	
  is	
  a	
  theory	
  of	
  learning	
  that	
  suggests	
  there	
  is	
  a	
  
dynamic	
  process	
  that	
  occurs	
  between	
  developing	
  declarative	
  and	
  procedural	
  knowledge	
  
(Anderson,	
  1983).	
  Anderson	
  states	
  that	
  this	
  transition	
  process	
  is	
  knowledge	
  creation.	
  This	
  
creation	
  of	
  new	
  procedural	
  knowledge	
  structures	
  may	
  be	
  critical	
  when	
  individuals	
  are	
  
faced	
  with	
  learning	
  or	
  adaptive	
  situations.	
  In	
  this	
  framework,	
  the	
  novel	
  	
  (or	
  unique)	
  
components	
  of	
  the	
  environment	
  are	
  driving	
  the	
  need	
  to	
  develop	
  new	
  or	
  modified	
  
knowledge	
  structures,	
  but	
  procedural	
  knowledge	
  will	
  eventually	
  become	
  automated	
  
behaviors	
  as	
  the	
  knowledge	
  is	
  utilized	
  over	
  and	
  over.	
  This	
  theory	
  provides	
  conceptual	
  
evidence	
  that	
  there	
  is	
  a	
  dynamic	
  change	
  in	
  the	
  formation	
  and	
  utilization	
  of	
  knowledge	
  
structures.	
  
Some	
  have	
  suggested	
  that	
  learning	
  and	
  adaptation	
  are	
  the	
  same	
  phenomenon	
  (e.g.,	
  
Edmondson,	
  1999;	
  Staddon,	
  1975),	
  and	
  I	
  agree	
  that	
  there	
  are	
  similarities	
  between	
  learning	
  
and	
  adaptation	
  (e.g.,	
  their	
  use	
  of	
  self-­‐regulatory	
  strategies	
  in	
  modifying	
  behaviors	
  in	
  light	
  of	
  
a	
  need	
  to	
  learn	
  new	
  information).	
  However,	
  the	
  main	
  distinction	
  I	
  see	
  between	
  learning	
  
and	
  adaptation	
  is	
  in	
  the	
  need	
  to	
  revise	
  current	
  knowledge	
  structures	
  to	
  meet	
  the	
  demands	
  
of	
  the	
  environment,	
  when	
  engaging	
  in	
  adaptation,	
  versus	
  having	
  to	
  create	
  new	
  structures,	
  
as	
  in	
  Anderson’s	
  explication	
  of	
  learning.	
  As	
  Burke	
  and	
  colleagues	
  aptly	
  stated,	
  “Learning	
  is	
  
an	
  essential	
  but	
  insufficient	
  condition	
  for	
  team	
  adaptation”	
  (Burke	
  et	
  al.,	
  2006,	
  p.	
  1190).	
  
Anderson	
  (1983)	
  discusses	
  learning	
  (i.e.,	
  knowledge	
  creation)	
  as	
  the	
  precursor	
  to	
  
routinized	
  behavior.	
  Therefore,	
  I	
  consider	
  learning	
  as	
  a	
  state	
  that	
  occurs	
  before	
  
performance	
  is	
  stabilized.	
  In	
  this	
  learning	
  environment,	
  individuals	
  are	
  obtaining	
  an	
  
understanding	
  about	
  the	
  relationships	
  that	
  exist	
  between	
  their	
  behaviors	
  and	
  outcomes.	
  
Once	
  these	
  links	
  are	
  created,	
  a	
  routine	
  performance	
  environment	
  is	
  established.	
  This	
  

	
  

15	
  

	
  
typical	
  environment	
  serves	
  a	
  critical	
  role	
  for	
  individuals.	
  It	
  allows	
  them	
  to	
  identify	
  a	
  
pattern	
  of	
  behavior	
  that	
  is	
  effective	
  and	
  a	
  reference	
  point	
  for	
  future	
  performance.	
  Without	
  
this	
  calibration,	
  individuals	
  will	
  be	
  unable	
  to	
  comprehend	
  whether	
  the	
  lack	
  of	
  performance	
  
stability	
  is	
  due	
  to	
  a	
  lack	
  of	
  understanding	
  of	
  the	
  task	
  or	
  a	
  novel	
  change	
  that	
  requires	
  
adaptation.	
  However,	
  after	
  learning	
  occurs	
  and	
  routinization	
  is	
  evident,	
  when	
  a	
  change	
  
occurs	
  individuals	
  can	
  adjust	
  their	
  behaviors	
  to	
  revise	
  or	
  relearn	
  how	
  their	
  actions	
  are	
  
influencing	
  the	
  outcomes.	
  
Self-­‐regulation	
  theory	
  provides	
  a	
  link	
  between	
  all	
  the	
  environments	
  in	
  which	
  
individuals	
  engage.	
  Whether	
  it	
  is	
  learning,	
  routine	
  performance,	
  adaptation,	
  or	
  re-­‐
stabilizing	
  performance,	
  self-­‐regulation	
  appears	
  to	
  be	
  key.	
  Regardless	
  of	
  what	
  environment	
  
individuals	
  are	
  in,	
  regulating	
  cognitions,	
  motivations,	
  and	
  behaviors	
  is	
  essential	
  in	
  
understanding	
  new	
  information,	
  revising	
  current	
  knowledge	
  structures,	
  or	
  simply	
  
performing	
  a	
  task	
  that	
  is	
  currently	
  understood.	
  Therefore,	
  I	
  consider	
  adaptation	
  as	
  
embedded	
  in	
  an	
  ongoing	
  self-­‐regulation	
  process	
  and	
  suggest	
  that	
  self-­‐regulation	
  is	
  the	
  
theoretical	
  driver	
  of	
  the	
  adaptation	
  process,	
  similar	
  to	
  previous	
  work	
  in	
  the	
  adaptation	
  
literature	
  (e.g.,	
  Kozlowski	
  et	
  al.,	
  2009;	
  Tsui	
  &	
  Ashford,	
  1994).	
  
As	
  suggested	
  by	
  this	
  discussion,	
  the	
  dynamics	
  of	
  the	
  adaptation	
  process	
  are	
  a	
  critical	
  
component	
  of	
  this	
  theory.	
  Adaptation	
  is	
  not	
  a	
  static	
  phenomenon.	
  It	
  does	
  not	
  occur	
  at	
  one	
  
singular	
  moment,	
  but	
  is	
  a	
  regulatory	
  process	
  individuals	
  engage	
  in	
  when	
  they	
  are	
  exposed	
  
to	
  a	
  change.	
  Figure	
  2	
  presents	
  a	
  graphical	
  representation	
  of	
  this	
  heuristic,	
  which	
  I	
  will	
  
explain	
  in	
  more	
  depth	
  presently.	
  Although	
  the	
  theory	
  is	
  represented	
  in	
  a	
  static	
  model	
  for	
  
the	
  purpose	
  of	
  clarity,	
  the	
  theory	
  suggests	
  that	
  individuals	
  move	
  through	
  this	
  process	
  many	
  
times	
  as	
  they	
  engage	
  in	
  work	
  environments.	
  	
  

	
  

16	
  

	
  
Figure	
  2	
  
Heuristic	
  Representation	
  of	
  the	
  Adaptation	
  Process	
  

Performance*

Evalua:on*
Learning*
Effort*

Metacogni:on*
Outcome*
Effort*
*
*

Self/*
efficacy*
Goals*

	
  

17	
  

	
  

	
  
Part	
  2:	
  The	
  Sub-­‐Cycles	
  
The	
  second	
  critical	
  element	
  of	
  the	
  definition	
  of	
  adaptation	
  adopted	
  in	
  this	
  research	
  
is	
  the	
  existence	
  of	
  multiple	
  components.	
  Specifically,	
  I	
  suggest	
  that	
  the	
  behaviors	
  of	
  
individuals	
  are	
  modified	
  through	
  engaging	
  in	
  two	
  interwoven	
  cycles:	
  cognitive	
  (the	
  
understanding	
  of	
  the	
  environment	
  and	
  the	
  impact	
  of	
  the	
  adaptive	
  change	
  on	
  behaviors	
  and	
  
outcomes)	
  and	
  motivational	
  (the	
  pursuit	
  of	
  regaining	
  pre-­‐change	
  performance	
  level).	
  The	
  
concept	
  of	
  a	
  two-­‐fold	
  subsystem	
  underlying	
  a	
  single	
  process	
  is	
  not	
  a	
  new	
  innovation	
  (see	
  
Karoly,	
  1993;	
  Klein,	
  1989;	
  Pintrich,	
  2000;	
  Zimmerman,	
  1989	
  for	
  examples).	
  The	
  adaptation	
  
literature	
  has	
  been	
  using	
  the	
  cognitive,	
  motivational,	
  and,	
  at	
  times,	
  affective	
  cycles	
  
described	
  in	
  the	
  self-­‐regulation	
  theory	
  to	
  describe	
  a	
  pre-­‐adaptation	
  learning	
  process	
  for	
  
many	
  years	
  (e.g.,	
  Bell,	
  2002;	
  Bell	
  &	
  Kozlowski,	
  2008;	
  Kozlowski	
  et	
  al.,	
  1999).	
  However,	
  as	
  
the	
  affective	
  cycle	
  has	
  not	
  received	
  support	
  in	
  empirical	
  investigations	
  (e.g.,	
  Bell	
  &	
  
Kozlowski,	
  2008),	
  this	
  pathway	
  will	
  not	
  be	
  discussed	
  in	
  more	
  detail	
  in	
  this	
  study.	
  Bell	
  and	
  
Kozlowski	
  (2008)	
  specify	
  that	
  the	
  cognitive	
  cycle	
  (or	
  pathway,	
  as	
  the	
  authors	
  describe	
  it)	
  
includes	
  the	
  self-­‐regulatory	
  mechanisms	
  of	
  metacognition,	
  evaluation,	
  and	
  knowledge,	
  
while	
  the	
  motivation	
  element	
  involves	
  goal	
  orientation,	
  self-­‐efficacy	
  and	
  intrinsic	
  
motivation.	
  This	
  multiple	
  pathway	
  framework	
  is	
  also	
  supported	
  by	
  some	
  of	
  the	
  early	
  
theoretical	
  conceptualizations	
  of	
  the	
  self-­‐regulation	
  process.	
  In	
  his	
  description	
  of	
  the	
  
triadic	
  model	
  of	
  self-­‐regulation,	
  Zimmerman	
  (1989)	
  discussed	
  the	
  role	
  of	
  cognitions	
  and	
  
motivations	
  in	
  the	
  context	
  of	
  how	
  individuals	
  modify	
  their	
  behaviors	
  in	
  task	
  environments.	
  
He	
  states	
  that	
  cognitive	
  components	
  are	
  critical	
  in	
  the	
  evaluation	
  of	
  strategies	
  and	
  
motivational	
  elements	
  are	
  essential	
  in	
  the	
  drive	
  to	
  perform	
  behaviors.	
  Other	
  researchers	
  
also	
  suggest	
  that	
  there	
  are	
  cognitive	
  and	
  motivational	
  components	
  in	
  the	
  self-­‐regulatory	
  

	
  

18	
  

	
  
process	
  involved	
  in	
  controlling	
  behavior	
  such	
  that	
  individuals	
  must	
  cognitively	
  develop	
  
goals	
  and	
  compare	
  the	
  current	
  state	
  to	
  the	
  desired	
  state	
  as	
  well	
  as	
  be	
  motivated	
  to	
  resolve	
  
any	
  discrepancies	
  found	
  in	
  the	
  comparison	
  (Carver	
  &	
  Scheier,	
  1982;	
  Klein,	
  1989).	
  	
  	
  	
  
Therefore,	
  I	
  describe	
  the	
  cognitive	
  cycle	
  of	
  the	
  adaptation	
  process	
  as	
  those	
  elements	
  
that	
  assist	
  individuals	
  in	
  understanding	
  the	
  impact	
  changes	
  (or	
  environmental	
  elements)	
  
have	
  on	
  current	
  or	
  previous	
  behaviors.	
  Cognitive	
  mechanisms	
  include	
  strategy	
  
development,	
  metacognitive	
  behaviors,	
  and	
  feedback	
  evaluation	
  behaviors	
  (Bell	
  &	
  
Kozlowski,	
  2008).	
  Thus,	
  after	
  a	
  change	
  is	
  introduced,	
  individuals	
  must	
  examine	
  
performance	
  feedback,	
  diagnose	
  feedback,	
  devote	
  effort	
  toward	
  learning	
  new	
  information	
  
to	
  resolve	
  any	
  issues	
  identified	
  in	
  the	
  feedback,	
  create	
  new	
  strategies,	
  and	
  analyze	
  the	
  
effectiveness	
  of	
  strategies	
  employed	
  (engaging	
  in	
  the	
  cognitive	
  cycle	
  of	
  the	
  adaptation	
  
process).	
  These	
  elements	
  are	
  most	
  critical	
  when	
  an	
  adaptive	
  change	
  is	
  introduced,	
  as	
  the	
  
reason	
  for	
  the	
  change	
  in	
  outcomes	
  must	
  be	
  identified	
  and	
  resolved.	
  In	
  a	
  routine	
  
performance	
  environment	
  there	
  is	
  less	
  of	
  a	
  need	
  for	
  evaluating	
  performance	
  and	
  
strategizing	
  behaviors	
  since	
  the	
  relationship	
  between	
  behaviors	
  and	
  outcomes	
  is	
  clear	
  and	
  
understood.	
  However,	
  when	
  a	
  novel	
  situation	
  arises,	
  there	
  is	
  a	
  need	
  to	
  re-­‐evaluate	
  the	
  
effectiveness	
  of	
  the	
  strategies	
  and	
  possibly	
  learn	
  new	
  information	
  to	
  develop	
  better	
  
methods	
  to	
  adapt	
  to	
  the	
  change.	
  Therefore,	
  the	
  cognitive	
  cycle	
  of	
  the	
  adaptation	
  process	
  is	
  
critical	
  in	
  novel	
  environments.	
  
The	
  motivational	
  cycle	
  of	
  the	
  adaptation	
  process,	
  on	
  the	
  other	
  hand,	
  incorporates	
  
elements	
  that	
  capture	
  overall	
  desire	
  to	
  continue	
  in	
  or	
  change	
  behaviors	
  to	
  address	
  the	
  
cognitive	
  strategy	
  one	
  is	
  pursuing.	
  In	
  other	
  words,	
  it	
  is	
  insufficient	
  to	
  simply	
  have	
  a	
  plan	
  for	
  
action;	
  individuals	
  must	
  also	
  be	
  motivated	
  to	
  devote	
  time	
  and	
  energy	
  to	
  achieve	
  the	
  desired	
  

	
  

19	
  

	
  
outcome.	
  For	
  this	
  reason,	
  goals,	
  self-­‐efficacy,	
  and	
  effort	
  are	
  key	
  for	
  this	
  aspect	
  of	
  self-­‐
regulation	
  (Bell	
  &	
  Kozlowski,	
  2008).	
  Therefore,	
  after	
  a	
  change,	
  individuals	
  must	
  set	
  goals	
  to	
  
drive	
  behavior,	
  have	
  a	
  certain	
  level	
  of	
  confidence	
  associated	
  with	
  that	
  goal,	
  and	
  devote	
  
behavioral	
  effort	
  toward	
  achieving	
  that	
  goal	
  (engaging	
  in	
  the	
  motivational	
  cycle	
  of	
  the	
  
adaptation	
  process).	
  Motivational	
  mechanisms	
  are	
  also	
  critical	
  in	
  the	
  adaptation	
  process	
  as	
  
it	
  is	
  necessary	
  for	
  individuals	
  to	
  be	
  motivated	
  to	
  devote	
  effort	
  toward	
  changing	
  cognitions	
  
and	
  behaviors	
  in	
  order	
  to	
  adapt	
  effectively.	
  The	
  motivational	
  cycle	
  is	
  even	
  more	
  important	
  
in	
  performance	
  environments	
  where	
  a	
  change	
  is	
  not	
  present	
  but	
  there	
  is	
  a	
  need	
  to	
  continue	
  
to	
  maintain	
  performance	
  despite	
  the	
  possible	
  monotony	
  a	
  routine	
  environment	
  may	
  bring.	
  
In	
  the	
  following	
  sections	
  I	
  describe,	
  in	
  more	
  depth,	
  the	
  specific	
  dynamics	
  of	
  these	
  cycles.	
  
The	
  Cognitive	
  Cycle.	
  The	
  cognitive	
  component	
  of	
  adaptation	
  process	
  suggests	
  that	
  
individuals	
  must	
  learn	
  the	
  impact	
  of	
  a	
  change	
  and	
  how	
  to	
  cope	
  with	
  it.	
  In	
  the	
  adaptation	
  
literature,	
  researchers	
  have	
  consistently	
  described	
  learning	
  and	
  adaptation	
  as	
  self-­‐
regulation	
  processes	
  containing	
  cognitive	
  elements	
  such	
  as	
  on	
  task	
  cognition	
  (Bell	
  &	
  
Kozlowski,	
  2002b),	
  cognitive	
  flexibility	
  (Griffin	
  &	
  Hesketh,	
  2003;	
  Mumford,	
  Baughman,	
  
Threlfall,	
  Uhlman	
  &	
  Costanza,	
  1993),	
  information	
  sharing	
  (Johnson,	
  Hollenbeck,	
  Humphrey,	
  
Ilgen,	
  Jundt	
  &	
  Meyer,	
  2006),	
  job	
  experience	
  and	
  knowledge	
  breadth	
  (Niessen,	
  Swarowsky	
  &	
  
Leiz,	
  2010;	
  Spiro	
  &	
  Weitz,	
  1990),	
  resource	
  allocation	
  (Porter,	
  Webb	
  &	
  Gogus,	
  2010),	
  
openness	
  to	
  change	
  (Pulakos,	
  Schmitt,	
  Dorsey,	
  Arad,	
  Borman	
  &	
  Hadge,	
  2002),	
  mental	
  
model	
  similarity	
  and	
  accuracy	
  (Randall,	
  Resick,	
  &	
  DeChurch,	
  2011),	
  error	
  use	
  (Woltz,	
  
Gardner	
  &	
  Gyll,	
  2000),	
  and	
  metacognition	
  (e.g.,	
  Bell	
  &	
  Kozlowski,	
  2008;	
  Ivancic	
  &	
  Hesketh,	
  
2000).	
  Clearly,	
  cognitive	
  mechanisms	
  are	
  critical	
  for	
  the	
  self-­‐regulation	
  of	
  behavior	
  and,	
  
although	
  there	
  have	
  been	
  a	
  variety	
  of	
  constructs	
  investigated,	
  it	
  is	
  evident	
  that	
  strategy-­‐

	
  

20	
  

	
  
focused	
  mechanisms	
  are	
  particularly	
  important	
  for	
  studies	
  considering	
  the	
  regulation	
  
process	
  prior	
  to	
  and	
  during	
  adaptation.	
  Specifically,	
  I	
  assert	
  that	
  there	
  are	
  three	
  key	
  
mechanisms	
  involved	
  in	
  the	
  cognitive	
  cycle	
  of	
  the	
  adaptation	
  process:	
  metacognition,	
  
feedback	
  evaluation,	
  and	
  learning-­‐oriented	
  effort.	
  The	
  inner	
  cycle	
  of	
  Figure	
  2	
  shows	
  a	
  
graphical	
  representation	
  of	
  the	
  interdependence	
  of	
  these	
  mechanisms.	
  
Although	
  metacognition	
  has	
  been	
  defined	
  and	
  investigated	
  many	
  ways,	
  Flavell,	
  one	
  
of	
  the	
  leading	
  thinkers	
  on	
  the	
  topic	
  of	
  metacognition,	
  proposed	
  that	
  this	
  phenomenon	
  has	
  
several	
  elements:	
  knowledge,	
  experiences,	
  goals/tasks,	
  and	
  actions/strategies	
  (Flavell,	
  
1979).	
  In	
  this	
  definition,	
  metacognition	
  could	
  entail	
  all	
  of	
  a	
  cognitive	
  self-­‐regulatory	
  
process,	
  but	
  in	
  this	
  study,	
  this	
  overall	
  metacognitive	
  process	
  is	
  being	
  separated	
  into	
  the	
  
components	
  in	
  order	
  to	
  more	
  specifically	
  investigate	
  the	
  impact	
  of	
  cognitive	
  variables	
  (i.e.,	
  
evaluation,	
  learning	
  effort,	
  and	
  strategy	
  development)	
  on	
  performance	
  behavior.	
  Many	
  
researchers	
  describe	
  metacognition	
  as	
  monitoring	
  one’s	
  cognitions	
  and	
  behaviors	
  
associated	
  with	
  those	
  ideas,	
  and	
  re-­‐evaluating	
  whether	
  these	
  behaviors	
  are	
  in	
  line	
  with	
  
one’s	
  strategy	
  (Bell	
  &	
  Kozlowski,	
  2008;	
  Cannon-­‐Bowers,	
  Rhodenizer,	
  Salas	
  &	
  Bowers,	
  1998;	
  
Flavell,	
  1979).	
  Researchers	
  have	
  suggested	
  that	
  metacognition	
  allows	
  for	
  a	
  more	
  rapid	
  
assessment	
  of	
  the	
  environment,	
  understanding	
  of	
  the	
  core	
  elements	
  and	
  better	
  success	
  in	
  
learning	
  (Ford,	
  Smith,	
  Weissbein,	
  Gully	
  &	
  Salas,	
  1998;	
  Ivancic	
  &	
  Hesketh,	
  1999).	
  Therefore,	
  
in	
  this	
  paper,	
  metacognition	
  refers	
  to	
  the	
  actions	
  individuals	
  take	
  to	
  understand	
  the	
  
effectiveness	
  and	
  impact	
  of	
  their	
  strategies	
  on	
  the	
  pursuit	
  of	
  their	
  goals.	
  	
  
Evaluation	
  behaviors	
  are	
  actions	
  devoted	
  to	
  analyzing	
  performance	
  feedback.	
  They	
  
can	
  be	
  as	
  simple	
  as	
  viewing	
  a	
  general	
  performance	
  score	
  from	
  a	
  task	
  simulation	
  or	
  as	
  
complex	
  as	
  seeking	
  360-­‐degree	
  evaluations	
  within	
  an	
  organization.	
  There	
  are	
  several	
  ways	
  

	
  

21	
  

	
  
in	
  which	
  evaluation	
  behaviors	
  are	
  operationalized	
  in	
  the	
  adaptation	
  literature.	
  Some	
  
researchers	
  label	
  it	
  behavioral	
  adaptation,	
  referring	
  to	
  monitoring	
  behaviors	
  that	
  are	
  
conducted	
  as	
  individuals	
  perform	
  (or	
  after	
  a	
  performance	
  period)	
  and	
  these	
  behaviors	
  are	
  
directed	
  toward	
  understanding	
  the	
  impact	
  of	
  a	
  change	
  on	
  performance	
  outcomes	
  (e.g.,	
  
LePine,	
  2003,	
  2005).	
  Other	
  researchers	
  in	
  the	
  adaptation	
  literature	
  operationalize	
  
evaluation	
  activity	
  as	
  time	
  spent	
  viewing	
  performance	
  feedback.	
  Still	
  others	
  consider	
  
evaluation	
  to	
  be	
  a	
  behavioral	
  manifestation	
  of	
  metacognition	
  in	
  that	
  the	
  focus	
  of	
  the	
  
performance	
  evaluation	
  behaviors	
  (i.e.,	
  which	
  parts	
  of	
  the	
  performance	
  information	
  are	
  
being	
  analyzed)	
  presents	
  insight	
  into	
  which	
  strategies	
  are	
  being	
  scrutinized	
  (e.g.,	
  Bell	
  &	
  
Kozlowski,	
  2008;	
  Ford	
  et	
  al.,	
  1998).	
  However,	
  in	
  all	
  of	
  them,	
  the	
  focus	
  is	
  on	
  the	
  cognitive	
  
understanding	
  of	
  the	
  impact	
  of	
  behaviors	
  on	
  performance	
  outcomes.	
  
Finally,	
  learning-­‐oriented	
  effort	
  is	
  a	
  component	
  of	
  effort	
  that	
  is	
  devoted	
  toward	
  
understanding	
  the	
  environment	
  by	
  testing	
  strategies	
  and	
  attaining	
  additional	
  information.	
  
Adaptation	
  researchers	
  state	
  that	
  “effort	
  alone	
  is	
  often	
  not	
  enough	
  to	
  perform	
  well	
  on	
  
difficult	
  and	
  complex	
  tasks.	
  Individuals	
  must	
  also	
  focus	
  their	
  effort	
  on	
  relevant	
  aspects	
  of	
  
the	
  task”	
  (Bell	
  &	
  Kozlowski,	
  2002a;	
  p.	
  278).	
  Fisher	
  and	
  Ford	
  (1998)	
  also	
  proposed	
  that	
  
effort	
  can	
  be	
  operationalized	
  as	
  the	
  number	
  of	
  behaviors	
  devoted	
  to	
  a	
  task,	
  but	
  can	
  also	
  be	
  
thought	
  of	
  as	
  the	
  complexity	
  of	
  the	
  cognitive	
  involvement	
  of	
  the	
  individuals	
  (i.e.,	
  the	
  
amount	
  of	
  cognitive	
  effort	
  is	
  dependent	
  on	
  the	
  type	
  of	
  process	
  chosen:	
  encoding,	
  
organization,	
  or	
  retrieval).	
  Together,	
  these	
  researchers	
  suggest	
  that	
  effort	
  is	
  more	
  than	
  the	
  
sum	
  of	
  behavioral	
  output;	
  effort	
  also	
  has	
  a	
  cognitive	
  component.	
  Some	
  label	
  this	
  form	
  of	
  
effort	
  in	
  adaptive	
  contexts	
  as	
  strategic	
  effort	
  as	
  they	
  differentiated	
  the	
  behaviors	
  in	
  the	
  task	
  
in	
  order	
  to	
  capture	
  effort	
  devoted	
  to	
  aspects	
  of	
  the	
  task	
  that	
  changed	
  (i.e.,	
  strategic	
  effort)	
  

	
  

22	
  

	
  
versus	
  aspects	
  that	
  did	
  not	
  (i.e.,	
  basic	
  effort;	
  see	
  Bell	
  &	
  Kozlowski,	
  2008;	
  Kozlowski	
  et	
  al.,	
  
2001).	
  However,	
  there	
  is	
  one	
  challenge	
  in	
  approaching	
  the	
  cognitively	
  focused	
  effort	
  in	
  
such	
  a	
  manner:	
  “strategic”	
  behaviors	
  conducted	
  in	
  an	
  adaptive	
  environment	
  (where	
  this	
  is	
  
an	
  unknown	
  change)	
  can	
  become	
  “basic”	
  behaviors	
  in	
  a	
  performance	
  environment	
  (where	
  
individuals	
  understand	
  the	
  change	
  and	
  are	
  optimizing	
  a	
  strategy).	
  Therefore,	
  I	
  specifically	
  
label	
  cognitively	
  focused	
  and	
  strategizing	
  behaviors	
  as	
  “learning-­‐oriented	
  effort”	
  as	
  both	
  
entail	
  obtaining	
  additional	
  understanding	
  and	
  information	
  from	
  the	
  environment.	
  Given	
  
this	
  conceptualization,	
  neither	
  the	
  definition	
  nor	
  the	
  operationalization	
  of	
  this	
  type	
  of	
  effort	
  
will	
  depend	
  on	
  the	
  kind	
  of	
  environment	
  in	
  which	
  individuals	
  are	
  engaged.	
  
The	
  Motivational	
  Cycle.	
  It	
  is	
  conceptually	
  insufficient	
  to	
  only	
  theorize	
  about	
  the	
  
cognitive	
  elements	
  involved	
  in	
  self-­‐regulation,	
  regardless	
  of	
  whether	
  the	
  discussion	
  is	
  on	
  
the	
  process	
  occurring	
  before	
  or	
  after	
  a	
  change.	
  Simply	
  investigating	
  the	
  cognitive	
  
components	
  of	
  adaptation	
  without	
  giving	
  pause	
  to	
  consider	
  the	
  impact	
  of	
  motivation	
  on	
  
seeking	
  out	
  information	
  for	
  strategy	
  development	
  or	
  behaviors	
  directed	
  at	
  enhancing	
  
performance	
  would	
  be	
  to	
  only	
  examine	
  one	
  side	
  of	
  a	
  coin	
  and	
  would	
  present	
  an	
  incomplete	
  
picture	
  of	
  the	
  phenomenon.	
  In	
  essence,	
  this	
  follows	
  from	
  the	
  logic	
  that	
  an	
  unmotivated	
  
individual,	
  although	
  perhaps	
  having	
  a	
  brilliant	
  strategy	
  of	
  how	
  to	
  excel	
  in	
  an	
  adaptive	
  
environment,	
  stills	
  need	
  to	
  be	
  motivated	
  and	
  devote	
  effort	
  toward	
  the	
  strategy.	
  Without	
  
such	
  motivated	
  and	
  directed	
  effort,	
  individuals	
  will	
  not	
  effectively	
  adapt	
  to	
  the	
  change.	
  It	
  is	
  
not	
  suggested	
  here	
  that	
  motivational	
  mechanisms	
  serve	
  as	
  moderators	
  of	
  the	
  cognitive	
  
mechanisms,	
  but	
  rather	
  that	
  each	
  must	
  be	
  considered	
  as	
  distinct	
  but	
  integrated	
  parts	
  of	
  the	
  
self-­‐regulatory	
  process.	
  	
  	
  
Several	
  motivational	
  and	
  behavioral	
  mechanisms	
  have	
  been	
  identified	
  and	
  studied	
  

	
  

23	
  

	
  
in	
  the	
  adaptation	
  literature,	
  including	
  effort	
  (DeRue,	
  Hollenbeck,	
  Johnson,	
  Ilgen	
  &	
  Jundt,	
  
2008),	
  emotional	
  stability	
  or	
  reactions	
  (Driskell,	
  Goodwin,	
  Salas	
  &	
  O’Shea,	
  2006;	
  Kozlowski,	
  
et	
  al.,	
  2001;	
  Rosen	
  et	
  al.,	
  2011;	
  Keith	
  &	
  Frese,	
  2005),	
  intrinsic	
  motivation	
  (e.g.,	
  Bell	
  &	
  
Kozlowski,	
  2008,	
  2010),	
  interpersonal	
  processes	
  (LePine,	
  2005),	
  willingness	
  to	
  adapt	
  
(Bröder	
  &	
  Schiffer,	
  2006),	
  internal	
  locus	
  of	
  control	
  (Mumford	
  et	
  al.,	
  1993;	
  Spiro	
  &	
  Weitz,	
  
1990),	
  self-­‐efficacy	
  (e.g.,	
  Griffin	
  &	
  Hesketh,	
  2003,	
  2004,	
  2005;	
  Bell	
  &	
  Kozlowski,	
  2002a,	
  
2002b,	
  2008,	
  2010),	
  and	
  goals	
  (e.g.,	
  Burke	
  et	
  al.,	
  2006;	
  Chen	
  et	
  al.,	
  2005).	
  	
  As	
  self-­‐efficacy,	
  
goals,	
  and	
  effort	
  have	
  been	
  consistently	
  identified	
  as	
  essential	
  mechanisms	
  in	
  the	
  self-­‐
regulation	
  process,	
  I	
  expect	
  that	
  these	
  three	
  mechanisms	
  are	
  the	
  key	
  motivational	
  
mechanisms	
  involved	
  in	
  the	
  adaptation	
  process.	
  See	
  the	
  outer	
  cycle	
  of	
  Figure	
  2	
  for	
  a	
  
representation	
  of	
  the	
  interrelationships	
  between	
  these	
  mechanisms.	
  
Goals	
  are	
  critical	
  for	
  regulatory	
  behavior	
  as	
  they	
  set	
  the	
  standard	
  for	
  subsequent	
  
comparisons	
  (Bandura,	
  1991;	
  Karoly,	
  1993).	
  They	
  provide	
  a	
  reference	
  point	
  from	
  which	
  
discrepancies	
  can	
  be	
  detected,	
  evaluated,	
  and	
  corrected.	
  As	
  Latham	
  and	
  Locke	
  (1991)	
  
suggest,	
  “[Goal	
  setting	
  theory]	
  states	
  that	
  the	
  simplest	
  and	
  most	
  direct	
  motivational	
  
explanation	
  of	
  why	
  some	
  people	
  perform	
  better	
  than	
  others	
  is	
  because	
  they	
  have	
  different	
  
performance	
  goals”	
  (p.213).	
  	
  
Self-­‐efficacy	
  is	
  the	
  well-­‐studied	
  self-­‐regulatory	
  mechanism,	
  which	
  is	
  defined	
  as	
  a	
  
belief	
  in	
  one’s	
  ability	
  to	
  control	
  goals,	
  effort,	
  and	
  performance	
  (Bandura,	
  1991).	
  	
  It	
  has	
  been	
  
found	
  to	
  impact	
  the	
  difficulty	
  of	
  the	
  goal	
  level	
  set,	
  the	
  commitment	
  and	
  effort	
  devoted	
  to	
  
that	
  goal,	
  the	
  performance	
  of	
  individuals	
  during	
  learning	
  and	
  the	
  performance	
  of	
  
individuals	
  after	
  an	
  adaptive	
  event	
  (Bandura,	
  1991;	
  Bell	
  &	
  Kozlowski,	
  2008;	
  Kozlowski,	
  
Gully,	
  et	
  al.,	
  2001;	
  Latham	
  &	
  Locke,	
  1991)	
  and	
  can	
  therefore	
  be	
  considered	
  a	
  vital	
  

	
  

24	
  

	
  
component	
  of	
  a	
  motivational	
  self-­‐regulatory	
  cycle.	
  
Effort	
  is	
  broadly	
  defined	
  as	
  behaviors	
  directed	
  at	
  completing	
  a	
  task	
  (Blau,	
  1993).	
  
Effort	
  has	
  been	
  operationalized	
  through	
  the	
  amount	
  of	
  time	
  spent	
  on	
  task	
  (e.g.,	
  Fisher	
  &	
  
Ford,	
  1998),	
  objective	
  assessments	
  of	
  behavioral	
  effort	
  (e.g.,	
  Bell	
  &	
  Kozlowski,	
  2008),	
  and	
  
self-­‐reports	
  of	
  effort	
  (e.g.,	
  “How	
  hard	
  were	
  you	
  trying	
  just	
  before	
  the	
  screen	
  froze?”	
  p.	
  263;	
  
Yeo	
  &	
  Neal,	
  2004).	
  However,	
  effort	
  has	
  several	
  components	
  as	
  discussed	
  above.	
  One	
  
component	
  of	
  effort	
  focuses	
  on	
  how	
  well	
  or	
  how	
  smart	
  an	
  individual	
  is	
  working	
  (i.e.,	
  
learning-­‐oriented	
  effort),	
  whereas	
  the	
  other	
  component	
  is	
  on	
  how	
  hard	
  an	
  individual	
  is	
  
working	
  (i.e.,	
  what	
  I	
  label	
  “outcome-­‐oriented	
  effort”).	
  Outcome-­‐oriented	
  effort	
  is	
  therefore	
  
a	
  description	
  of	
  the	
  effort	
  individuals	
  specifically	
  devote	
  to	
  achieving	
  a	
  desired	
  
performance	
  level.	
  Cognitive,	
  learning-­‐oriented	
  effort	
  may	
  impact	
  the	
  specific	
  behaviors	
  of	
  
the	
  outcome-­‐oriented	
  effort.	
  For	
  instance,	
  certain	
  effort	
  behaviors	
  may	
  be	
  more	
  strategic	
  in	
  
nature	
  in	
  an	
  adaptation	
  environment	
  when	
  strategies	
  are	
  being	
  tested	
  (i.e.,	
  labeled	
  
“strategic	
  effort”	
  in	
  previous	
  adaptation	
  research;	
  Bell	
  &	
  Kozlowski,	
  2008),	
  but	
  if	
  these	
  
behaviors	
  are	
  determined	
  to	
  be	
  effective,	
  they	
  will	
  continue	
  to	
  be	
  executed	
  in	
  performance	
  
environments	
  even	
  though	
  they	
  are	
  no	
  longer	
  strategic	
  in	
  nature.	
  In	
  order	
  to	
  maintain	
  
consistency	
  in	
  the	
  meaning	
  of	
  effort	
  across	
  environments,	
  I	
  label	
  the	
  effort	
  devoted	
  to	
  
performance	
  outcomes,	
  whether	
  strategic	
  or	
  not,	
  “outcome-­‐oriented”	
  effort.	
  	
  Although	
  this	
  
distinction	
  may	
  seem	
  nuanced,	
  since	
  I	
  theorize	
  that	
  the	
  adaptation	
  process	
  is	
  cyclic	
  and	
  
dynamic	
  over	
  time,	
  it	
  is	
  necessary	
  that	
  the	
  definitions	
  of	
  the	
  mechanisms	
  involved	
  are	
  not	
  
dependent	
  on	
  the	
  type	
  of	
  environment.	
  
Behavioral	
  Mechanisms.	
  As	
  is	
  evident	
  in	
  the	
  discussions	
  above,	
  behaviors	
  play	
  a	
  
unique	
  role	
  in	
  self-­‐regulation	
  as	
  they	
  provide	
  external	
  observational	
  insights	
  into	
  the	
  

	
  

25	
  

	
  
internal	
  regulatory	
  process	
  of	
  individuals	
  (e.g.,	
  changes	
  in	
  cognition	
  or	
  motivation	
  that	
  
would	
  otherwise	
  remain	
  hidden	
  phenomena).	
  Thus,	
  behavioral	
  mechanisms	
  are	
  inherently	
  
tied	
  to	
  the	
  cognitive	
  and	
  motivational	
  cycles	
  of	
  adaptation.	
  Specifically,	
  with	
  regard	
  to	
  
cognition,	
  the	
  emphasis	
  is:	
  how	
  well	
  did	
  the	
  individuals	
  perform?	
  The	
  behavioral	
  
manifestation	
  of	
  this	
  is	
  in	
  feedback	
  evaluation	
  and	
  effort	
  devoted	
  to	
  understanding	
  the	
  
effectiveness	
  of	
  strategies	
  used.	
  This	
  emphasis	
  contrasts	
  with	
  the	
  foci	
  of	
  the	
  motivational	
  
pathways	
  behaviors,	
  which	
  is:	
  how	
  much	
  did	
  individuals	
  do?	
  One	
  way	
  of	
  distinguishing	
  
these	
  paths	
  is	
  by	
  differentiating	
  the	
  meaning	
  of	
  effort	
  within	
  both	
  of	
  these	
  paths.	
  Overall,	
  
whereas	
  the	
  motivational	
  path	
  is	
  concerned	
  with	
  working	
  harder,	
  the	
  cognitive	
  path	
  is	
  
focused	
  on	
  working	
  smarter.	
  
	
  
Two	
  Orders	
  of	
  Change	
  
The	
  following	
  sections	
  of	
  this	
  paper	
  will	
  specify	
  the	
  dynamics	
  of	
  the	
  adaptation	
  
process.	
  Both	
  Powers	
  (1973)	
  and	
  Carver	
  and	
  Scheier	
  (1982)	
  suggest	
  that	
  dynamics	
  can	
  be	
  
investigated	
  through	
  examining	
  change	
  in	
  two	
  ways.	
  They	
  state	
  that	
  first	
  order	
  changes	
  are	
  
evident	
  in	
  the	
  changing	
  trajectories	
  of	
  each	
  variable.	
  Trajectories	
  allow	
  for	
  insight	
  into	
  how	
  
the	
  levels	
  of	
  the	
  variables	
  are	
  impacted	
  by	
  the	
  environment	
  and	
  by	
  time.	
  Therefore,	
  the	
  
discussion	
  of	
  the	
  dynamics	
  of	
  the	
  adaptation	
  process	
  will	
  begin	
  at	
  this	
  level.	
  Powers	
  (1973)	
  
and	
  Carver	
  and	
  Scheier	
  (1982)	
  also	
  state	
  that	
  second	
  order	
  changes	
  are	
  observed	
  through	
  
changes	
  in	
  the	
  bivariate	
  relationships	
  of	
  the	
  variables.	
  Thus,	
  once	
  first	
  order	
  change	
  
trajectories	
  are	
  discussed,	
  the	
  second	
  order	
  relationship	
  changes	
  will	
  be	
  explicated,	
  
specifically	
  describing	
  the	
  direction	
  and	
  strength	
  of	
  the	
  relationships	
  of	
  the	
  adaptation	
  
process	
  mechanisms	
  as	
  seen	
  in	
  Figure	
  2.	
  

	
  

26	
  

	
  
FIRST	
  ORDER	
  CHANGES:	
  TRAJECTORIES	
  
Since	
  level,	
  or	
  trajectory,	
  changes	
  in	
  variables	
  are	
  considered	
  first	
  order	
  changes	
  
(Powers,	
  1973),	
  I	
  will	
  begin	
  my	
  explication	
  of	
  the	
  adaptation	
  processes	
  with	
  a	
  description	
  
of	
  the	
  expected	
  level	
  and	
  slopes	
  of	
  the	
  variables	
  in	
  these	
  processes	
  as	
  determined	
  by	
  the	
  
environment	
  in	
  which	
  they	
  are	
  engaged.	
  Examining	
  the	
  trajectories	
  allows	
  for	
  developing	
  
an	
  understanding	
  of	
  how	
  the	
  variables	
  change.	
  From	
  this	
  knowledge,	
  one	
  can	
  then	
  ask	
  why	
  
the	
  variables	
  change	
  in	
  the	
  manner	
  they	
  do	
  and	
  what	
  determines	
  the	
  trajectory	
  changes.	
  
The	
  ‘why’	
  is	
  the	
  second	
  order	
  change;	
  I	
  will	
  come	
  back	
  to	
  this	
  thought	
  in	
  the	
  following	
  
section.	
  	
  
Prior	
  to	
  specifying	
  the	
  trajectory	
  changes	
  of	
  each	
  mechanism	
  within	
  the	
  two	
  cycles	
  
of	
  the	
  adaptation	
  process,	
  it	
  is	
  first	
  necessary	
  to	
  investigate	
  the	
  demands	
  of	
  the	
  
environments	
  that	
  the	
  process	
  must	
  adjust	
  to.	
  As	
  referred	
  to	
  earlier,	
  I	
  suggest	
  that	
  
individuals	
  face	
  two	
  types	
  of	
  environments:	
  adaptive	
  (i.e.,	
  novel)	
  and	
  performance	
  (i.e.,	
  
typical	
  or	
  routine)	
  environments.	
  For	
  example,	
  an	
  individual	
  working	
  on	
  his	
  job	
  is	
  typically	
  
performing	
  his	
  tasks	
  with	
  high	
  levels	
  of	
  routine	
  and	
  performance	
  is	
  very	
  stable	
  –	
  this	
  is	
  the	
  
first	
  situation	
  where	
  a	
  performance	
  environment	
  is	
  evident.	
  Then,	
  for	
  instance,	
  this	
  
individual	
  is	
  given	
  the	
  work	
  of	
  a	
  coworker	
  who	
  went	
  on	
  vacation	
  and	
  so	
  his	
  workload	
  has	
  
doubled.	
  This	
  change	
  has	
  presented	
  the	
  employee	
  with	
  an	
  adaptive	
  environment	
  where	
  
performance	
  is,	
  at	
  least	
  temporarily,	
  very	
  unstable	
  as	
  the	
  ramifications	
  of	
  such	
  an	
  increase	
  
in	
  workload	
  may	
  require	
  him	
  to	
  re-­‐prioritize	
  and	
  re-­‐strategize	
  his	
  other	
  projects.	
  Finally,	
  
once	
  he	
  obtains	
  an	
  understanding	
  of	
  how	
  he	
  should	
  adapt	
  to	
  this	
  change	
  and	
  performance	
  
begins	
  to	
  increase,	
  the	
  individual	
  will	
  re-­‐enter	
  a	
  performance	
  environment	
  where	
  
performance	
  levels	
  begin	
  to	
  stabilize	
  again.	
  To	
  summarize,	
  individuals	
  begin	
  in	
  a	
  stable	
  

	
  

27	
  

	
  
performance	
  environment,	
  shift	
  to	
  an	
  unstable	
  adaptive	
  environment	
  when	
  a	
  change	
  
occurs,	
  and	
  then	
  transition	
  into	
  a	
  second	
  performance	
  environment	
  where	
  performance	
  is	
  
re-­‐stabilizing.	
  Since	
  the	
  dynamics	
  of	
  the	
  first	
  performance	
  environment	
  are	
  not	
  expected	
  to	
  
be	
  as	
  interesting	
  as	
  those	
  in	
  the	
  second	
  performance	
  environment	
  (given	
  the	
  lack	
  of	
  
variability	
  likely	
  to	
  exist	
  in	
  the	
  first	
  stable	
  situation)	
  I	
  will	
  devote	
  my	
  attention	
  to	
  the	
  re-­‐
stabilizing	
  performance	
  environment	
  that	
  occurs	
  after	
  adaptation.	
  	
  
	
  
The	
  Trajectory	
  of	
  Performance	
  
Based	
  on	
  the	
  above	
  description,	
  performance	
  change	
  is	
  a	
  key	
  indicator	
  of	
  a	
  shift	
  in	
  
to	
  and	
  out	
  of	
  an	
  adaptive	
  environment.	
  The	
  adaptation	
  literature	
  has	
  oftentimes	
  defined	
  
adaptation	
  as	
  the	
  change	
  in	
  performance	
  from	
  a	
  routine	
  to	
  a	
  changed	
  setting	
  (e.g.,	
  
Kozlowski	
  et	
  al.,	
  1999;	
  LePine,	
  2003).	
  Adaptive	
  individuals	
  are	
  those	
  who	
  did	
  not	
  have	
  as	
  
large	
  of	
  a	
  decrease	
  in	
  performance	
  when	
  faced	
  with	
  a	
  change.	
  Researchers	
  in	
  the	
  
adaptation	
  literature	
  assert	
  that	
  individuals	
  likely	
  engage	
  in	
  a	
  process	
  of	
  regulating	
  
behavior	
  in	
  order	
  to	
  effectively	
  adapt,	
  but	
  they	
  have	
  not	
  examined	
  that	
  process	
  after	
  a	
  
change	
  is	
  introduced.	
  However,	
  simply	
  because	
  the	
  empirical	
  research	
  has	
  not	
  yet	
  
advanced	
  to	
  that	
  point,	
  it	
  does	
  not	
  mean	
  that	
  currently	
  existing	
  theories	
  are	
  not	
  capable	
  of	
  
informing	
  the	
  situation.	
  
Control	
  theory	
  suggests	
  that	
  changes	
  are	
  detected	
  through	
  the	
  recognition	
  of	
  a	
  large	
  
gap	
  between	
  desired	
  performance	
  level	
  and	
  actual	
  performance	
  level.	
  Carver	
  and	
  Scheier	
  
(1982)	
  link	
  control	
  theory	
  to	
  self-­‐regulatory	
  theory	
  in	
  their	
  extension	
  of	
  Wiener’s	
  (1948)	
  
cybernetic	
  framework.	
  The	
  key	
  principle	
  of	
  control	
  theory	
  is	
  the	
  influence	
  of	
  negative	
  
feedback	
  loops	
  in	
  controlling	
  behaviors.	
  Carver	
  and	
  Scheier	
  (1982)	
  suggest	
  that	
  when	
  

	
  

28	
  

	
  
individuals	
  are	
  faced	
  with	
  a	
  large	
  goal-­‐feedback	
  gap,	
  they	
  engage	
  in	
  cognitively	
  focused	
  
behaviors	
  that	
  are	
  devoted	
  to	
  “discrepancy	
  reduction”.	
  	
  In	
  this	
  cognitive	
  cycle,	
  as	
  in	
  my	
  
model	
  (see	
  Figure	
  2),	
  individuals	
  are	
  focused	
  on	
  examining	
  feedback	
  in	
  order	
  to	
  evaluate	
  
the	
  source	
  of	
  the	
  change,	
  to	
  obtain	
  more	
  information	
  about	
  the	
  environment,	
  and	
  finally	
  to	
  
develop	
  and	
  test	
  new	
  strategies	
  to	
  cope	
  with	
  the	
  change.	
  Carver	
  and	
  Scheier	
  (1982)	
  also	
  
propose	
  that	
  there	
  is	
  a	
  motivational	
  component	
  involved	
  in	
  that	
  individuals	
  must	
  
determine	
  the	
  “expectancy-­‐assessment”	
  of	
  behaviors,	
  which	
  is	
  the	
  extent	
  to	
  which	
  
individuals	
  expect	
  their	
  behaviors	
  to	
  be	
  able	
  to	
  correct	
  the	
  discrepancy	
  in	
  performance.	
  
Therefore,	
  a	
  motivational	
  process	
  is	
  also	
  activated	
  by	
  a	
  performance	
  change,	
  similar	
  to	
  the	
  
theorized	
  model	
  of	
  this	
  research	
  endeavor	
  (see	
  Figure	
  2).	
  	
  
In	
  addition	
  to	
  defining	
  effective	
  adaptation	
  as	
  higher	
  performance	
  at	
  one	
  time	
  point	
  
after	
  a	
  change,	
  researchers	
  have	
  also	
  investigated	
  the	
  longitudinal	
  effects	
  of	
  change	
  
through	
  examining	
  performance	
  trajectories	
  before	
  and	
  after	
  a	
  change.	
  Through	
  this	
  line	
  of	
  
research,	
  they	
  have	
  established	
  that	
  performance	
  not	
  only	
  significantly	
  decreases	
  with	
  a	
  
change,	
  but	
  more	
  adaptive	
  individuals	
  will	
  increase	
  their	
  performance	
  more	
  rapidly	
  in	
  an	
  
adaptive	
  environment	
  (e.g.,	
  Lang	
  &	
  Bliese,	
  2009;	
  LePine,	
  2003,	
  2005).	
  Therefore,	
  it	
  is	
  
expected	
  that	
  after	
  a	
  change	
  is	
  introduced,	
  individuals	
  will	
  initially	
  have	
  very	
  low	
  
performance,	
  but	
  this	
  will	
  rapidly	
  increase	
  as	
  they	
  engage	
  in	
  the	
  adaptation	
  process	
  (which	
  
I	
  will	
  describe	
  in	
  more	
  detail	
  below).	
  Figure	
  3	
  presents	
  a	
  graphical	
  representation	
  of	
  the	
  
expected	
  trajectories	
  of	
  performance	
  and	
  all	
  process	
  mechanisms	
  in	
  both	
  the	
  adaptation	
  
and	
  performance	
  environments.	
  	
  
Hypothesis	
  1a:	
  In	
  the	
  adaptation	
  environment,	
  performance	
  will	
  have	
  a	
  low	
  
intercept	
  and	
  a	
  strong	
  positive	
  slope.	
  

	
  

29	
  

	
  
Performance	
  environments,	
  on	
  the	
  other	
  hand,	
  are	
  characterized	
  by	
  efficiency,	
  
productivity	
  and,	
  eventually,	
  automaticity	
  (Anderson,	
  1983;	
  March,	
  1991).	
  Individuals	
  
transition	
  into	
  this	
  type	
  of	
  environment	
  when	
  they	
  begin	
  to	
  see	
  improvements	
  in	
  
performance,	
  and	
  their	
  behaviors	
  begin	
  to	
  stabilize.	
  Unlike	
  in	
  adaptive	
  environments,	
  
individuals	
  will	
  not	
  be	
  as	
  confused	
  about	
  what	
  behaviors	
  lead	
  to	
  what	
  outcomes.	
  Extending	
  
control	
  theory	
  to	
  describe	
  the	
  rationale	
  of	
  why	
  this	
  is	
  the	
  case,	
  some	
  original	
  work	
  by	
  
Carver	
  and	
  Scheier	
  (1982)	
  would	
  suggest	
  that	
  once	
  the	
  reason	
  behind	
  a	
  discrepancy	
  
between	
  a	
  goal	
  and	
  performance	
  feedback	
  is	
  understood,	
  the	
  focus	
  shifts	
  to	
  performing	
  the	
  
behaviors	
  that	
  will	
  correct	
  this	
  change.	
  This	
  point	
  is	
  determined	
  by	
  either	
  the	
  dissipation	
  of	
  
the	
  gap	
  or	
  when	
  the	
  rate	
  at	
  which	
  it	
  is	
  decreasing	
  is	
  at	
  an	
  acceptable	
  pace	
  so	
  that	
  the	
  
individual	
  is	
  confident	
  that	
  further	
  adaptation	
  is	
  no	
  longer	
  required.	
  For	
  this	
  reason,	
  it	
  is	
  
expected	
  that	
  when	
  individuals	
  shift	
  to	
  a	
  performance	
  environment,	
  performance	
  will	
  be	
  at	
  
a	
  moderately	
  high	
  level	
  and	
  will	
  increase	
  as	
  more	
  focus	
  is	
  devoted	
  to	
  executing	
  the	
  new	
  
strategy	
  rather	
  than	
  continuing	
  to	
  strategize	
  and	
  determine	
  how	
  to	
  adapt.	
  This	
  increase	
  is	
  
based	
  on	
  self-­‐regulation	
  and	
  learning	
  theories,	
  which	
  suggest	
  that	
  individuals	
  increase	
  in	
  
performance	
  as	
  they	
  practice	
  (Bröder	
  &	
  Shiffer,	
  2006;	
  Kozlowski,	
  et	
  al.,	
  2001).	
  However,	
  
given	
  that	
  it	
  is	
  unreasonable	
  for	
  individuals	
  to	
  consistently	
  increase	
  in	
  performance	
  levels	
  
indefinitely,	
  the	
  rate	
  of	
  growth	
  will	
  slow,	
  until	
  they	
  reach	
  a	
  plateau,	
  or	
  a	
  point	
  where	
  
performance	
  will	
  remain	
  relatively	
  stable	
  as	
  they	
  engage	
  in	
  the	
  performance	
  environment.	
  
Hypothesis	
  1b:	
  In	
  the	
  performance	
  environment,	
  performance	
  will	
  have	
  a	
  
high	
  intercept	
  and	
  a	
  weak	
  positive	
  slope.	
  

	
  

30	
  

	
  
Figure	
  3	
  
Example	
  Trajectories	
  of	
  Key	
  Mechanisms	
  Across	
  the	
  Adaptation	
  and	
  Performance	
  Environment	
  
!"#$%&'()*&(+

!"

#"

$"

%"

&"

'"

("

)"

Performance(Environment(

*" !+" !!" !#" !$" !%" !&" !'" !(" !)" !*" #+"

!+"
*"
)"
("
'"
&"
%"
$"
#"
!"

P"h"a"s"e""""S"h"i"f"t"

Adap%ve(Environment(

!"#$%"&'(')*

P"h"a"s"e""""S"h"i"f"t"

!+"
*"
)"
("
'"
&"
%"
$"
#"
!"

Adap%ve(Environment(

!"

#"

$"

%"

&"

'"

("

)"

!"#$%#&'()

!"

#"

$"

%"

&"

'"

("

)"

Performance(Environment(

*" !+" !!" !#" !$" !%" !&" !'" !(" !)" !*" #+"

!+"
*"
)"
("
'"
&"
%"
$"
#"
!"

Adap%ve(Environment(

!"

#"

$"

%"

&"

'"

("

#"

$"

%"

&"

'"

("

)"

*" !+" !!" !#" !$" !%" !&" !'" !(" !)" !*" #+"

!+"
*"
)"
("
'"
&"
%"
$"
#"
!"

Performance(Environment(

Adap%ve(Environment(

!"

*" !+" !!" !#" !$" !%" !&" !'" !(" !)" !*" #+"

P"h"a"s"e""""S"h"i"f"t"

!"

P"h"a"s"e""""S"h"i"f"t"

Adap%ve(Environment(

)"

#"

$"

%"

&"

'"

("

)"

Ra(ng&

P"h"a"s"e""""S"h"i"f"t"

Adap%ve(Environment(

!"

	
  

#"

$"

%"

&"

'"

("

)"

Performance(Environment(

*" !+" !!" !#" !$" !%" !&" !'" !(" !)" !*" #+"

Trial&

31	
  

Performance(Environment(

*" !+" !!" !#" !$" !%" !&" !'" !(" !)" !*" #+"

!"#$%#&'()"*
!+"
*"
)"
("
'"
&"
%"
$"
#"
!"

Performance(Environment(

!"#$%&'()*%+#(

!"#$%&%'()*+$,(
!+"
*"
)"
("
'"
&"
%"
$"
#"
!"

P"h"a"s"e""""S"h"i"f"t"

Adap%ve(Environment(

*" !+" !!" !#" !$" !%" !&" !'" !(" !)" !*" #+"

!"#$%&

P"h"a"s"e""""S"h"i"f"t"

!+"
*"
)"
("
'"
&"
%"
$"
#"
!"

Performance(Environment(

	
  
The	
  Driver	
  of	
  Performance	
  Trajectory	
  Changes	
  	
  
As	
  mentioned	
  earlier,	
  control	
  theory	
  suggests	
  that	
  understanding	
  the	
  source	
  and	
  
impact	
  of	
  a	
  change	
  is	
  vital	
  to	
  adaptation.	
  Theorists	
  suggest	
  that	
  adaptation	
  is	
  needed	
  where	
  
there	
  is	
  a	
  large	
  discrepancy	
  between	
  the	
  original	
  goals	
  made	
  and	
  the	
  performance	
  feedback	
  
received.	
  When	
  such	
  a	
  situation	
  arises,	
  a	
  discrepancy-­‐reduction	
  process	
  is	
  initiated	
  (here,	
  it	
  
is	
  called	
  an	
  adaptation	
  process)	
  where	
  individuals	
  engage	
  in	
  evaluation,	
  re-­‐strategizing,	
  
and	
  testing	
  strategies	
  in	
  order	
  to	
  determine	
  an	
  effective	
  solution.	
  A	
  failure	
  to	
  engage	
  in	
  an	
  
adaptation	
  process	
  would	
  likely	
  result	
  in	
  optimizing	
  an	
  ineffective	
  strategy,	
  resulting	
  in	
  
decreased	
  performance.	
  This	
  is	
  consistent	
  with	
  claims	
  made	
  by	
  March	
  (1991)	
  when	
  he	
  
discussed	
  the	
  need	
  for	
  organizations	
  to	
  engage	
  in	
  learning	
  and	
  strategy	
  behaviors	
  when	
  
faced	
  with	
  a	
  change	
  rather	
  than	
  immediately	
  choosing	
  a	
  solution	
  and	
  exploiting	
  it.	
  
However,	
  in	
  both	
  theories	
  it	
  is	
  clear	
  that	
  once	
  individuals	
  choose	
  a	
  strategy	
  and	
  begin	
  to	
  
exploit	
  it,	
  they	
  transition	
  from	
  an	
  adaptation-­‐focused	
  environment	
  to	
  a	
  performance-­‐
execution	
  environment.	
  	
  
Hypothesis	
  2:	
  Individuals	
  shift	
  from	
  an	
  adaptive	
  to	
  a	
  performance	
  
environment	
  when	
  a	
  strategy	
  is	
  chosen.	
  
	
  
The	
  speed	
  at	
  which	
  individuals	
  are	
  able	
  to	
  identify	
  the	
  source	
  of	
  the	
  change,	
  develop	
  
a	
  strategy	
  to	
  deal	
  with	
  it,	
  and	
  execute	
  that	
  strategy	
  will	
  be	
  different	
  based	
  on	
  the	
  type	
  of	
  
complexity	
  introduced,	
  given	
  the	
  differences	
  Wood	
  (1986)	
  describes	
  in	
  his	
  theory.	
  This	
  will	
  
impact	
  the	
  rate	
  that	
  individuals	
  switch	
  from	
  an	
  adaptive	
  environment	
  to	
  a	
  performance	
  
environment.	
  Since	
  dynamic	
  complexity	
  change	
  is	
  a	
  complex	
  combination	
  and	
  extension	
  of	
  
component	
  and	
  coordinative	
  change,	
  this	
  type	
  will	
  not	
  be	
  a	
  focus	
  of	
  this	
  research	
  endeavor.	
  
As	
  component	
  complexity	
  is	
  the	
  simplest	
  type	
  of	
  change	
  Wood	
  discusses,	
  it	
  is	
  
	
  

32	
  

	
  
expected	
  that	
  individuals	
  will	
  more	
  rapidly	
  identify	
  the	
  source	
  of	
  the	
  change	
  and	
  a	
  
plausible	
  solution	
  for	
  it	
  (Wood,	
  1986).	
  This	
  would	
  result	
  in	
  a	
  faster	
  shift	
  from	
  an	
  adaptive	
  
environment,	
  where	
  individuals	
  are	
  searching	
  to	
  understand	
  the	
  impact	
  of	
  their	
  behaviors	
  
in	
  the	
  new	
  situation,	
  to	
  a	
  performance	
  environment,	
  where	
  individuals	
  are	
  executing	
  
behaviors	
  based	
  on	
  the	
  strategy	
  they	
  chose	
  to	
  deal	
  with	
  the	
  change.	
  Coordinative	
  
complexity	
  change,	
  on	
  the	
  other	
  hand,	
  is	
  not	
  only	
  more	
  difficult	
  to	
  detect,	
  but	
  it	
  is	
  also	
  more	
  
challenging	
  to	
  understand	
  what	
  an	
  effective	
  new	
  strategy	
  would	
  be,	
  given	
  the	
  more	
  
nuanced	
  nature	
  of	
  this	
  type	
  of	
  change	
  (Wood,	
  1986).	
  This	
  would	
  result	
  in	
  a	
  slower	
  shift	
  
from	
  an	
  adaptive	
  environment	
  to	
  a	
  performance	
  environment.	
  
Hypothesis	
  3:	
  Individuals	
  exposed	
  to	
  a	
  component	
  complexity	
  change	
  will	
  
shift	
  from	
  an	
  adaptive	
  environment	
  to	
  a	
  performance	
  environment	
  faster	
  
than	
  individuals	
  exposed	
  to	
  a	
  coordinative	
  complexity	
  change.	
  
	
  
The	
  Trajectories	
  of	
  the	
  Process	
  Mechanisms	
  
	
  

Given	
  that	
  individuals	
  engage	
  in	
  self-­‐regulatory	
  behaviors	
  in	
  both	
  adaptation	
  and	
  

performance	
  environments,	
  the	
  sections	
  below	
  will	
  specify	
  the	
  expected	
  trajectories	
  of	
  the	
  
cognitive	
  and	
  motivational	
  mechanisms	
  in	
  both	
  settings.	
  As	
  discussed	
  earlier,	
  the	
  
mechanisms	
  involved	
  in	
  the	
  cognitive	
  cycle	
  of	
  the	
  adaptation	
  process	
  (metacognition,	
  
feedback	
  evaluation,	
  and	
  learning-­‐oriented	
  effort)	
  are	
  primarily	
  concerned	
  with	
  
understanding	
  the	
  reason	
  for	
  the	
  change	
  and	
  devoting	
  effort	
  toward	
  the	
  best	
  solution.	
  
Therefore,	
  these	
  mechanisms	
  are	
  the	
  primary	
  ones	
  activated	
  in	
  an	
  adaptation	
  environment.	
  
However,	
  once	
  individuals	
  transition	
  to	
  a	
  performance	
  environment,	
  where	
  a	
  strategy	
  has	
  
been	
  chosen	
  to	
  handle	
  the	
  change,	
  the	
  motivational	
  cycle	
  (self-­‐efficacy,	
  goals,	
  and	
  outcome-­‐

	
  

33	
  

	
  
oriented	
  effort)	
  will	
  be	
  the	
  source	
  of	
  most	
  influential	
  process	
  mechanisms.	
  Both	
  cycles	
  are	
  
critical	
  in	
  both	
  environments,	
  but	
  the	
  trajectories	
  they	
  follow	
  will	
  differ	
  based	
  on	
  the	
  
environment	
  in	
  which	
  the	
  individuals	
  are	
  engaged.	
  The	
  following	
  sections	
  will	
  specify	
  how.	
  
Longitudinal	
  research	
  on	
  self-­‐regulatory	
  mechanisms	
  has	
  both	
  historically	
  and	
  
recently	
  adopted	
  one	
  of	
  the	
  following	
  two	
  characteristics	
  (see	
  DeShon,	
  2012	
  for	
  a	
  recent	
  
discussion	
  of	
  dynamic	
  research	
  in	
  organizational	
  science).	
  Researchers	
  tend	
  to	
  use	
  three	
  to	
  
five	
  measurement	
  time	
  points	
  to	
  capture	
  the	
  phenomenon	
  of	
  interest,	
  and/or	
  they	
  will	
  
separate	
  the	
  measurement	
  of	
  processes	
  and	
  outcomes	
  between	
  the	
  time	
  points	
  in	
  order	
  to	
  
present	
  a	
  case	
  for	
  mediation	
  (e.g.,	
  Converse,	
  Piccone,	
  Lockamy,	
  Miloslavic,	
  Mysiak,	
  &	
  
Pathak,	
  2014;	
  Cron,	
  Slocum,	
  VandeWalle	
  &	
  Fu,	
  2005;	
  Thomas	
  &	
  Mathieu,	
  1994).	
  Thus,	
  there	
  
are	
  few	
  instances	
  of	
  the	
  process	
  variables	
  being	
  repeatedly	
  measured	
  (see	
  Vancouver,	
  
Thompson,	
  Tischner	
  &	
  Putka,	
  2002	
  for	
  an	
  example	
  of	
  an	
  exception).	
  The	
  lack	
  of	
  empirical	
  
investigations	
  of	
  the	
  changes	
  in	
  self-­‐regulatory	
  variables	
  makes	
  it	
  difficult	
  to	
  ascertain	
  the	
  
specifics	
  of	
  the	
  trajectories	
  of	
  the	
  mechanisms	
  of	
  the	
  adaptation	
  process	
  over	
  time.	
  
However,	
  there	
  is	
  considerable	
  work	
  devoted	
  to	
  the	
  theoretical	
  understanding	
  of	
  the	
  
impact	
  of	
  cognitions	
  and	
  motivations	
  in	
  both	
  the	
  adaptation	
  and	
  self-­‐regulation	
  literatures.	
  
The	
  following	
  discussion	
  of	
  the	
  trajectory	
  changes	
  of	
  the	
  adaptation	
  process	
  will	
  therefore	
  
be	
  based	
  on	
  the	
  empirical	
  evidence	
  available	
  as	
  well	
  as	
  extrapolations	
  of	
  the	
  dynamics	
  
inherent	
  in	
  some	
  of	
  the	
  theoretical	
  works.	
  
	
  
The	
  Cognitive	
  Cycle	
  
Adaptive	
  Environment.	
  The	
  cognitive	
  aspects	
  of	
  the	
  adaptation	
  process	
  are	
  all	
  
focused	
  on	
  evaluating	
  the	
  change	
  that	
  occurred,	
  determining	
  the	
  source	
  of	
  it,	
  examining	
  the	
  

	
  

34	
  

	
  
result	
  of	
  current	
  behaviors,	
  and	
  analyzing	
  what	
  strategy	
  would	
  result	
  in	
  the	
  best	
  adaptive	
  
performance.	
  McGrath	
  (2001)	
  also	
  states,	
  “in	
  highly	
  novel	
  settings,	
  groups	
  should	
  follow	
  
the	
  variety-­‐generating	
  approach.	
  This	
  is	
  because,	
  absent	
  a	
  base	
  of	
  cause-­‐and-­‐effect	
  
understanding,	
  experimentation	
  generates	
  information	
  that	
  cannot	
  be	
  obtained	
  any	
  other	
  
way”	
  (p.118).	
  Research	
  in	
  that	
  adaptation	
  literature	
  suggests	
  that	
  employing	
  a	
  learning	
  
focus	
  prior	
  to	
  a	
  novel	
  change	
  results	
  in	
  increased	
  performance	
  in	
  such	
  an	
  environment	
  as	
  a	
  
deeper	
  understanding	
  of	
  the	
  task	
  is	
  developed	
  (Bell	
  &	
  Kozlowski,	
  2008).	
  Through	
  this	
  
deeper	
  understanding,	
  individuals	
  can	
  more	
  quickly	
  identify	
  the	
  source	
  of	
  the	
  change	
  and	
  
adjust	
  subsequent	
  behaviors	
  accordingly.	
  Given	
  that	
  they	
  are	
  all	
  very	
  interrelated	
  in	
  
achieving	
  these	
  objectives,	
  it	
  is	
  expected	
  that	
  they	
  would	
  all	
  follow	
  similar	
  trajectories	
  in	
  an	
  
adaptive	
  environment.	
  Specifically,	
  the	
  initial	
  levels	
  of	
  these	
  variables	
  will	
  be	
  high,	
  as	
  it	
  is	
  
necessary	
  to	
  engage	
  in	
  these	
  behaviors	
  in	
  order	
  to	
  effectively	
  adapt.	
  As	
  individuals	
  
continue	
  to	
  perform	
  in	
  an	
  adaptation	
  environment,	
  the	
  behaviors	
  will	
  decrease	
  slightly	
  as	
  
the	
  objectives	
  are	
  slowly	
  met.	
  Another	
  way	
  of	
  thinking	
  of	
  the	
  reason	
  for	
  the	
  decrease	
  in	
  
levels	
  is	
  based	
  on	
  the	
  fact	
  that	
  individuals	
  have	
  limited	
  resources	
  that	
  they	
  can	
  devote	
  
toward	
  a	
  task.	
  Initially,	
  individuals	
  need	
  to	
  allocate	
  more	
  resources	
  toward	
  cognitively	
  
understanding	
  the	
  environment,	
  but	
  this	
  resource	
  allocation	
  must	
  eventually	
  shift	
  to	
  
motivational	
  behaviors	
  as	
  a	
  strategy	
  is	
  chosen	
  and	
  tested	
  (and	
  as	
  they	
  transition	
  to	
  a	
  
performance	
  environment).	
  See	
  Figure	
  3	
  for	
  a	
  representation	
  of	
  these	
  trajectories.	
  
Research	
  in	
  the	
  adaptation	
  literature	
  has	
  not	
  yet	
  advanced	
  to	
  empirically	
  examining	
  how	
  
process	
  mechanisms	
  change	
  over	
  time.	
  However,	
  several	
  studies	
  have	
  suggested	
  that	
  
metacognition	
  is	
  related	
  to	
  increased	
  adaptive	
  performance	
  immediately	
  after	
  a	
  change	
  
(Bell	
  &	
  Kozlowski,	
  2008;	
  Ivancic	
  &	
  Hesketh,	
  2000;	
  Keith	
  &	
  Frese,	
  2005).	
  This	
  supports	
  the	
  

	
  

35	
  

	
  
statement	
  that	
  metacognitive	
  behaviors	
  will	
  be	
  initially	
  elevated	
  when	
  a	
  change	
  occurs.	
  
Authors	
  have	
  also	
  intimated	
  that	
  metacognitive	
  behaviors	
  will	
  decrease	
  over	
  time.	
  Some	
  
researchers	
  describe	
  metacognition	
  as	
  a	
  central	
  component	
  to	
  self-­‐regulating	
  cognition	
  
during	
  a	
  task	
  through	
  monitoring	
  the	
  environment	
  and	
  revising	
  strategies	
  so	
  that	
  changes	
  
are	
  more	
  quickly	
  recognized,	
  understood	
  and	
  addressed	
  (Bell	
  &	
  Kozlowski,	
  2008;	
  Ivancic	
  &	
  
Hesketh,	
  2000;	
  Keith	
  &	
  Frese,	
  2005).	
  Therefore,	
  it	
  is	
  expected	
  that	
  as	
  individuals	
  engage	
  in	
  
high	
  levels	
  of	
  metacognition,	
  the	
  change	
  will	
  be	
  understood	
  more	
  rapidly,	
  reducing	
  the	
  
need	
  to	
  continue	
  performing	
  these	
  behaviors,	
  resulting	
  in	
  their	
  decrease	
  over	
  time	
  in	
  an	
  
adaptation	
  environment.	
  	
  
Feedback	
  evaluation	
  behaviors	
  have	
  a	
  purpose	
  similar	
  to	
  metacognition.	
  
Performance	
  outcomes	
  are	
  able	
  to	
  serve	
  as	
  a	
  comparator	
  between	
  the	
  initial	
  goal	
  and	
  the	
  
result	
  of	
  actions	
  (e.g.,	
  effort)	
  directed	
  toward	
  that	
  goal.	
  Evaluation	
  is	
  the	
  behavioral	
  
manifestation	
  of	
  that	
  comparison,	
  which	
  informs	
  future	
  goal	
  formation	
  and	
  strategizing	
  
(i.e.,	
  metacognition;	
  Bell	
  &	
  Kozlowski,	
  2008;	
  Ford	
  et	
  al.,	
  1998).	
  Those	
  new	
  goals	
  and	
  
strategies	
  then	
  form	
  a	
  new	
  comparator	
  against	
  which	
  future	
  performance	
  outcomes	
  will	
  be	
  
evaluated.	
  Thus,	
  evaluation	
  behaviors	
  are	
  essential	
  for	
  effectively	
  engaging	
  in	
  the	
  cognitive	
  
cycle	
  of	
  adaptation	
  (i.e.,	
  goals	
  leading	
  to	
  behavior,	
  resulting	
  in	
  performance	
  and	
  evaluation,	
  
leading	
  to	
  revised	
  goals	
  and	
  behaviors).	
  Using	
  a	
  problem-­‐solving	
  framework	
  as	
  a	
  means	
  of	
  
understanding	
  the	
  adaptation	
  process,	
  Zaccaro,	
  Banks,	
  Kiechel-­‐Koles,	
  Kemp	
  and	
  Bader	
  
(2009)	
  described	
  performance	
  evaluation	
  as	
  “affirming	
  the	
  realignment	
  between	
  the	
  unit	
  
(organization	
  or	
  team)	
  and	
  its	
  environment”	
  (p.	
  7).	
  This	
  suggests	
  that	
  high	
  levels	
  of	
  
evaluation	
  activity	
  allow	
  the	
  team	
  to	
  identify	
  whether	
  a	
  strategy	
  is	
  effective	
  in	
  light	
  of	
  an	
  
adaptive	
  change,	
  which	
  then	
  leads	
  to	
  more	
  effective	
  performance	
  behaviors.	
  	
  In	
  addition	
  to	
  

	
  

36	
  

	
  
the	
  positive	
  impact	
  of	
  evaluation	
  behaviors	
  on	
  performance	
  outcomes,	
  researchers	
  have	
  
also	
  suggested	
  that	
  higher	
  levels	
  of	
  evaluation	
  behaviors	
  have	
  resulted	
  in	
  a	
  deeper	
  
understanding	
  of	
  the	
  task	
  (Bell	
  &	
  Kozlowski,	
  2008;	
  Ford	
  et	
  al.,	
  1998;	
  Kozlowski	
  &	
  Bell,	
  
2006).	
  As	
  individuals	
  develop	
  an	
  understanding	
  of	
  how	
  behaviors	
  impact	
  performance	
  
outcomes,	
  fewer	
  evaluation	
  behaviors	
  are	
  needed	
  over	
  time.	
  Therefore,	
  it	
  is	
  expected	
  the	
  
evaluation	
  will	
  decrease	
  as	
  individuals	
  continue	
  to	
  engage	
  in	
  adaptation	
  environments.	
  
	
  

The	
  third	
  mechanism	
  involved	
  in	
  the	
  cognitive	
  cycle	
  of	
  adaptation	
  is	
  learning-­‐

oriented	
  effort.	
  Similar	
  to	
  the	
  others,	
  this	
  form	
  of	
  effort	
  is	
  focused	
  on	
  gathering	
  information	
  
about	
  the	
  environment,	
  the	
  change,	
  and	
  an	
  appropriate	
  response	
  to	
  it.	
  Research	
  in	
  the	
  
adaptation	
  literature	
  typically	
  does	
  not	
  investigate	
  cognitive	
  effort	
  itself,	
  but	
  rather	
  the	
  
impact	
  of	
  different	
  goals	
  or	
  strategies	
  on	
  effort	
  behaviors.	
  However,	
  few	
  have	
  provided	
  
insight	
  on	
  how	
  these	
  effort	
  behaviors	
  may	
  change	
  over	
  time.	
  Yeo	
  and	
  Neal	
  (2004)	
  
examined	
  the	
  relationship	
  between	
  effort	
  and	
  practice	
  and	
  found	
  that	
  it	
  is	
  increased	
  during	
  
the	
  early	
  stages	
  of	
  a	
  novel	
  task.	
  	
  This	
  provides	
  initial	
  evidence	
  that	
  cognitive	
  effort	
  will	
  
likely	
  be	
  high	
  as	
  individuals	
  are	
  exposed	
  to	
  a	
  change	
  that	
  is	
  not	
  understood.	
  Research	
  has	
  
also	
  suggested	
  that	
  the	
  relationship	
  between	
  effort	
  and	
  performance	
  weakens	
  with	
  practice	
  
(Kanfer	
  &	
  Ackerman,	
  1989).	
  Yeo	
  and	
  Neal	
  (2004)	
  also	
  propose	
  that	
  the	
  relationship	
  
between	
  effort	
  and	
  performance	
  is	
  established	
  early	
  in	
  the	
  task	
  such	
  that	
  if	
  individuals	
  do	
  
not	
  devote	
  effort	
  immediately	
  after	
  a	
  change,	
  an	
  appropriate	
  strategy	
  will	
  not	
  be	
  
established	
  and	
  thus	
  increasing	
  effort	
  will	
  not	
  show	
  increased	
  performance.	
  These	
  findings	
  
support	
  the	
  assertions	
  made	
  by	
  Bell	
  and	
  Kozlowski	
  (2002a,	
  2008)	
  that	
  effort	
  not	
  only	
  has	
  a	
  
motivational	
  component	
  but	
  also	
  a	
  cognitive	
  one	
  –	
  individuals	
  must	
  recognize	
  that	
  a	
  new	
  
strategy	
  is	
  required	
  in	
  addition	
  to	
  maintaining	
  or	
  increasing	
  motivation	
  in	
  order	
  to	
  deal	
  

	
  

37	
  

	
  
with	
  the	
  change.	
  Therefore,	
  it	
  is	
  expected	
  that	
  although	
  learning-­‐oriented	
  effort	
  will	
  be	
  high	
  
initially	
  after	
  the	
  change,	
  these	
  behaviors	
  will	
  slowly	
  decrease	
  as	
  individuals	
  engage	
  in	
  the	
  
adaptation	
  environment.	
  	
  
Hypothesis	
  4a:	
  In	
  the	
  adaptation	
  environment,	
  metacognition	
  will	
  have	
  a	
  
high	
  intercept	
  and	
  a	
  weak	
  negative	
  slope.	
  
Hypothesis	
  4b:	
  In	
  the	
  adaptation	
  environment,	
  evaluation	
  will	
  have	
  a	
  high	
  
intercept	
  and	
  a	
  weak	
  negative	
  slope.	
  
Hypothesis	
  4c:	
  In	
  the	
  adaptation	
  environment,	
  learning-­‐oriented	
  effort	
  will	
  
have	
  a	
  high	
  intercept	
  and	
  a	
  weak	
  negative	
  slope.	
  
	
  
Performance	
  Environment.	
  As	
  individuals	
  move	
  out	
  of	
  an	
  adaptation-­‐focused	
  
environment	
  and	
  transition	
  to	
  a	
  performance	
  environment,	
  where	
  the	
  focus	
  is	
  on	
  executing	
  
a	
  strategy,	
  the	
  impact	
  of	
  cognitive	
  mechanisms	
  changes.	
  Since	
  the	
  cognitive	
  mechanisms	
  
are	
  primarily	
  centered	
  on	
  understanding	
  the	
  environment,	
  once	
  individuals	
  enter	
  a	
  
performance	
  environment,	
  continuing	
  to	
  engage	
  in	
  such	
  behaviors	
  is	
  less	
  necessary	
  and	
  
could	
  possibly	
  be	
  harmful	
  to	
  performance,	
  as	
  the	
  goal	
  is	
  now	
  to	
  execute	
  behaviors	
  rather	
  
than	
  continuing	
  to	
  reflect	
  on	
  strategies	
  (Kluger	
  &	
  DeNisi,	
  1996).	
  Considering	
  that	
  research	
  
has	
  not	
  yet	
  examined	
  the	
  trajectories	
  of	
  the	
  process	
  mechanisms	
  during	
  adaptation,	
  it	
  is	
  
not	
  surprising	
  that	
  research	
  has	
  not	
  yet	
  moved	
  toward	
  understanding	
  the	
  re-­‐stabilization	
  
of	
  these	
  regulatory	
  variables	
  after	
  a	
  change	
  is	
  addressed.	
  However,	
  through	
  extending	
  
adaptation	
  and	
  self-­‐regulation	
  research,	
  it	
  is	
  expected	
  that	
  these	
  cognitive	
  mechanisms	
  will	
  
decrease	
  in	
  importance	
  in	
  a	
  performance	
  environment.	
  Therefore,	
  it	
  is	
  expected	
  that	
  these	
  
cognitive	
  mechanisms	
  will	
  initially	
  be	
  at	
  a	
  moderate	
  level,	
  but	
  will	
  decrease	
  rapidly	
  as	
  

	
  

38	
  

	
  
individuals	
  continue	
  to	
  recognize	
  that	
  the	
  strategies	
  chosen	
  is	
  a	
  viable	
  solution	
  to	
  deal	
  with	
  
the	
  change	
  in	
  the	
  environment.	
  
Metacognition	
  is	
  discussed	
  as	
  regulation	
  of	
  strategy	
  effectiveness	
  (Bell	
  &	
  Kozlowski,	
  
2008).	
  Individuals	
  engage	
  in	
  this	
  activity	
  when	
  developing	
  an	
  understanding	
  of	
  the	
  
environment	
  or	
  the	
  impact	
  of	
  their	
  behaviors.	
  Initially	
  upon	
  transitioning	
  to	
  a	
  performance	
  
environment,	
  it	
  is	
  possible	
  that	
  metacognition	
  would	
  remain	
  slightly	
  elevated	
  as	
  
individuals	
  continue	
  to	
  test	
  whether	
  the	
  strategy	
  is	
  effective.	
  However,	
  when	
  the	
  
environment	
  is	
  understood,	
  it	
  is	
  logical	
  to	
  hypothesize	
  that	
  metacognitive	
  activity	
  would	
  
decrease	
  significantly	
  as	
  it	
  would	
  no	
  longer	
  be	
  necessary	
  to	
  re-­‐strategize	
  in	
  order	
  to	
  
maintain	
  effective	
  performance.	
  In	
  fact,	
  some	
  suggest	
  that	
  a	
  failure	
  to	
  reduce	
  metacognitive	
  
activity	
  would	
  result	
  in	
  a	
  lack	
  of	
  available	
  resources	
  that	
  could	
  be	
  devoted	
  to	
  executing	
  the	
  
task	
  since	
  cognitive	
  and	
  motivational	
  resources	
  can	
  be	
  in	
  competition	
  (e.g.,	
  March,	
  1991).	
  
Feedback	
  evaluation	
  is	
  a	
  critical	
  element	
  in	
  adjusting	
  behaviors	
  when	
  a	
  goal	
  is	
  not	
  
met	
  according	
  to	
  control	
  theory	
  (Carver	
  &	
  Scheier,	
  1982).	
  However,	
  when	
  the	
  goal-­‐
performance	
  gap	
  is	
  acceptable,	
  it	
  is	
  no	
  longer	
  necessary	
  to	
  engage	
  in	
  as	
  many	
  feedback	
  
evaluation	
  behaviors	
  in	
  order	
  to	
  be	
  effective	
  in	
  that	
  environment.	
  In	
  other	
  words,	
  once	
  
individuals	
  transition	
  to	
  a	
  performance	
  environment,	
  evaluation	
  behaviors	
  may	
  be	
  initially	
  
elevated	
  as	
  the	
  strategy	
  being	
  used	
  is	
  still	
  being	
  checked	
  for	
  accuracy,	
  but	
  feedback	
  seeking	
  
will	
  quickly	
  diminish	
  as	
  performance	
  continues	
  to	
  increase	
  as	
  expected.	
  
Learning-­‐oriented	
  effort	
  will	
  also	
  be	
  less	
  critical	
  in	
  performance	
  environments.	
  This	
  
form	
  of	
  effort	
  is	
  dedicated	
  to	
  understanding	
  the	
  environment.	
  However,	
  when	
  individuals	
  
transition	
  to	
  a	
  performance	
  environment,	
  they	
  have	
  chosen	
  a	
  strategy	
  to	
  deal	
  with	
  the	
  
change	
  and,	
  therefore,	
  are	
  no	
  longer	
  in	
  need	
  of	
  additional	
  information	
  to	
  inform	
  that	
  

	
  

39	
  

	
  
decision.	
  Given	
  this,	
  it	
  is	
  expected	
  that	
  learning-­‐oriented	
  effort,	
  similar	
  to	
  metacognition	
  
and	
  evaluation	
  behaviors,	
  will	
  initially	
  be	
  slightly	
  elevated	
  but	
  will	
  rapidly	
  increase	
  as	
  
individuals	
  engage	
  in	
  the	
  performance	
  environment.	
  
Hypothesis	
  5a:	
  In	
  the	
  performance	
  environment,	
  metacognition	
  will	
  have	
  a	
  
mid-­‐level	
  intercept	
  and	
  a	
  strong	
  negative	
  slope.	
  
Hypothesis	
  5b:	
  In	
  the	
  performance	
  environment,	
  evaluation	
  will	
  have	
  a	
  mid-­‐
level	
  intercept	
  and	
  a	
  strong	
  negative	
  slope.	
  
Hypothesis	
  5c:	
  In	
  the	
  performance	
  environment,	
  learning-­‐oriented	
  effort	
  
will	
  have	
  a	
  mid-­‐level	
  intercept	
  and	
  a	
  strong	
  negative	
  slope.	
  
	
  
The	
  Motivational	
  Cycle	
  
	
  

Adaptive	
  Environment.	
  Having	
  a	
  cognitive	
  understanding	
  of	
  an	
  environment	
  is	
  

essential	
  but	
  insufficient	
  for	
  effective	
  performance.	
  Individuals	
  must	
  also	
  be	
  motivated	
  to	
  
change	
  their	
  behaviors	
  in	
  the	
  face	
  of	
  a	
  change,	
  as	
  well	
  as	
  to	
  maintain	
  or	
  increase	
  effort	
  
levels,	
  when	
  transitioning	
  to	
  a	
  more	
  typical	
  performance	
  environment.	
  Each	
  of	
  the	
  three	
  
mechanisms	
  involved	
  in	
  the	
  motivational	
  cycle	
  of	
  the	
  adaptation	
  process	
  (goals,	
  self-­‐
efficacy,	
  outcome-­‐oriented	
  effort)	
  are	
  expected	
  to	
  follow	
  a	
  similar	
  trajectory	
  as	
  they	
  are	
  all	
  
focused	
  on	
  achieving	
  a	
  common	
  outcome:	
  the	
  calibration	
  of	
  effort	
  to	
  achieve	
  effective	
  
performance.	
  In	
  a	
  novel	
  environment,	
  the	
  likelihood	
  of	
  previous	
  goals	
  and	
  behaviors	
  being	
  
an	
  effective	
  performance	
  strategy	
  is	
  extremely	
  low.	
  This	
  will	
  impact	
  not	
  only	
  initial	
  goal	
  
levels,	
  but	
  also	
  efficacy	
  and	
  effort	
  allocation	
  behaviors.	
  Specifically,	
  it	
  is	
  expected	
  that	
  
individuals	
  will	
  initially	
  have	
  low	
  levels	
  of	
  these	
  mechanisms	
  as	
  they	
  determine	
  how	
  their	
  
behaviors	
  are	
  calibrated	
  to	
  the	
  outcomes.	
  However,	
  as	
  individuals	
  continue	
  to	
  engage	
  in	
  the	
  

	
  

40	
  

	
  
adaptation	
  environment,	
  it	
  is	
  expected	
  that	
  the	
  trajectories	
  of	
  these	
  motivational	
  
mechanisms	
  will	
  be	
  weakly	
  positive	
  as	
  individuals	
  gain	
  more	
  confidence,	
  set	
  higher	
  goals,	
  
and	
  put	
  more	
  effort	
  toward	
  the	
  task.	
  See	
  Figure	
  3	
  for	
  a	
  representation	
  of	
  these	
  trajectories.	
  
Similar	
  to	
  the	
  research	
  on	
  the	
  cognitive	
  cycle,	
  the	
  adaptation	
  literature	
  has	
  not	
  yet	
  
advanced	
  to	
  empirically	
  investigating	
  the	
  trajectories	
  of	
  these	
  mechanisms.	
  However,	
  
researchers	
  have	
  taken	
  steps	
  to	
  examine	
  how	
  goals	
  are	
  part	
  of	
  the	
  learning	
  process	
  and	
  
have	
  theoretically	
  applied	
  the	
  results	
  to	
  the	
  adaptation	
  process	
  after	
  a	
  change	
  is	
  initiated	
  as	
  
well.	
  	
  In	
  a	
  multilevel	
  study,	
  Chen	
  and	
  colleagues	
  (2005)	
  found	
  that	
  both	
  individual	
  and	
  
team	
  regulatory	
  goal	
  processes	
  fully-­‐mediated	
  the	
  relationship	
  between	
  self-­‐efficacy	
  (and	
  
collective	
  efficacy)	
  and	
  individual	
  (and	
  team)	
  adaptive	
  performance	
  (respectively).	
  At	
  the	
  
individual	
  level,	
  goal	
  choice	
  (defined	
  as	
  the	
  selection	
  of	
  a	
  goal)	
  positively	
  and	
  significantly	
  
predicted	
  goal	
  striving	
  (defined	
  as	
  the	
  effort	
  devoted	
  toward	
  achieving	
  that	
  goal),	
  which	
  in	
  
turn	
  predicted	
  adaptive	
  performance	
  after	
  the	
  change.	
  This	
  research	
  suggests	
  that	
  goals	
  
are	
  a	
  critical	
  part	
  of	
  a	
  regulation	
  process	
  that	
  influences	
  adaptive	
  performance.	
  However,	
  
what	
  is	
  unclear	
  is	
  the	
  relative	
  level	
  of	
  the	
  goals	
  made	
  after	
  a	
  change,	
  as	
  compared	
  to	
  goals	
  
made	
  before	
  a	
  change.	
  In	
  other	
  words,	
  Chen	
  et	
  al.	
  (2005)	
  show	
  that	
  goals	
  created	
  before	
  a	
  
change	
  are	
  related	
  to	
  performance	
  after	
  a	
  change,	
  but	
  they	
  did	
  not	
  provide	
  insight	
  on	
  
whether	
  the	
  goals	
  vary	
  based	
  on	
  whether	
  they	
  are	
  made	
  before	
  or	
  after	
  a	
  change	
  is	
  
introduced.	
  I	
  extend	
  their	
  research	
  by	
  suggesting	
  that	
  although	
  goals	
  will	
  increase	
  with	
  
performance	
  levels	
  over	
  time	
  as	
  individuals	
  engage	
  in	
  the	
  adaptive	
  environment,	
  goals	
  (like	
  
performance)	
  will	
  be	
  initially	
  low.	
  Ilies	
  (2003)	
  supports	
  this	
  claim	
  through	
  examining	
  the	
  
impact	
  of	
  performance	
  on	
  goal	
  levels	
  over	
  multiple	
  trials	
  in	
  a	
  task.	
  He	
  found	
  that	
  when	
  
individuals	
  experienced	
  negative	
  feedback,	
  goals	
  were	
  decreased,	
  as	
  would	
  be	
  expected	
  

	
  

41	
  

	
  
initially	
  after	
  a	
  change	
  is	
  introduced;	
  however,	
  goals	
  were	
  increased	
  following	
  positive	
  
feedback,	
  which	
  would	
  be	
  likely	
  once	
  individuals	
  start	
  to	
  increase	
  their	
  performance	
  in	
  the	
  
adaptation	
  environment.	
  
Researchers	
  have	
  consistently	
  described	
  the	
  utility	
  of	
  self-­‐efficacy	
  in	
  self-­‐regulation	
  
(e.g.,	
  Bandura,	
  1991;	
  Bell	
  &	
  Kozlowski,	
  2008;	
  Zimmerman,	
  1989).	
  Although	
  the	
  specific	
  
variables	
  involved	
  in	
  the	
  self-­‐regulatory	
  process	
  described	
  by	
  these	
  researchers	
  changes	
  
slightly	
  between	
  studies,	
  self-­‐efficacy	
  is	
  consistently	
  conceptualized	
  as	
  a	
  motivational	
  
mechanism	
  (along	
  with	
  intrinsic	
  motivation	
  and	
  goal	
  orientation)	
  that	
  works	
  in	
  concert	
  
with	
  cognitive	
  (e.g.,	
  knowledge	
  acquisition,	
  metacognitive	
  activity)	
  and	
  behavioral	
  
elements	
  (e.g.,	
  effort,	
  self-­‐evaluation	
  activity;	
  Bell	
  &	
  Kozlowski,	
  2008,	
  2010).	
  Research	
  has	
  
shown	
  that	
  self-­‐efficacy	
  has	
  a	
  unique,	
  positive,	
  and	
  significant	
  impact	
  on	
  adaptive	
  
performance	
  (e.g.,	
  Kozlowski,	
  Gully	
  et	
  al.,	
  2001;	
  Bell	
  &	
  Kozlowski,	
  2002a,	
  2002b,	
  2008).	
  
Using	
  a	
  cyclical	
  and	
  longitudinal	
  framework,	
  this	
  would	
  suggest	
  that	
  changes	
  in	
  self-­‐
efficacy	
  result	
  in	
  changes	
  in	
  performance,	
  and	
  vice	
  versa.	
  Therefore,	
  when	
  a	
  change	
  is	
  
introduced,	
  it	
  is	
  expected	
  that	
  both	
  performance	
  and	
  self-­‐efficacy	
  will	
  be	
  low.	
  Other	
  studies	
  
also	
  provide	
  initial	
  evidence	
  that	
  practicing	
  a	
  different	
  version	
  of	
  a	
  task	
  during	
  training	
  
results	
  in	
  increased	
  self-­‐efficacy	
  in	
  an	
  adaptive	
  situation	
  (Holladay	
  &	
  Quiñones,	
  2003).	
  
Extending	
  this	
  finding	
  to	
  variations	
  being	
  presented	
  in	
  the	
  adaptive	
  environment	
  itself,	
  it	
  
may	
  be	
  that	
  efficacy	
  will	
  increase	
  as	
  individuals	
  are	
  continually	
  exposed	
  to	
  such	
  changes.	
  	
  
The	
  third	
  motivational	
  mechanism,	
  outcome-­‐oriented	
  effort,	
  is	
  the	
  key	
  driver	
  of	
  
performance	
  behavior.	
  The	
  self-­‐regulation	
  literature	
  suggests	
  that	
  individuals	
  who	
  are	
  
committed	
  to	
  their	
  goals	
  are	
  more	
  likely	
  to	
  regulate	
  their	
  behaviors	
  to	
  achieve	
  those	
  goals.	
  
They	
  do	
  this	
  by	
  comparing	
  their	
  current	
  performance	
  state	
  to	
  the	
  previous	
  one	
  and	
  

	
  

42	
  

	
  
determining	
  if	
  there	
  is	
  a	
  discrepancy	
  between	
  the	
  effort	
  they	
  are	
  devoting	
  to	
  the	
  task	
  and	
  
the	
  outcomes	
  of	
  those	
  behaviors	
  (Bandura,	
  1993).	
  In	
  the	
  words	
  of	
  Bandura,	
  “[individuals]	
  
seek	
  self-­‐satisfaction	
  from	
  fulfilling	
  valued	
  goals	
  and	
  are	
  prompted	
  to	
  intensify	
  their	
  efforts	
  
by	
  discontent	
  with	
  substandard	
  performances”	
  (p.	
  130).	
  When	
  performance	
  is	
  drastically	
  
increased	
  (i.e.,	
  when	
  a	
  change	
  that	
  requires	
  adaptation	
  is	
  introduced),	
  although	
  individuals	
  
may	
  be	
  motivated	
  to	
  increase	
  performance,	
  a	
  lack	
  of	
  understanding	
  about	
  the	
  reason	
  for	
  
the	
  performance	
  discrepancy	
  may	
  result	
  in	
  low	
  outcome-­‐oriented	
  effort	
  behaviors	
  initially.	
  
For	
  this	
  reason,	
  I	
  delineate	
  two	
  elements	
  of	
  effort:	
  a	
  learning-­‐oriented	
  component	
  (which	
  
will	
  be	
  increased	
  in	
  the	
  face	
  of	
  a	
  change)	
  and	
  an	
  outcome-­‐oriented	
  component	
  (which	
  will	
  
be	
  low	
  initially,	
  as	
  there	
  is	
  limited	
  understanding	
  of	
  the	
  impact	
  of	
  effort	
  behaviors).	
  Support	
  
for	
  this	
  concept	
  can	
  be	
  found	
  in	
  some	
  initial	
  research	
  in	
  the	
  adaptation	
  literature.	
  Ford	
  et	
  al.	
  
(1998)	
  determined	
  that	
  influencing	
  the	
  type	
  of	
  goals	
  set	
  by	
  individuals	
  led	
  to	
  different	
  
levels	
  of	
  effort	
  devoted	
  to	
  the	
  learning	
  process.	
  Those	
  with	
  mastery	
  goals	
  were	
  more	
  
motivated	
  during	
  learning	
  and	
  therefore	
  devoted	
  more	
  effort	
  to	
  the	
  task	
  at	
  hand.	
  This	
  
suggests	
  that	
  more	
  effective	
  individuals	
  are	
  those	
  who	
  devote	
  more	
  learning-­‐oriented	
  
effort	
  as	
  opposed	
  to	
  outcome-­‐oriented	
  effort	
  when	
  exposed	
  to	
  an	
  adaptive	
  change.	
  
However,	
  in	
  other	
  research,	
  behavioral	
  effort	
  (as	
  described	
  by	
  the	
  amount	
  of	
  behavior	
  
devoted	
  to	
  the	
  task,	
  i.e.,	
  outcome-­‐oriented	
  effort)	
  was	
  related	
  to	
  enhanced	
  performance	
  in	
  
an	
  adaptive	
  scenario	
  (e.g.,	
  Bell	
  &	
  Kozlowski,	
  2008;	
  Kozlowski,	
  et	
  al.,	
  1999).	
  Therefore,	
  it	
  is	
  
likely	
  that	
  as	
  individuals	
  continue	
  to	
  engage	
  in	
  an	
  adaptive	
  environment,	
  both	
  effort	
  and	
  
performance	
  will	
  increase.	
  	
  
Hypothesis	
  6a:	
  In	
  the	
  adaptation	
  environment,	
  goals	
  will	
  have	
  a	
  low	
  
intercept	
  and	
  a	
  weak	
  positive	
  slope.	
  

	
  

43	
  

	
  
Hypothesis	
  6b:	
  In	
  the	
  adaptation	
  environment,	
  self-­‐efficacy	
  will	
  have	
  a	
  low	
  
intercept	
  and	
  a	
  weak	
  positive	
  slope.	
  
Hypothesis	
  6c:	
  In	
  the	
  adaptation	
  environment,	
  outcome-­‐oriented	
  effort	
  will	
  
have	
  a	
  low	
  intercept	
  and	
  a	
  weak	
  positive	
  slope.	
  
	
  
	
  

Performance	
  Environment.	
  As	
  was	
  true	
  with	
  the	
  cognitive	
  cycle,	
  the	
  trajectories	
  of	
  

the	
  mechanisms	
  involved	
  in	
  the	
  motivational	
  cycle	
  change	
  once	
  individuals	
  shift	
  from	
  an	
  
adaptive	
  to	
  a	
  performance	
  environment	
  and	
  resources	
  must	
  be	
  re-­‐allocated.	
  Unlike	
  the	
  
cognitive	
  elements,	
  which	
  become	
  less	
  critical,	
  the	
  motivational	
  mechanisms	
  take	
  on	
  a	
  new	
  
level	
  of	
  importance.	
  In	
  a	
  performance	
  environment,	
  a	
  cognitively	
  driven	
  strategy	
  is	
  chosen	
  
that	
  has	
  shown	
  initial	
  evidence	
  of	
  increasing	
  performance.	
  However,	
  an	
  effective	
  strategy	
  is	
  
only	
  as	
  effective	
  as	
  the	
  behaviors	
  devoted	
  toward	
  the	
  strategy.	
  Therefore,	
  the	
  mechanisms	
  
of	
  the	
  motivational	
  cycle,	
  as	
  seen	
  in	
  setting	
  goals,	
  feeling	
  efficacious	
  and	
  behaviorally	
  
devoting	
  effort	
  toward	
  performance	
  outcomes,	
  will	
  start	
  relatively	
  low,	
  but	
  rapidly	
  
increase	
  as	
  the	
  need	
  for	
  maintaining	
  motivation	
  in	
  performance	
  becomes	
  all	
  the	
  more	
  
critical	
  for	
  effectiveness.	
  	
  
	
  

Research	
  in	
  the	
  goal	
  setting	
  literature	
  has	
  long	
  suggested	
  that	
  goals	
  are	
  critical	
  for	
  

maintaining	
  and	
  increasing	
  performance	
  (e.g.,	
  Brown,	
  2005;	
  Latham	
  &	
  Locke,	
  1991).	
  In	
  a	
  
performance	
  environment	
  where	
  change	
  is	
  no	
  longer	
  the	
  driving	
  force	
  behind	
  behaviors,	
  it	
  
is	
  critical	
  to	
  set	
  difficult	
  and	
  specific	
  goals	
  in	
  order	
  for	
  performance	
  to	
  continue	
  to	
  increase	
  
and	
  not	
  remain	
  the	
  same	
  or	
  decrease	
  due	
  to	
  fatigue.	
  As	
  goals	
  will	
  only	
  just	
  be	
  starting	
  to	
  be	
  
calibrated	
  initially	
  upon	
  transitioning	
  to	
  a	
  performance	
  environment,	
  it	
  is	
  expected	
  that	
  
they	
  will	
  be	
  at	
  only	
  a	
  moderate	
  level.	
  However,	
  as	
  performance	
  continues	
  to	
  increase	
  and	
  

	
  

44	
  

	
  
goals	
  are	
  met,	
  these	
  goals	
  are	
  expected	
  to	
  increase.	
  	
  
	
  

Similarly,	
  self-­‐regulation	
  literature	
  suggests	
  that	
  as	
  individuals	
  practice	
  a	
  task	
  more,	
  

they	
  develop	
  more	
  expertise,	
  which	
  results	
  in	
  higher	
  levels	
  of	
  self-­‐efficacy	
  (e.g.,	
  Holladay	
  &	
  
Quiñones,	
  2003).	
  As	
  discussed	
  earlier,	
  when	
  a	
  novel	
  change	
  is	
  introduced,	
  efficacy	
  will	
  be	
  
negatively	
  impacted.	
  However,	
  as	
  individuals	
  transition	
  to	
  a	
  performance	
  environment,	
  
confidence	
  in	
  their	
  ability	
  to	
  continue	
  performing	
  the	
  task	
  will	
  likely	
  be	
  re-­‐established	
  and	
  
increasing.	
  Therefore,	
  it	
  is	
  expected	
  that	
  the	
  trajectory	
  of	
  this	
  mechanism	
  will	
  also	
  be	
  
positive	
  throughout	
  the	
  performance	
  environment	
  as	
  individuals	
  gain	
  more	
  expertise	
  in	
  
the	
  task	
  and	
  have	
  opportunity	
  to	
  show	
  their	
  capabilities	
  in	
  it.	
  
Finally,	
  outcome-­‐oriented	
  effort,	
  or	
  effort	
  directed	
  at	
  completing	
  a	
  task	
  (Blau,	
  1993),	
  
is	
  essential	
  in	
  a	
  performance	
  environment.	
  While	
  cognitive	
  behaviors	
  are	
  the	
  critical	
  
components	
  in	
  adaptive	
  environments,	
  when	
  the	
  situation	
  has	
  stabilized	
  and	
  the	
  change	
  is	
  
understood,	
  individuals	
  need	
  to	
  re-­‐allocate	
  resources	
  to	
  devote	
  behavioral	
  effort	
  toward	
  
achieving	
  the	
  highest	
  level	
  of	
  performance	
  as	
  possible.	
  Yeo	
  and	
  Neal	
  (2008)	
  examined	
  the	
  
intra-­‐individual	
  dynamics	
  of	
  the	
  relationship	
  between	
  effort	
  and	
  performance	
  and	
  found	
  
that	
  changes	
  in	
  effort	
  predicted	
  changes	
  in	
  performance.	
  The	
  authors	
  suggested	
  that	
  an	
  
increase	
  in	
  effort	
  reflected	
  an	
  increase	
  in	
  motivation	
  to	
  work	
  harder	
  to	
  reach	
  the	
  desired	
  
performance	
  level.	
  In	
  the	
  second	
  study	
  they	
  manipulated	
  the	
  difficulty	
  of	
  the	
  task	
  at	
  the	
  
midway	
  point.	
  They	
  found	
  that	
  the	
  relationship	
  between	
  effort	
  and	
  performance	
  did	
  not	
  
replicate	
  in	
  the	
  more	
  difficult	
  scenario	
  perhaps	
  due	
  to	
  the	
  situation	
  being	
  too	
  difficult	
  to	
  
handle	
  simply	
  by	
  increasing	
  effort.	
  Taken	
  together,	
  I	
  suggest	
  that	
  the	
  initial	
  level	
  of	
  effort	
  
may	
  only	
  be	
  at	
  a	
  moderate	
  level	
  as	
  individuals	
  are	
  only	
  beginning	
  to	
  transition	
  to	
  a	
  more	
  
stable	
  performance	
  environment;	
  however,	
  it	
  is	
  expected	
  that	
  outcome-­‐oriented	
  effort	
  will	
  

	
  

45	
  

	
  
increase	
  as	
  individuals	
  engage	
  in	
  a	
  performance	
  environment	
  and	
  continue	
  to	
  optimize	
  on	
  
the	
  strategy	
  they	
  have	
  chosen	
  to	
  adapt	
  to	
  the	
  change	
  previously	
  introduced	
  into	
  their	
  task.	
  
Hypothesis	
  7a:	
  In	
  the	
  performance	
  environment,	
  goals	
  will	
  have	
  a	
  mid-­‐level	
  
intercept	
  and	
  a	
  strong	
  positive	
  slope.	
  
Hypothesis	
  7b:	
  In	
  the	
  performance	
  environment,	
  self-­‐efficacy	
  will	
  have	
  a	
  
mid-­‐level	
  intercept	
  and	
  a	
  strong	
  positive	
  slope.	
  
Hypothesis	
  7c:	
  In	
  the	
  performance	
  environment,	
  outcome-­‐oriented	
  effort	
  
will	
  have	
  a	
  mid-­‐level	
  intercept	
  and	
  a	
  strong	
  positive	
  slope.	
  

	
  

46	
  

	
  
SECOND	
  ORDER	
  CHANGES:	
  RELATIONSHIPS	
  
Considering	
  that	
  the	
  mechanisms	
  of	
  the	
  adaptation	
  process	
  (i.e.,	
  goals,	
  
metacognition,	
  self-­‐efficacy,	
  effort,	
  and	
  evaluation)	
  will	
  clearly	
  change	
  over	
  time,	
  it	
  is	
  now	
  
necessary	
  to	
  discuss	
  the	
  causes	
  of	
  the	
  trajectories.	
  A	
  process	
  perspective	
  implies	
  dynamic	
  
and	
  cyclical	
  changes.	
  However,	
  the	
  current	
  longitudinal	
  analyses	
  employed	
  in	
  the	
  
adaptation	
  and	
  self-­‐regulation	
  literature	
  typically	
  investigate	
  solely	
  the	
  level	
  changes	
  in	
  the	
  
variables	
  involved	
  in	
  the	
  process	
  (e.g.,	
  latent	
  growth	
  curves),	
  or	
  SEM	
  frameworks	
  are	
  
employed	
  when	
  investigating	
  the	
  process	
  at	
  one	
  or	
  two	
  points	
  in	
  time.	
  Level,	
  or	
  trajectory,	
  
changes	
  in	
  the	
  process	
  of	
  adaptation	
  only	
  tell	
  part	
  of	
  the	
  story	
  of	
  this	
  dynamic	
  
phenomenon.	
  The	
  first	
  order	
  changes	
  provide	
  insight	
  into	
  how	
  variables	
  change,	
  but	
  not	
  
into	
  why	
  they	
  change.	
  The	
  reason	
  why	
  certain	
  trajectories	
  are	
  seen	
  is	
  based	
  on	
  the	
  
relationships	
  a	
  variable	
  has	
  with	
  previous	
  levels	
  of	
  itself	
  (an	
  autoregressive	
  component)	
  
and	
  with	
  previous	
  levels	
  of	
  other	
  related	
  variables	
  (a	
  cross-­‐variable	
  lagged	
  component).	
  
Therefore,	
  detailed	
  theoretical	
  descriptions	
  of	
  these	
  relationships,	
  as	
  well	
  as	
  analyses	
  (e.g.,	
  
vector	
  autoregression,	
  longitudinal	
  cross-­‐lag	
  panel	
  regression),	
  that	
  can	
  capture	
  cyclical	
  
relationships	
  are	
  needed.	
  
Processes	
  are	
  sometimes	
  referred	
  to	
  as	
  a	
  “black	
  box”	
  as	
  they	
  are	
  not	
  only	
  difficult	
  to	
  
theorize	
  but	
  very	
  difficult	
  to	
  test.	
  I	
  argue	
  that	
  both	
  the	
  trajectories	
  and	
  relationships	
  of	
  the	
  
variables	
  involved	
  in	
  a	
  process	
  need	
  to	
  be	
  investigated	
  longitudinally	
  in	
  order	
  to	
  
understand	
  the	
  phenomenon.	
  Therefore,	
  I	
  turn	
  my	
  attention	
  to	
  the	
  relationships	
  involved	
  
within	
  each	
  process,	
  describing	
  the	
  overall	
  relationships	
  of	
  the	
  mechanisms	
  involved	
  in	
  the	
  
cognitive	
  and	
  motivational	
  cycles,	
  as	
  well	
  as	
  specifying	
  the	
  changes	
  in	
  the	
  bivariate	
  
relationship	
  embedded	
  in	
  the	
  cycle,	
  in	
  order	
  to	
  define	
  the	
  expected	
  dynamics	
  of	
  the	
  

	
  

47	
  

	
  
process.	
  I	
  will	
  begin	
  my	
  discussion	
  with	
  the	
  dynamics	
  of	
  the	
  mechanisms	
  in	
  the	
  cognitive	
  
cycle	
  (containing	
  evaluation,	
  learning-­‐oriented	
  effort,	
  and	
  metacognition)	
  both	
  in	
  the	
  
adaptation	
  and	
  performance	
  processes,	
  as	
  seen	
  in	
  the	
  inner	
  circle	
  of	
  Figures	
  3	
  and	
  4,	
  
respectively.	
  Then	
  I	
  will	
  move	
  to	
  present	
  the	
  dynamics	
  of	
  the	
  motivational	
  cycle	
  
(containing	
  goals,	
  self-­‐efficacy,	
  and	
  outcome-­‐oriented	
  effort)	
  in	
  these	
  processes,	
  as	
  
depicted	
  in	
  the	
  outer	
  circles	
  of	
  Figure	
  4.	
  	
  	
  
	
  
The	
  Cognitive	
  Cycle	
  
As	
  described	
  earlier,	
  the	
  cognitive	
  cycle	
  has	
  three	
  key	
  mechanisms:	
  evaluation,	
  
metacognition,	
  and	
  learning-­‐oriented	
  effort	
  behaviors.	
  Evaluation	
  is	
  primarily	
  described	
  as	
  
seeking	
  feedback	
  information	
  about	
  performance	
  in	
  order	
  to	
  determine	
  the	
  source	
  and	
  
impact	
  of	
  the	
  change	
  based	
  on	
  performance	
  change.	
  Metacognition	
  is	
  an	
  examination	
  of	
  the	
  
effectiveness	
  of	
  a	
  strategy	
  used	
  in	
  a	
  performance	
  situation.	
  Learning-­‐oriented	
  effort	
  is	
  the	
  
component	
  of	
  effort	
  that	
  is	
  not	
  directly	
  relevant	
  to	
  performance	
  that	
  is	
  devoted	
  to	
  
understanding	
  the	
  source	
  and	
  impact	
  of	
  the	
  change.	
  	
  These	
  mechanisms	
  dynamically	
  
interact	
  with	
  each	
  other,	
  where	
  performance	
  impacts	
  learning-­‐oriented	
  effort,	
  which	
  then	
  
impacts	
  metacognitive	
  behaviors,	
  finally	
  impacting	
  subsequent	
  performance.	
  In	
  a	
  new	
  
environment,	
  individuals	
  are	
  faced	
  with	
  a	
  decrease	
  in	
  performance.	
  This	
  large	
  goal-­‐
performance	
  discrepancy	
  results	
  in	
  the	
  need	
  for	
  effort	
  to	
  be	
  devoted	
  to	
  understanding	
  the	
  
“why”	
  behind	
  the	
  change.	
  Based	
  on	
  control	
  theory,	
  this	
  suggests	
  that	
  individuals	
  are	
  
required	
  to	
  increase	
  in	
  their	
  reflection	
  on	
  the	
  strategies	
  employed	
  in	
  order	
  to	
  correct	
  
performance	
  decreases	
  in	
  order	
  to	
  re-­‐attain	
  the	
  desired	
  goal	
  level.	
  Evaluation	
  behaviors	
  
are	
  also	
  a	
  critical	
  component	
  of	
  the	
  cognitive	
  cycle	
  in	
  its	
  impact	
  on	
  both	
  learning-­‐oriented	
  

	
  

48	
  

	
  
Figure	
  4	
  

	
  
	
  

	
  

Heuristic	
  Representations	
  of	
  the	
  Relationships	
  in	
  the	
  Adaptation	
  and	
  Performance	
  Processes	
  
	
  
	
  
	
  
	
  	
  	
  	
  Adaptation	
  Process	
  
	
  
	
  
	
  
	
  
	
  	
  	
  	
  	
  	
  	
  Performance	
  Process	
  

Performance*

Performance*

Evalua:on*
Learning*
Effort*

	
  	
  	
  	
  	
  	
  	
  	
  

	
  

Evalua:on*

Metacogni:on*
Outcome*
Effort*
*
*

Learning*
Effort*

Metacogni:on*
Outcome*
Effort*
*
*

Self/*
efficacy*

Self/*
efficacy*

Goals*

Goals*

49	
  

	
  

	
  
effort	
  and	
  metacognitive	
  behaviors	
  as	
  it	
  is	
  necessary	
  to	
  reflect	
  on	
  performance	
  levels	
  in	
  
order	
  to	
  determine	
  that	
  there	
  is	
  a	
  need	
  for	
  learning	
  and	
  strategy	
  revision.	
  The	
  following	
  
sections	
  specify	
  the	
  bivariate	
  relationships	
  involved	
  in	
  the	
  dynamic	
  cognitive	
  process	
  
occurring	
  in	
  both	
  adaptation	
  and	
  routine	
  performance	
  processes	
  (see	
  Figure	
  4),	
  based	
  on	
  
the	
  trajectory	
  changes	
  previously	
  discussed	
  (see	
  Figure	
  3).	
  Bivariate	
  relationships	
  (as	
  
opposed	
  to	
  multivariate)	
  are	
  described	
  in	
  order	
  to	
  present	
  the	
  theoretical	
  workings	
  of	
  the	
  
dynamic	
  adaptation	
  process	
  as	
  clearly	
  as	
  possible.	
  First,	
  I	
  will	
  describe	
  the	
  cycle	
  where	
  
performance	
  impacts	
  learning-­‐oriented	
  effort,	
  which	
  then	
  influences	
  metacognition,	
  which	
  
cycles	
  back	
  to	
  effect	
  performance.	
  Then,	
  I	
  will	
  present	
  how	
  evaluation	
  behaviors	
  are	
  
intertwined	
  in	
  this	
  cycle	
  (in	
  that	
  performance	
  influences	
  evaluation),	
  which	
  then	
  impacts	
  
effort	
  and	
  metacognitive	
  behaviors.	
  
	
  
Performance	
  and	
  Learning-­‐Oriented	
  Effort	
  
One	
  challenge	
  associated	
  with	
  dynamic	
  theory	
  is	
  the	
  inherent	
  lagged	
  relationship	
  
between	
  variables.	
  In	
  this	
  case,	
  the	
  first	
  link	
  in	
  the	
  cognitive	
  cycle	
  is	
  that	
  performance	
  will	
  
influence	
  subsequent	
  effort	
  behaviors,	
  but	
  static	
  research	
  typically	
  only	
  investigates	
  the	
  
opposite	
  direction	
  of	
  this	
  relationship.	
  Yeo	
  and	
  Neal	
  (2008)	
  found	
  that	
  the	
  amount	
  of	
  
cognitive	
  effort	
  individuals	
  devoted	
  to	
  a	
  task	
  fluctuated	
  with	
  respect	
  to	
  performance,	
  
expertise,	
  difficulty,	
  and	
  practice.	
  Less	
  cognitive	
  effort	
  was	
  associated	
  with	
  more	
  practice	
  
and	
  more	
  expertise,	
  but	
  more	
  cognitive	
  effort	
  was	
  associated	
  with	
  more	
  difficult	
  
environments.	
  The	
  results	
  of	
  their	
  study	
  suggest	
  that	
  cognitive	
  effort	
  and	
  performance	
  had	
  
a	
  strong	
  positive	
  correlation	
  between	
  people	
  but	
  only	
  a	
  weak	
  positive	
  one	
  within	
  the	
  
person.	
  Given	
  that	
  the	
  situation	
  will	
  become	
  more	
  complex	
  after	
  the	
  adaptive	
  change,	
  it	
  is	
  

	
  

50	
  

	
  
expected	
  that,	
  initially,	
  individuals	
  will	
  need	
  high	
  levels	
  of	
  cognitive	
  effort	
  (supported	
  by	
  
Yeo	
  &	
  Neal,	
  2008),	
  as	
  performance	
  will	
  be	
  negatively	
  impacted.	
  Therefore,	
  it	
  is	
  expected	
  
that	
  there	
  will	
  be	
  a	
  weak	
  negative	
  relationship	
  between	
  performance	
  and	
  cognitive	
  effort	
  
such	
  that	
  as	
  performance	
  begins	
  to	
  increase,	
  less	
  cognitive	
  effort	
  (which	
  can	
  be	
  considered	
  
learning-­‐oriented	
  effort)	
  will	
  be	
  needed.	
  
Hypothesis	
  8a:	
  In	
  the	
  adaptation	
  process,	
  performance	
  will	
  have	
  a	
  weak	
  and	
  
negative	
  impact	
  on	
  subsequent	
  learning-­‐oriented	
  effort	
  behaviors.	
  
	
  
However,	
  as	
  individuals	
  understand	
  the	
  source	
  of	
  the	
  change	
  and	
  choose	
  a	
  new	
  
strategy,	
  individuals	
  will	
  transition	
  to	
  a	
  performance	
  process	
  where	
  effort	
  and	
  
performance	
  will	
  be	
  more	
  strongly	
  related.	
  In	
  this	
  environment,	
  higher	
  performance	
  will	
  
lead	
  to	
  less	
  effort	
  devoted	
  toward	
  learning	
  since	
  additional	
  strategizing	
  and	
  information	
  
gathering	
  will	
  not	
  be	
  as	
  necessary.	
  This	
  is	
  supported	
  by	
  Yeo	
  and	
  Neal	
  (2008)	
  in	
  their	
  
finding	
  that	
  more	
  practice	
  and	
  decreased	
  difficulty	
  resulted	
  in	
  less	
  cognitive	
  effort	
  devoted	
  
to	
  the	
  task.	
  This	
  is	
  likely	
  due	
  to	
  the	
  individual	
  understanding	
  the	
  source	
  of	
  the	
  change,	
  
recognizing	
  the	
  strategy	
  that	
  needed	
  to	
  be	
  employed,	
  and	
  realizing	
  that	
  any	
  additional	
  
resources	
  devoted	
  to	
  re-­‐strategizing	
  would	
  be	
  wasted	
  effort	
  that	
  should	
  instead	
  be	
  devoted	
  
to	
  executing	
  the	
  task	
  (March,	
  1991).	
  
Hypothesis	
  8b:	
  In	
  the	
  performance	
  process,	
  performance	
  will	
  have	
  a	
  strong	
  
and	
  negative	
  impact	
  on	
  subsequent	
  learning-­‐oriented	
  effort	
  behaviors.	
  
	
  
Learning-­‐Oriented	
  Effort	
  and	
  Metacognition	
  
Continuing	
  in	
  the	
  cognitive	
  cycle,	
  when	
  individuals	
  are	
  faced	
  with	
  a	
  change,	
  not	
  only	
  

	
  

51	
  

	
  
does	
  performance	
  changes	
  cue	
  the	
  need	
  for	
  additional	
  information	
  and	
  understanding,	
  but	
  
individuals	
  who	
  devote	
  more	
  effort	
  to	
  these	
  behaviors	
  will	
  also	
  likely	
  engage	
  in	
  more	
  
effective	
  strategizing.	
  This	
  suggests	
  that	
  learning-­‐oriented	
  effort	
  and	
  metacognitive	
  activity	
  
will	
  both	
  be	
  devoted	
  toward	
  obtaining	
  more	
  information	
  about	
  what	
  strategy	
  would	
  be	
  
most	
  effective.	
  If	
  effort	
  is	
  not	
  devoted	
  to	
  learning,	
  it	
  is	
  likely	
  that	
  the	
  individual	
  is	
  not	
  
analyzing	
  the	
  strategy,	
  resulting	
  in	
  less	
  metacognitive	
  activity.	
  Zimmerman	
  (1989)	
  
indirectly	
  discussed	
  this	
  relationship	
  in	
  the	
  triadic	
  model	
  of	
  self-­‐regulation	
  where	
  cognitive	
  
behaviors	
  are	
  thought	
  to	
  provide	
  the	
  strategic	
  links	
  among	
  the	
  internal	
  regulatory	
  process	
  
(of	
  goal	
  formation	
  and	
  motivation),	
  the	
  external	
  behaviors	
  employed,	
  and	
  feedback	
  that	
  is	
  
required	
  to	
  test	
  the	
  strategy.	
  Empirically,	
  Clause,	
  Delbridge,	
  Schmitt,	
  Chan,	
  and	
  Jennings	
  
(2001)	
  investigated	
  the	
  relationship	
  between	
  motivational	
  factors,	
  metacognition,	
  effort	
  
and	
  performance	
  in	
  police	
  officer	
  candidates	
  applying	
  for	
  a	
  position.	
  Results	
  from	
  this	
  
study	
  suggest	
  that	
  the	
  more	
  metacognitive	
  activity	
  the	
  applicant	
  engaged	
  in,	
  the	
  more	
  
effort	
  they	
  devoted	
  to	
  engaging	
  in	
  the	
  task	
  and	
  the	
  higher	
  the	
  performance	
  they	
  achieved.	
  
These	
  studies	
  support	
  the	
  relationship	
  seen	
  in	
  Figure	
  4,	
  where	
  the	
  relationship	
  between	
  
effort	
  devoted	
  to	
  learning	
  and	
  metacognitive	
  behaviors	
  is	
  expected	
  to	
  be	
  positive	
  given	
  that	
  
as	
  individuals	
  engage	
  in	
  strategic	
  thinking	
  about	
  the	
  task,	
  they	
  are	
  likely	
  to	
  devote	
  effort	
  to	
  
determine	
  the	
  effectiveness	
  of	
  their	
  strategy	
  through	
  engaging	
  in	
  metacognitive	
  activity.	
  
Hypothesis	
  9a:	
  In	
  the	
  adaptation	
  process,	
  learning-­‐oriented	
  effort	
  will	
  have	
  
a	
  strong	
  and	
  positive	
  impact	
  on	
  subsequent	
  metacognition	
  behaviors.	
  
	
  
However,	
  as	
  individuals	
  are	
  exposed	
  to	
  the	
  changed	
  environment	
  for	
  longer	
  periods,	
  
performance	
  will	
  increase,	
  resulting	
  in	
  a	
  decreased	
  need	
  for	
  learning-­‐oriented	
  effort	
  and	
  

	
  

52	
  

	
  
metacognitive	
  activity	
  (see	
  Figure	
  3	
  for	
  a	
  reminder	
  of	
  the	
  expected	
  trajectory	
  changes).	
  
Given	
  that	
  learning	
  and	
  metacognition	
  are	
  consistently	
  related	
  to	
  each	
  other	
  in	
  the	
  
development	
  of	
  knowledge	
  and	
  strategies,	
  and	
  that	
  their	
  relationship	
  spans	
  across	
  domains	
  
(Veenman,	
  Van	
  Hout-­‐Wolters	
  &	
  Afflerbach,	
  2006),	
  it	
  is	
  expected	
  that	
  this	
  relationship	
  
would	
  remain	
  positive	
  regardless	
  of	
  whether	
  an	
  individual	
  is	
  initially	
  learning	
  a	
  task,	
  re-­‐
strategizing	
  based	
  on	
  a	
  change,	
  or	
  monitoring	
  the	
  effectiveness	
  of	
  established	
  strategies.	
  A	
  
recent	
  meta-­‐analysis	
  reported	
  that	
  the	
  relationship	
  between	
  metacognition	
  and	
  learning	
  
strategies	
  was	
  one	
  of	
  the	
  strongest	
  effects	
  among	
  self-­‐regulatory	
  variables	
  (ρ	
  =	
  .83,	
  k	
  =	
  39,	
  
N	
  =	
  9,529;	
  Sitzmann	
  &	
  Ely,	
  2011),	
  suggesting	
  that	
  devoting	
  effort	
  toward	
  learning	
  presents	
  
a	
  natural	
  inclination	
  for	
  individuals	
  to	
  test	
  and	
  evaluate	
  that	
  information	
  through	
  
metacognitive	
  behavior.	
  Therefore,	
  in	
  the	
  performance	
  environment,	
  it	
  is	
  anticipated	
  that	
  
there	
  will	
  be	
  a	
  weaker,	
  though	
  still	
  positive,	
  relationship	
  between	
  effort	
  and	
  metacognition	
  
(see	
  Figure	
  4),	
  since	
  a	
  certain	
  level	
  of	
  learning	
  will	
  remain	
  as	
  individuals	
  must	
  maintain	
  
their	
  understanding	
  of	
  the	
  environment	
  and	
  monitor	
  the	
  effectiveness	
  of	
  strategies.	
  
Hypothesis	
  9b:	
  In	
  the	
  performance	
  process,	
  learning-­‐oriented	
  effort	
  will	
  
have	
  a	
  weak	
  and	
  positive	
  impact	
  on	
  subsequent	
  metacognitive	
  behaviors.	
  
	
  
Metacognition	
  and	
  Performance	
  
Unlike	
  some	
  of	
  the	
  research	
  on	
  the	
  other	
  relationships	
  in	
  the	
  cognitive	
  cycle	
  
described	
  above,	
  work	
  on	
  metacognition	
  typically	
  incorporates	
  some	
  complex	
  situation	
  
where	
  individuals	
  or	
  groups	
  need	
  to	
  re-­‐strategize	
  in	
  order	
  to	
  improve	
  their	
  performance.	
  
While	
  some	
  of	
  this	
  work	
  is	
  in	
  the	
  education	
  literature,	
  examining	
  the	
  impact	
  of	
  
metacognition	
  and	
  intelligence	
  on	
  performance	
  (e.g.,	
  Landine	
  &	
  Stewart,	
  1998;	
  Pintrich	
  &	
  

	
  

53	
  

	
  
DeGroot,	
  1990),	
  some	
  work	
  has	
  also	
  been	
  done	
  in	
  the	
  adaptation	
  literature.	
  In	
  an	
  
investigation	
  of	
  the	
  team	
  adaptation	
  process,	
  Zaccaro	
  and	
  colleagues	
  (2009)	
  proposed	
  that	
  
effective	
  strategies	
  and	
  performance	
  are	
  involved	
  in	
  a	
  problem	
  solving	
  process	
  where	
  the	
  
strategy	
  is	
  assessed	
  as	
  performance	
  occurs,	
  suggesting	
  that	
  this	
  metacognitive	
  activity	
  is	
  
critical	
  for	
  effective	
  team	
  adaptive	
  performance.	
  Researchers	
  have	
  also	
  indirectly	
  
determined	
  that	
  there	
  is	
  a	
  positive	
  relationship	
  between	
  metacognition	
  and	
  performance	
  in	
  
both	
  field	
  and	
  laboratory	
  settings	
  where	
  complexity	
  increased	
  through	
  investigating	
  the	
  
impact	
  of	
  metacognition	
  on	
  learning	
  knowledge	
  outcomes,	
  which	
  then	
  were	
  found	
  to	
  be	
  
positively	
  associated	
  with	
  adaptive	
  performance	
  (e.g.,	
  Bell	
  &	
  Kozlowski,	
  2008;	
  Keith	
  &	
  
Frese,	
  2005;	
  White,	
  Mueller-­‐Hanson,	
  Dorsey,	
  Pulakos,	
  Wisecarver,	
  Deagle	
  &	
  Medini,	
  2005).	
  
Thus,	
  given	
  the	
  increasing	
  levels	
  of	
  metacognition	
  and	
  performance	
  (see	
  Figure	
  2),	
  it	
  is	
  
hypothesized	
  that	
  metacognition	
  will	
  have	
  a	
  strong	
  and	
  positive	
  impact	
  on	
  performance	
  in	
  
adaptive	
  environments	
  (see	
  Figure	
  4).	
  
Hypothesis	
  10a:	
  In	
  the	
  adaptation	
  process,	
  metacognition	
  will	
  have	
  a	
  strong	
  
and	
  positive	
  impact	
  on	
  subsequent	
  performance.	
  
	
  
However,	
  once	
  a	
  strategy	
  is	
  chosen	
  and	
  individuals	
  shift	
  to	
  a	
  performance	
  process,	
  
metacognitive	
  activity	
  will	
  no	
  longer	
  be	
  needed	
  in	
  order	
  to	
  continue	
  to	
  improve	
  
performance.	
  In	
  organizational	
  theory,	
  March	
  (1991)	
  and	
  McGrath	
  (2001)	
  both	
  suggested	
  
that	
  extensive	
  and	
  unneeded	
  exploration	
  (or	
  learning-­‐oriented)	
  behaviors	
  result	
  in	
  wasted	
  
time	
  and	
  resources	
  when	
  those	
  cognitive	
  and	
  motivational	
  efforts	
  could	
  be	
  devoted	
  to	
  
exploiting	
  the	
  strategy	
  that	
  is	
  being	
  effective.	
  Therefore,	
  it	
  is	
  expected	
  that	
  increases	
  in	
  
metacognitive	
  activity	
  would	
  lead	
  to	
  decreases	
  in	
  performance	
  as	
  resources	
  that	
  should	
  be	
  

	
  

54	
  

	
  
devoted	
  to	
  performance	
  execution	
  is	
  instead	
  being	
  devoted	
  to	
  unneeded	
  strategy	
  
contemplation.	
  This	
  will	
  result	
  in	
  a	
  weakly	
  negative	
  relationship	
  between	
  metacognition	
  
and	
  performance	
  in	
  the	
  performance	
  process	
  (see	
  Figure	
  4).	
  
Hypothesis	
  10b:	
  In	
  the	
  performance	
  process,	
  metacognition	
  will	
  have	
  a	
  
weak	
  and	
  negative	
  impact	
  on	
  subsequent	
  performance.	
  
	
  
Performance	
  and	
  Evaluation	
  
Transitioning	
  to	
  focusing	
  on	
  the	
  role	
  of	
  evaluation	
  behaviors	
  in	
  the	
  cognitive	
  cycle,	
  
research	
  in	
  the	
  feedback	
  seeking	
  literature	
  suggests	
  that	
  the	
  more	
  feedback	
  that	
  is	
  
evaluated,	
  the	
  better	
  performance	
  will	
  be	
  in	
  the	
  future	
  (Kluger	
  &	
  DeNisi,	
  1996).	
  Similar	
  to	
  
the	
  issue	
  described	
  with	
  the	
  relationship	
  between	
  effort	
  and	
  performance,	
  the	
  direction	
  of	
  
the	
  relationship	
  posed	
  in	
  the	
  larger	
  literature	
  is	
  typically	
  the	
  inverse	
  of	
  what	
  is	
  expected	
  in	
  
the	
  present	
  theory.	
  Investigating	
  the	
  impact	
  of	
  performance	
  on	
  evaluation	
  behaviors	
  
requires	
  a	
  cyclical	
  and	
  longitudinal	
  approach,	
  which	
  is	
  largely	
  not	
  the	
  focus	
  of	
  
investigations.	
  Unlike	
  some	
  of	
  the	
  relationships	
  involved	
  in	
  the	
  adaptation	
  process,	
  
research	
  in	
  the	
  extent	
  literature	
  has	
  empirically	
  investigated	
  the	
  relationship	
  between	
  
evaluation	
  and	
  performance.	
  In	
  a	
  series	
  of	
  articles	
  by	
  Kozlowski	
  and	
  colleagues,	
  evaluative	
  
activity	
  (measured	
  as	
  time	
  spent	
  on	
  investigating	
  performance	
  feedback)	
  has	
  been	
  
positively	
  related	
  to	
  increased	
  knowledge	
  and	
  performance	
  after	
  an	
  adaptive	
  event	
  is	
  
introduced.	
  They	
  attributed	
  this	
  to	
  a	
  desire	
  to	
  develop	
  a	
  deeper	
  understanding	
  of	
  the	
  task	
  
that	
  one	
  attains	
  from	
  evaluating	
  feedback	
  (Bell	
  &	
  Kozlowski,	
  2008;	
  Ford	
  et	
  al.,	
  1998;	
  
Kozlowski	
  &	
  Bell,	
  2006).	
  These	
  findings	
  suggest	
  that	
  evaluative	
  activity	
  is	
  useful	
  to	
  the	
  
extent	
  that	
  it	
  impacts	
  metacognitive	
  activity	
  and	
  strategy	
  development,	
  but	
  it	
  also	
  suggests	
  

	
  

55	
  

	
  
that	
  there	
  is	
  a	
  relationship	
  between	
  evaluation	
  activity	
  and	
  performance	
  level.	
  Evaluation	
  
activities	
  are	
  needed	
  to	
  determine	
  why	
  performance	
  decreased	
  and	
  how	
  to	
  respond.	
  
Given	
  that	
  these	
  studies	
  did	
  not	
  look	
  at	
  multiple	
  iterations	
  of	
  performance	
  in	
  the	
  
adaptive	
  environment,	
  it	
  is	
  unknown	
  what	
  the	
  cyclical	
  relationship	
  between	
  evaluation	
  and	
  
performance	
  would	
  be.	
  However,	
  given	
  their	
  findings,	
  it	
  is	
  expected	
  that	
  when	
  individuals	
  
are	
  faced	
  with	
  an	
  unacceptable	
  goal-­‐performance	
  gap	
  (i.e.,	
  when	
  a	
  novel	
  change	
  occurs	
  and	
  
adaptation	
  is	
  required),	
  they	
  will	
  seek	
  feedback	
  in	
  order	
  to	
  diagnose	
  the	
  change	
  and	
  the	
  
new	
  strategy	
  that	
  should	
  be	
  employed.	
  As	
  individuals	
  continue	
  to	
  perform	
  and	
  develop	
  an	
  
understanding	
  of	
  the	
  environment,	
  less	
  evaluation	
  activity	
  will	
  be	
  needed,	
  resulting	
  in	
  a	
  
negative	
  relationship	
  between	
  evaluation	
  and	
  performance	
  in	
  adaptive	
  environments.	
  This	
  
is	
  based	
  on	
  the	
  general	
  principle	
  of	
  a	
  negative	
  feedback	
  loop	
  as	
  discussed	
  in	
  control	
  theory	
  
(Klein,	
  1989).	
  
Hypothesis	
  11a:	
  In	
  the	
  adaptation	
  process,	
  performance	
  will	
  have	
  a	
  weak	
  
and	
  negative	
  impact	
  on	
  subsequent	
  evaluation	
  behaviors.	
  
	
  
The	
  negative	
  feedback	
  loop,	
  as	
  described	
  by	
  control	
  theory,	
  will	
  continue	
  to	
  result	
  in	
  
a	
  negative	
  relationship	
  between	
  performance	
  and	
  evaluation	
  behaviors	
  in	
  a	
  performance	
  
process	
  because	
  individuals	
  will	
  continue	
  to	
  reduce	
  their	
  feedback	
  seeking	
  behaviors	
  once	
  
their	
  performance	
  begins	
  to	
  stabilize.	
  Researchers	
  have	
  suggested	
  that	
  feedback	
  seeking	
  
behavior	
  is	
  only	
  useful	
  when	
  there	
  is	
  a	
  need	
  or	
  an	
  uncertainty	
  to	
  which	
  the	
  individual	
  is	
  
responding	
  (Ashford	
  &	
  Cummings,	
  1983).	
  Therefore,	
  in	
  non-­‐adaptive	
  environments,	
  
increases	
  in	
  performance	
  will	
  trigger	
  decreases	
  in	
  evaluation	
  behaviors,	
  as	
  the	
  level	
  of	
  
performance	
  will	
  inform	
  individuals	
  on	
  the	
  effectiveness	
  of	
  their	
  behaviors.	
  Therefore,	
  they	
  

	
  

56	
  

	
  
will	
  not	
  need	
  to	
  continue	
  searching	
  for	
  the	
  underlying	
  meaning	
  of	
  performance	
  changes	
  
through	
  evaluating	
  feedback.	
  If	
  this	
  evaluation	
  behavior	
  continues,	
  then	
  it	
  is	
  likely	
  that	
  this	
  
is	
  due	
  to	
  continuing	
  low	
  performance.	
  
Hypothesis	
  11b:	
  In	
  the	
  performance	
  process,	
  performance	
  will	
  have	
  a	
  
strong	
  and	
  negative	
  impact	
  on	
  subsequent	
  evaluation	
  behaviors.	
  
	
  
Evaluation	
  and	
  Learning-­‐Oriented	
  Effort	
  	
  
Evaluation	
  is	
  the	
  examining	
  of	
  feedback,	
  and	
  this	
  investigation	
  of	
  performance	
  is	
  
expected	
  to	
  impact	
  strategizing	
  in	
  two	
  ways.	
  The	
  first	
  way	
  evaluation	
  impacts	
  the	
  cognitive	
  
cycle	
  is	
  by	
  influencing	
  the	
  cognitive	
  effort	
  devoted	
  to	
  filling	
  in	
  any	
  gaps	
  in	
  knowledge.	
  Both	
  
evaluation	
  and	
  effort	
  are	
  behavioral	
  mechanisms	
  that	
  have	
  been	
  discussed	
  as	
  critical	
  
components	
  in	
  understanding	
  performance	
  (Jundt,	
  2009).	
  Evaluating	
  feedback	
  without	
  
devoting	
  effort	
  based	
  on	
  that	
  information	
  is	
  a	
  waste	
  of	
  cognitive	
  resources,	
  which,	
  in	
  
adaptive	
  environments,	
  is	
  just	
  as	
  critical	
  as	
  the	
  behavioral	
  resources	
  used	
  to	
  adapt	
  to	
  a	
  
change.	
  In	
  self-­‐regulatory	
  research,	
  the	
  feedback	
  and	
  strategic	
  goals	
  were	
  experimentally	
  
shown	
  to	
  be	
  associated	
  with	
  the	
  effort	
  devoted	
  by	
  individuals.	
  Both	
  goals	
  and	
  evaluation	
  
were	
  needed	
  in	
  order	
  for	
  performance	
  to	
  increase	
  due	
  to	
  effort	
  (Bandura	
  &	
  Cervone,	
  1983).	
  
Therefore,	
  these	
  researchers	
  not	
  only	
  show	
  the	
  critical	
  link	
  of	
  the	
  relationship	
  between	
  
evaluation	
  and	
  strategic	
  effort	
  behaviors,	
  but	
  it	
  also	
  supports	
  the	
  importance	
  of	
  the	
  
dynamics	
  of	
  the	
  cognitive	
  cycle	
  of	
  which	
  this	
  relationship	
  is	
  a	
  part.	
  Therefore,	
  it	
  is	
  expected	
  
that,	
  in	
  an	
  adaptive	
  environment,	
  increases	
  in	
  evaluation	
  behaviors	
  (see	
  Figure	
  2)	
  will	
  be	
  
strongly	
  and	
  positively	
  associated	
  with	
  effort	
  devoted	
  to	
  understanding	
  the	
  reason	
  behind	
  
the	
  performance	
  feedback	
  (see	
  Figure	
  4).	
  	
  

	
  

57	
  

	
  
Hypothesis	
  12a:	
  In	
  the	
  adaptation	
  process,	
  evaluation	
  behaviors	
  will	
  have	
  a	
  
strong	
  and	
  positive	
  impact	
  on	
  subsequent	
  learning-­‐oriented	
  effort	
  behaviors.	
  
	
  
However,	
  over	
  time,	
  as	
  individuals	
  are	
  exposed	
  to	
  the	
  environment,	
  develop	
  a	
  
strategy,	
  and	
  experience	
  performance	
  increases,	
  evaluation	
  and	
  effort	
  behaviors	
  would	
  
both	
  decrease	
  as	
  cognitive	
  resources	
  are	
  not	
  as	
  central	
  a	
  need	
  in	
  the	
  performance	
  
environment.	
  Hattie	
  and	
  Timperly	
  (2007)	
  describe	
  the	
  importance	
  of	
  evaluation	
  behaviors	
  
in	
  determining	
  the	
  extent	
  to	
  which	
  further	
  understanding	
  and	
  learning	
  are	
  needed.	
  Ashford	
  
and	
  Cummings	
  (1983)	
  also	
  state	
  that	
  one	
  of	
  the	
  most	
  critical	
  reasons	
  why	
  feedback	
  is	
  
sought	
  is	
  to	
  reduce	
  uncertainty	
  about	
  goal-­‐performance	
  discrepancies.	
  In	
  addition	
  to	
  
learning	
  from	
  performance	
  feedback,	
  individuals	
  must	
  also	
  learn	
  by	
  obtaining	
  more	
  
information	
  from	
  the	
  environment.	
  Therefore,	
  to	
  the	
  extent	
  that	
  evaluation	
  behaviors	
  are	
  
engaged	
  in	
  (see	
  Figure	
  2),	
  learning-­‐oriented	
  effort	
  will	
  follow;	
  however,	
  given	
  that	
  fewer	
  of	
  
these	
  behaviors	
  will	
  be	
  performed,	
  there	
  would	
  be	
  a	
  weaker,	
  but	
  still	
  positive,	
  relationship	
  
between	
  these	
  mechanisms	
  (see	
  Figure	
  4).	
  	
  
Hypothesis	
  12b:	
  In	
  the	
  performance	
  process,	
  evaluation	
  behaviors	
  will	
  have	
  
a	
  weak	
  and	
  positive	
  impact	
  on	
  subsequent	
  learning-­‐oriented	
  effort	
  
behaviors.	
  
	
  
Evaluation	
  and	
  Metacognition	
  	
  
The	
  second	
  way	
  evaluation	
  impacts	
  the	
  cognitive	
  cycle	
  is	
  by	
  impacting	
  the	
  
development	
  and	
  testing	
  of	
  strategies	
  through	
  enhancing	
  metacognition.	
  Both	
  feedback	
  
evaluation	
  and	
  metacognitive	
  behaviors	
  are	
  concerned	
  with	
  understanding	
  the	
  underlying	
  

	
  

58	
  

	
  
strategy	
  of	
  the	
  task,	
  although	
  one	
  is	
  a	
  direct	
  investigation	
  of	
  performance	
  feedback	
  while	
  
the	
  other	
  is	
  an	
  analysis	
  of	
  the	
  strategy	
  used	
  to	
  attain	
  that	
  performance.	
  Therefore,	
  it	
  is	
  
logical	
  that	
  they	
  would	
  be	
  positively	
  associated	
  with	
  each	
  other.	
  This	
  is	
  supported	
  by	
  early	
  
self-­‐regulatory	
  research,	
  where	
  the	
  relationship	
  between	
  evaluation	
  behaviors	
  and	
  
metacognition	
  were	
  indirectly	
  discussed	
  in	
  the	
  triadic	
  model	
  of	
  self-­‐regulation	
  
(Zimmerman,	
  1989).	
  Strategy	
  development	
  and	
  testing	
  (i.e.,	
  metacognition)	
  provided	
  an	
  
important	
  link	
  between	
  the	
  internal	
  process	
  Zimmerman	
  described	
  (containing	
  goal	
  
formation	
  and	
  motivation)	
  and	
  the	
  external	
  process	
  (of	
  behaviors	
  and	
  feedback).	
  This	
  was	
  
supported	
  by	
  Ertmer	
  and	
  Newby	
  (1996)	
  in	
  their	
  inclusion	
  of	
  evaluation	
  activities	
  within	
  an	
  
overall	
  discussion	
  of	
  metacognitive	
  behaviors.	
  The	
  authors	
  suggest	
  that	
  without	
  having	
  re-­‐
evaluated	
  the	
  strategy	
  employed,	
  when	
  evaluating	
  feedback	
  information,	
  the	
  feedback	
  is	
  
less	
  useful.	
  More	
  specifically	
  associated	
  with	
  adaptation,	
  Ford,	
  Bell,	
  Kozlowski	
  and	
  
colleagues	
  conducted	
  a	
  series	
  of	
  studies	
  where	
  evaluation	
  behaviors	
  were	
  related	
  to	
  
increased	
  strategic	
  knowledge	
  and	
  adaptive	
  performance.	
  They	
  further	
  claimed	
  that	
  this	
  
strategic	
  knowledge	
  and	
  adaptive	
  performance	
  was	
  reflective	
  of	
  a	
  deeper	
  processing	
  of	
  the	
  
task	
  (Bell	
  &	
  Kozlowski,	
  2008;	
  Ford	
  et	
  al.,	
  1998;	
  Kozlowski	
  &	
  Bell,	
  2006).	
  The	
  positive	
  
relationship	
  between	
  evaluation	
  and	
  metacognition	
  also	
  appears	
  in	
  team-­‐level	
  theories	
  of	
  
the	
  adaptation	
  process	
  where	
  strategies	
  currently	
  being	
  employed	
  are	
  analyzed	
  through	
  
performance	
  information,	
  which	
  then	
  informs	
  subsequent	
  strategy	
  development	
  (Rosen	
  et	
  
al.,	
  2011;	
  Zaccaro	
  et	
  al.,	
  2009).	
  Therefore,	
  individuals	
  who	
  engage	
  in	
  evaluation	
  behaviors	
  
will	
  engage	
  in	
  more	
  strategic	
  thinking	
  (i.e.,	
  metacognitive	
  activity)	
  about	
  the	
  change	
  that	
  
occurred,	
  resulting	
  in	
  a	
  strong	
  positive	
  relationship.	
  	
  Similarly,	
  individuals	
  who	
  choose	
  not	
  
to	
  evaluate	
  their	
  performance	
  would	
  likely	
  not	
  think	
  about	
  their	
  strategy	
  as	
  they	
  likely	
  did	
  

	
  

59	
  

	
  
not	
  recognize	
  any	
  discrepancies	
  in	
  the	
  feedback.	
  
Hypothesis	
  13a:	
  In	
  the	
  adaptation	
  process,	
  evaluation	
  behaviors	
  will	
  have	
  a	
  
strong	
  and	
  positive	
  impact	
  on	
  subsequent	
  metacognition	
  behaviors.	
  
	
  
Even	
  once	
  a	
  strategy	
  is	
  chosen	
  and	
  individuals	
  shift	
  to	
  a	
  performance	
  process,	
  the	
  
extent	
  to	
  which	
  individuals	
  devote	
  energy	
  and	
  effort	
  to	
  evaluating	
  performance	
  feedback	
  
will	
  be	
  reflected	
  in	
  the	
  extent	
  to	
  which	
  they	
  engage	
  in	
  metacognitive	
  activity.	
  Hattie	
  and	
  
Timperly	
  (2007)	
  discuss	
  the	
  positive	
  impact	
  of	
  evaluation	
  behaviors	
  in	
  the	
  engagement	
  in	
  
metacognitive	
  activity.	
  They	
  describe	
  one	
  of	
  the	
  key	
  benefits	
  of	
  evaluating	
  performance	
  
feedback	
  as	
  being	
  able	
  to	
  detect	
  errors,	
  which	
  provides	
  the	
  opportunity	
  for	
  individuals	
  to	
  
seek	
  out	
  better	
  strategies	
  and	
  problem	
  solve	
  more	
  effectively.	
  Given	
  that	
  both	
  evaluation	
  
and	
  metacognitive	
  activity	
  should	
  not	
  be	
  as	
  required	
  in	
  a	
  performance	
  process,	
  there	
  will	
  
be	
  a	
  lack	
  of	
  variance	
  in	
  both	
  of	
  these	
  variables,	
  resulting	
  in	
  a	
  weaker,	
  though	
  still	
  positive,	
  
relationship.	
  
Hypothesis	
  13b:	
  In	
  the	
  performance	
  process,	
  evaluation	
  behaviors	
  will	
  have	
  
a	
  weak	
  and	
  positive	
  impact	
  on	
  subsequent	
  metacognition	
  behaviors.	
  
	
  
The	
  Motivational	
  Cycle	
  
The	
  motivational	
  cycle	
  has	
  three	
  key	
  mechanisms,	
  as	
  discussed	
  earlier:	
  goals,	
  self-­‐
efficacy,	
  and	
  outcome-­‐oriented	
  effort	
  behaviors.	
  Goals	
  are	
  used	
  to	
  calibrate	
  motivation	
  
through	
  setting	
  a	
  particular	
  outcome	
  to	
  achieve.	
  Self-­‐efficacy	
  is	
  the	
  confidence	
  an	
  individual	
  
has	
  in	
  achieving	
  a	
  goal.	
  Outcome-­‐oriented	
  effort	
  is	
  another	
  component	
  of	
  effort	
  that	
  is	
  
directly	
  relevant	
  to	
  performance	
  by	
  focusing	
  on	
  executing	
  behaviors	
  to	
  complete	
  the	
  task.	
  

	
  

60	
  

	
  
These	
  mechanisms	
  dynamically	
  interact	
  with	
  each	
  other	
  where	
  performance	
  influences	
  the	
  
subsequent	
  goal	
  that	
  is	
  set,	
  which	
  drives	
  behaviors	
  directed	
  toward	
  achieving	
  that	
  goal,	
  
resulting	
  in	
  performance	
  based	
  on	
  those	
  behaviors.	
  Self-­‐efficacy	
  is	
  also	
  a	
  critical	
  
component	
  of	
  this	
  process	
  as	
  the	
  level	
  of	
  confidence	
  associated	
  with	
  performance	
  impacts	
  
both	
  the	
  level	
  of	
  the	
  goal	
  and	
  the	
  behavior	
  that	
  is	
  devoted	
  to	
  achieving	
  it.	
  In	
  a	
  new	
  
environment,	
  goals	
  may	
  not	
  initially	
  be	
  set	
  at	
  a	
  high	
  level	
  considering	
  the	
  unknown	
  nature	
  
of	
  the	
  environment;	
  however,	
  as	
  confidence	
  increases,	
  the	
  amount	
  of	
  effort	
  devoted	
  to	
  
achieving	
  the	
  outcome	
  will	
  be	
  enhanced.	
  Similar	
  to	
  the	
  discussion	
  of	
  the	
  cognitive	
  cycle,	
  the	
  
following	
  sections	
  specify	
  the	
  bivariate	
  relationships	
  involved	
  in	
  the	
  dynamic	
  motivational	
  
process	
  occurring	
  in	
  both	
  adaptation	
  and	
  routine	
  performance	
  processes	
  (Figure	
  4).	
  The	
  
bivariate	
  relationships	
  are	
  based	
  on	
  the	
  logic	
  discussed	
  earlier	
  in	
  the	
  overall	
  description	
  of	
  
the	
  process	
  as	
  well	
  as	
  the	
  expected	
  trajectories	
  of	
  the	
  mechanisms	
  involved	
  (see	
  Figure	
  3).	
  
First	
  I	
  begin	
  my	
  discussion	
  of	
  the	
  motivational	
  cycle	
  with	
  the	
  description	
  of	
  how	
  
performance	
  impacts	
  goals,	
  which	
  then	
  influences	
  the	
  outcome-­‐oriented	
  effort	
  devoted	
  to	
  
the	
  task,	
  which	
  finally	
  cycles	
  back	
  to	
  effect	
  performance.	
  Then,	
  I	
  will	
  explicate	
  how	
  self-­‐
efficacy	
  is	
  intertwined	
  in	
  this	
  motivational	
  cycle	
  through	
  describing	
  how	
  performance	
  
impacts	
  individuals’	
  confidence	
  in	
  their	
  abilities,	
  which	
  then	
  influences	
  subsequent	
  goal	
  
levels	
  and	
  effort	
  behaviors.	
  
	
  
Performance	
  and	
  Goals	
  
	
  

Beginning	
  with	
  the	
  well-­‐examined	
  relationship	
  between	
  performance	
  and	
  goals,	
  

researchers	
  have,	
  for	
  many	
  decades,	
  supported	
  the	
  positive	
  impact	
  of	
  goals	
  on	
  
performance.	
  Work	
  in	
  the	
  goal-­‐setting	
  literature,	
  in	
  particular,	
  suggests	
  that	
  individuals	
  

	
  

61	
  

	
  
calibrate	
  their	
  performance	
  to	
  their	
  goals	
  (Bandura,	
  1991;	
  Latham	
  &	
  Locke,	
  1991).	
  
However,	
  it	
  is	
  also	
  possible	
  that	
  performance	
  impacts	
  goals	
  as	
  well.	
  Carver	
  and	
  Scheier	
  
(1998)	
  suggest	
  that	
  goals	
  may	
  initially	
  fluctuate	
  in	
  response	
  to	
  an	
  increase	
  in	
  complexity	
  as	
  
individuals	
  attempt	
  to	
  identify	
  how	
  to	
  respond.	
  Evaluating	
  performance	
  changes	
  is	
  critical	
  
in	
  order	
  to	
  understand	
  how	
  to	
  create	
  effective	
  and	
  reasonable	
  goals.	
  Therefore,	
  it	
  is	
  likely	
  
that	
  in	
  novel	
  environments,	
  there	
  would	
  initially	
  be	
  a	
  weak	
  relationship	
  between	
  
performance	
  and	
  goals.	
  Research	
  in	
  the	
  adaptation	
  literature	
  also	
  shows	
  that	
  goals	
  are	
  
positively	
  related	
  to	
  adaptive	
  performance	
  measured	
  in	
  a	
  variety	
  of	
  settings	
  incorporating	
  
changes	
  in	
  component,	
  coordinative	
  and	
  dynamic	
  complexity	
  (Bell	
  &	
  Kozlowski,	
  2008;	
  
Drach-­‐Zahavy	
  &	
  Somech,	
  1999;	
  Washburn,	
  Smith	
  &	
  Taglialatela,	
  2005).	
  At	
  the	
  team	
  level,	
  
LePine	
  (2005)	
  found	
  that	
  goals,	
  by	
  themselves,	
  did	
  not	
  effectively	
  predict	
  initial	
  levels	
  of	
  
team	
  performance	
  adaptation.	
  However,	
  LePine	
  did	
  find	
  that	
  when	
  difficult	
  goals	
  were	
  set,	
  
only	
  teams	
  of	
  individuals	
  low	
  in	
  performance	
  orientation	
  were	
  able	
  to	
  adapt	
  effectively,	
  
and	
  easy	
  goals	
  were	
  equally	
  ineffective	
  for	
  both	
  goal	
  orientation	
  types.	
  This	
  was	
  also	
  true	
  in	
  
his	
  examination	
  of	
  the	
  moderation	
  of	
  goal	
  orientation	
  with	
  respect	
  to	
  transition	
  and	
  action	
  
processes.	
  Although	
  research	
  typically	
  focuses	
  on	
  the	
  impact	
  of	
  goals	
  on	
  performance,	
  
there	
  is	
  initial	
  evidence	
  that	
  performance	
  will	
  also	
  impact	
  subsequent	
  goals,	
  with	
  this	
  
relationship	
  being	
  weak	
  when	
  an	
  individual	
  is	
  initially	
  faced	
  with	
  the	
  adaptive	
  change	
  (see	
  
Figure	
  4).	
  This	
  weak	
  relationship	
  is	
  due	
  to	
  individuals	
  having	
  insufficient	
  information	
  
about	
  the	
  source	
  of,	
  or	
  the	
  appropriate	
  strategy	
  to	
  deal	
  with,	
  the	
  adaptive	
  change,	
  resulting	
  
in	
  goals	
  that	
  may	
  not	
  be	
  feasible.	
  
Hypothesis	
  14a:	
  In	
  the	
  adaptation	
  process,	
  performance	
  will	
  have	
  a	
  weak	
  
and	
  positive	
  impact	
  on	
  subsequent	
  goals.	
  

	
  

62	
  

	
  
However,	
  as	
  individuals	
  shift	
  to	
  a	
  performance	
  process	
  where	
  performance	
  is	
  more	
  
stable	
  and	
  a	
  strategy	
  is	
  chosen,	
  the	
  relationship	
  between	
  performance	
  and	
  goals	
  will	
  
become	
  stronger.	
  In	
  this	
  environment,	
  individuals	
  can	
  gain	
  more	
  confidence	
  in	
  their	
  
performance,	
  resulting	
  in	
  higher	
  goals.	
  Although	
  some	
  researchers	
  suggest	
  that	
  there	
  is	
  a	
  
negative	
  relationship	
  between	
  goals	
  and	
  performance	
  (e.g.,	
  Vancouver,	
  Thompson,	
  &	
  
Williams,	
  2001),	
  others	
  suggest	
  that	
  this	
  is	
  not	
  the	
  case	
  in	
  highly	
  complex	
  environments	
  
where	
  feedback	
  is	
  provided	
  (Schmidt	
  &	
  DeShon,	
  2010;	
  Sitzmann	
  &	
  Yeo,	
  2013).	
  
Furthermore,	
  given	
  that	
  the	
  direction	
  of	
  the	
  relationship	
  is	
  from	
  performance	
  onto	
  goals,	
  it	
  
is	
  expected	
  that	
  even	
  in	
  a	
  performance	
  environment,	
  the	
  relationship	
  will	
  remain	
  positive	
  
and	
  strong	
  (see	
  Figure	
  4).	
  
Hypothesis	
  14b:	
  In	
  the	
  performance	
  process,	
  performance	
  will	
  have	
  a	
  
strong	
  and	
  positive	
  impact	
  on	
  subsequent	
  goals.	
  
	
  
Goals	
  and	
  Outcome-­‐Oriented	
  Effort	
  
	
  

Not	
  only	
  does	
  the	
  literature	
  support	
  the	
  impact	
  of	
  performance	
  on	
  goals,	
  but	
  goals	
  

also	
  impact	
  the	
  effort	
  individuals	
  devote	
  to	
  a	
  task.	
  Self-­‐regulation	
  theories	
  describe	
  goals	
  as	
  
a	
  key	
  motivational	
  mechanism	
  that	
  impacts	
  how	
  much	
  effort	
  individuals	
  allocate	
  in	
  a	
  task	
  
(e.g.,	
  Latham	
  &	
  Locke,	
  1991;	
  Yeo	
  &	
  Neal,	
  2004).	
  Bandura	
  and	
  Cervone	
  (1983)	
  
experimentally	
  and	
  behaviorally	
  examined	
  the	
  effort	
  individuals	
  devoted	
  to	
  a	
  task	
  and	
  
found	
  that	
  when	
  individuals	
  increased	
  goals,	
  they	
  also	
  increased	
  their	
  effort	
  in	
  the	
  
subsequent	
  trial	
  of	
  the	
  task.	
  Converse	
  and	
  colleagues	
  (2014)	
  also	
  examined	
  the	
  
relationship	
  over	
  time	
  and	
  found	
  a	
  within-­‐person	
  positive	
  relationship	
  where	
  previous	
  goal	
  
level	
  influenced	
  subsequent	
  effort	
  levels.	
  In	
  the	
  adaptation	
  process	
  that	
  occurs	
  right	
  after	
  a	
  

	
  

63	
  

	
  
novel	
  change	
  is	
  introduced,	
  it	
  is	
  expected	
  that	
  goals	
  and	
  effort	
  will	
  have	
  a	
  weak	
  but	
  positive	
  
relationship	
  since	
  the	
  reason	
  for	
  the	
  change	
  will	
  be	
  initially	
  unknown	
  and	
  goals	
  may	
  not	
  be	
  
calibrated	
  appropriately.	
  
Hypothesis	
  15a:	
  In	
  the	
  adaptation	
  process,	
  goals	
  will	
  have	
  a	
  weak	
  and	
  
positive	
  impact	
  on	
  subsequent	
  outcome-­‐oriented	
  effort	
  behaviors.	
  
	
  
However,	
  as	
  individuals	
  have	
  more	
  exposure	
  to	
  the	
  environment	
  and	
  shift	
  to	
  a	
  
performance	
  process	
  where	
  performance	
  is	
  increasing,	
  goals	
  will	
  be	
  more	
  effectively	
  set	
  
and	
  will	
  likely	
  be	
  a	
  better	
  estimate	
  of	
  what	
  the	
  individual	
  can	
  realistically	
  accomplish,	
  
resulting	
  in	
  a	
  stronger	
  positive	
  relationship	
  with	
  effort	
  devoted	
  to	
  those	
  goals.	
  This	
  is	
  
supported	
  by	
  the	
  model	
  discussed	
  by	
  Kluger	
  and	
  DeNisi	
  (1996)	
  that	
  suggests	
  that	
  
individuals	
  engage	
  in	
  different	
  behaviors	
  based	
  on	
  their	
  response	
  to	
  the	
  feedback-­‐standard	
  
discrepancy.	
  This	
  claim	
  was	
  supported	
  by	
  self-­‐regulation	
  researchers	
  who	
  state	
  that	
  
individuals	
  will	
  devote	
  the	
  effort	
  that	
  is	
  needed	
  to	
  goals	
  that	
  they	
  are	
  committed	
  to	
  and	
  
pursuing	
  (Bandura	
  &	
  Locke,	
  2003;	
  Sitzmann	
  &	
  Ely,	
  2011).	
  DeShon,	
  Kozlowski,	
  Schmidt,	
  
Milner	
  &	
  Wiechmann	
  (2004)	
  experimentally	
  investigated	
  the	
  impact	
  of	
  goals	
  on	
  
subsequent	
  effort	
  in	
  a	
  computer	
  simulation	
  task	
  over	
  multiple	
  trials	
  and	
  determined	
  that	
  
goals	
  were	
  positively	
  related	
  to	
  subsequent	
  self-­‐focused	
  effort	
  (or	
  effort	
  that	
  was	
  related	
  to	
  
performing	
  the	
  task).	
  This	
  effort	
  was	
  then	
  related	
  to	
  increases	
  in	
  performance	
  levels.	
  
Therefore,	
  it	
  is	
  expected	
  that	
  when	
  there	
  is	
  an	
  increase	
  in	
  performance,	
  there	
  would	
  be	
  a	
  
positive	
  relationship	
  between	
  goals	
  and	
  effort	
  such	
  that	
  goals	
  would	
  increase,	
  resulting	
  in	
  
more	
  effort	
  devoted	
  to	
  that	
  desired	
  outcome	
  level.	
  
Hypothesis	
  15b:	
  In	
  the	
  performance	
  process,	
  goals	
  will	
  have	
  a	
  strong	
  and	
  
positive	
  impact	
  on	
  subsequent	
  outcome-­‐oriented	
  effort	
  behaviors.	
  
	
  

64	
  

	
  
Outcome-­‐Oriented	
  Effort	
  and	
  Performance	
  	
  
	
  

The	
  final	
  relationship	
  in	
  the	
  motivational	
  cycle	
  is	
  the	
  impact	
  effort	
  has	
  on	
  

performance	
  (see	
  Figures	
  3	
  and	
  4).	
  Self-­‐regulatory	
  research	
  has	
  found	
  strong	
  support	
  for	
  
the	
  relationship	
  between	
  effort	
  and	
  performance.	
  Schmidt,	
  Dolis	
  &	
  Tolli	
  (2009)	
  found	
  that	
  
individuals	
  were	
  more	
  likely	
  to	
  devote	
  time	
  and	
  energy	
  toward	
  goals	
  that	
  were	
  influenced	
  
by	
  an	
  unpredicted	
  event.	
  Yeo	
  and	
  Neal	
  (2004)	
  also	
  found	
  that	
  the	
  relationship	
  between	
  
effort	
  and	
  performance	
  increased	
  initially	
  in	
  a	
  novel	
  task.	
  Other	
  researchers	
  suggest	
  the	
  
relationship	
  becomes	
  weaker	
  over	
  time	
  (e.g.,	
  Kanfer	
  &	
  Ackerman,	
  1989),	
  but	
  Yeo	
  and	
  Neal	
  
suggest	
  that	
  this	
  is	
  due	
  to	
  the	
  lack	
  of	
  effective	
  strategy	
  or	
  the	
  lack	
  of	
  strategic	
  effort,	
  not	
  the	
  
raw	
  effort	
  devoted	
  to	
  performing	
  the	
  task.	
  Good	
  and	
  Michel	
  (2013)	
  investigated	
  the	
  impact	
  
of	
  outcome-­‐oriented	
  effort	
  among	
  undergraduate	
  students	
  performing	
  a	
  decision-­‐making	
  
task	
  in	
  an	
  adaptive	
  environment.	
  They	
  found	
  that	
  the	
  effort	
  had	
  a	
  weak	
  positive	
  
relationship	
  with	
  adaptive	
  performance,	
  although	
  their	
  measurement	
  of	
  this	
  type	
  of	
  effort	
  
was	
  through	
  an	
  individual	
  difference	
  (i.e.,	
  the	
  amount	
  to	
  which	
  individuals	
  are	
  prone	
  to	
  be	
  
focused	
  in	
  their	
  attention).	
  These	
  studies	
  suggest	
  that,	
  when	
  individuals	
  are	
  faced	
  with	
  an	
  
unexpected	
  change,	
  they	
  will	
  put	
  forth	
  more	
  effort	
  to	
  achieve	
  their	
  previous	
  level	
  of	
  
performance	
  (see	
  Figure	
  2),	
  but	
  this	
  effort	
  may	
  not	
  be	
  well-­‐calibrated,	
  as	
  the	
  reason	
  for	
  
performance	
  changes	
  may	
  not	
  be	
  initially	
  clear.	
  Therefore,	
  it	
  is	
  expected	
  that	
  effort	
  and	
  
performance	
  will	
  have	
  a	
  weak	
  and	
  positive	
  relationship	
  in	
  adaptive	
  environments.	
  
Hypothesis	
  16a:	
  In	
  the	
  adaptation	
  process,	
  outcome-­‐oriented	
  effort	
  
behaviors	
  will	
  have	
  a	
  weak	
  and	
  positive	
  impact	
  on	
  subsequent	
  performance.	
  
	
  
The	
  studies	
  discussed	
  above	
  have	
  shown	
  that	
  effort	
  and	
  performance	
  have	
  an	
  

	
  

65	
  

	
  
inherently	
  positive	
  relationship,	
  even	
  in	
  the	
  face	
  of	
  a	
  novel	
  change.	
  This	
  is	
  also	
  based	
  on	
  
some	
  of	
  Campbell’s	
  original	
  work	
  describing	
  performance	
  as	
  being	
  impacted	
  by	
  the	
  
motivation	
  of	
  the	
  individual	
  (Campbell,	
  1990).	
  In	
  this	
  work,	
  he	
  presented	
  a	
  framework	
  
where	
  the	
  motivation	
  of	
  the	
  individual	
  was	
  based	
  on	
  the	
  goal	
  choice,	
  the	
  level	
  of	
  effort,	
  and	
  
the	
  persistence	
  of	
  effort.	
  These	
  factors	
  then	
  impacted	
  the	
  performance	
  level	
  the	
  individuals	
  
were	
  able	
  to	
  achieve.	
  This	
  principle	
  was	
  reiterated	
  in	
  more	
  recent	
  work	
  summarizing	
  the	
  
relationship	
  between	
  effort	
  and	
  performance,	
  with	
  Campbell	
  specifically	
  stating	
  that	
  effort	
  
is	
  so	
  inherently	
  tied	
  to	
  performance	
  that	
  the	
  two	
  should	
  not	
  be	
  separated	
  (Campbell,	
  2012).	
  
Instead,	
  he	
  proposes	
  that	
  effort,	
  as	
  a	
  distinct	
  entity,	
  should	
  be	
  discussed	
  as	
  persistence	
  or	
  
extra	
  effort	
  that	
  is	
  devoted	
  to	
  a	
  task.	
  This	
  strong	
  positive	
  relationship	
  between	
  effort	
  
behaviors	
  and	
  performance	
  outcomes	
  is	
  supported	
  by	
  empirical	
  evidence	
  that	
  suggests	
  
that	
  increasing	
  effort	
  behaviors	
  will	
  increase	
  performance	
  (Baard,	
  2013;	
  DeShon	
  et	
  al.,	
  
2004).	
  Given	
  the	
  strong	
  relationship	
  between	
  effort	
  and	
  performance	
  in	
  the	
  literature,	
  it	
  is	
  
expected	
  that	
  once	
  individuals	
  understand	
  the	
  source	
  of	
  the	
  change	
  and	
  choose	
  a	
  new	
  
strategy,	
  they	
  will	
  transition	
  to	
  a	
  performance	
  process	
  where	
  effort	
  and	
  performance	
  will	
  
be	
  more	
  strongly	
  related,	
  given	
  that	
  the	
  impact	
  of	
  effort	
  is	
  more	
  clearly	
  understood.	
  
Hypothesis	
  16b:	
  In	
  the	
  performance	
  process,	
  outcome-­‐oriented	
  effort	
  
behaviors	
  will	
  have	
  a	
  strong	
  and	
  positive	
  impact	
  on	
  subsequent	
  performance.	
  
	
  
Performance	
  and	
  Self-­‐efficacy	
  
	
  

It	
  would	
  be	
  conceptually	
  incomplete	
  to	
  neglect	
  a	
  discussion	
  of	
  the	
  impact	
  self-­‐

efficacy	
  has	
  on	
  the	
  motivational	
  cycle.	
  The	
  literature	
  on	
  self-­‐efficacy	
  and	
  performance	
  
reflects	
  a	
  very	
  complex	
  relationship.	
  The	
  classic	
  between-­‐person	
  theory	
  suggests	
  that	
  

	
  

66	
  

	
  
increases	
  in	
  self-­‐efficacy	
  lead	
  to	
  increased	
  performance	
  (Bandura,	
  1991).	
  However,	
  within-­‐
person	
  empirical	
  findings	
  suggests	
  that	
  there	
  is	
  a	
  negative	
  within-­‐person	
  effect	
  (Vancouver	
  
et	
  al.,	
  2001;	
  Yeo	
  &	
  Neal,	
  2008)	
  such	
  that	
  increases	
  in	
  self-­‐efficacy	
  result	
  in	
  lower	
  
performance	
  due	
  to	
  individuals	
  having	
  a	
  sense	
  of	
  overconfidence	
  in	
  their	
  ability	
  and	
  
devoting	
  less	
  effort	
  toward	
  their	
  goals.	
  More	
  recent	
  within-­‐person	
  research	
  suggests	
  that	
  
this	
  effect	
  is	
  true	
  only	
  when	
  the	
  environment	
  is	
  highly	
  ambiguous.	
  However,	
  there	
  is	
  a	
  
positive	
  relationship	
  between	
  self-­‐efficacy	
  and	
  performance	
  when	
  feedback	
  information	
  is	
  
provided	
  (Schmidt	
  &	
  DeShon,	
  2010).	
  Other	
  researchers	
  have	
  also	
  suggested	
  that	
  self-­‐
efficacy	
  impacts	
  performance	
  through	
  calibrating	
  their	
  goals	
  and	
  that	
  performance	
  serves	
  
as	
  a	
  way	
  to	
  calibrate	
  subsequent	
  self-­‐efficacy	
  (Sitzmann	
  &	
  Yeo,	
  2013;	
  Tolli	
  &	
  Schmidt,	
  
2008).	
  This	
  is	
  also	
  supported	
  by	
  within-­‐individual	
  analyses	
  (e.g.,	
  Seo	
  &	
  Ilies,	
  2009).	
  	
  In	
  the	
  
adaptation	
  literature,	
  Chen	
  et	
  al.	
  (2005),	
  using	
  an	
  adaptation	
  performance	
  scenario,	
  found	
  
that	
  self-­‐efficacy	
  was	
  related	
  to	
  adaptive	
  performance	
  through	
  impacting	
  goal	
  choice	
  and	
  
goal	
  striving	
  activities	
  such	
  that	
  more	
  efficacious	
  individuals	
  set	
  more	
  challenging	
  goals	
  
and	
  pursued	
  them	
  with	
  more	
  effort.	
  Several	
  studies	
  have	
  shown	
  that	
  self-­‐efficacy	
  also	
  
positively	
  impacts	
  adaptive	
  performance	
  directly	
  (e.g.,	
  Bell	
  &	
  Kozlowski,	
  2008;	
  Kozlowski,	
  
Gully,	
  Brown,	
  Salas,	
  Smith	
  &	
  Nason,	
  2001).	
  Some	
  measure	
  self-­‐efficacy	
  as	
  a	
  confidence	
  in	
  
their	
  ability	
  to	
  perform	
  in	
  an	
  adaptive	
  setting	
  and	
  found	
  that	
  it	
  does	
  predict	
  adaptive	
  
performance	
  (Pulakos	
  et	
  al.,	
  2002;	
  Griffin	
  &	
  Hesketh,	
  2005;	
  White	
  et	
  al.,	
  2005).	
  As	
  
evidenced	
  in	
  the	
  literature	
  above,	
  it	
  is	
  more	
  typical	
  for	
  self-­‐efficacy	
  to	
  be	
  predictive	
  of	
  
performance,	
  rather	
  than	
  performance	
  on	
  self-­‐efficacy.	
  However,	
  it	
  may	
  also	
  be	
  that	
  
performance	
  impacts	
  self-­‐efficacy	
  such	
  that	
  as	
  performance	
  increases,	
  it	
  will	
  inform	
  the	
  
individuals	
  about	
  their	
  ability	
  levels	
  and	
  therefore	
  their	
  self-­‐efficacy	
  will	
  directly	
  reflect	
  

	
  

67	
  

	
  
fluctuations	
  in	
  performance.	
  Given	
  that	
  the	
  research	
  by	
  Schmidt	
  and	
  DeShon	
  (2010)	
  
suggest	
  that	
  this	
  positive	
  association	
  is	
  possible	
  in	
  highly	
  complex	
  situations,	
  it	
  is	
  expected	
  
that	
  this	
  would	
  hold	
  true	
  in	
  adaptive	
  settings.	
  
Hypothesis	
  17a:	
  In	
  the	
  adaptation	
  process,	
  performance	
  will	
  have	
  a	
  strong	
  
and	
  positive	
  impact	
  on	
  subsequent	
  self-­‐efficacy.	
  
	
  
It	
  is	
  expected	
  that	
  once	
  individuals	
  start	
  to	
  increase	
  performance	
  and	
  determine	
  an	
  
effective	
  strategy	
  for	
  adaptation,	
  they	
  will	
  shift	
  into	
  a	
  performance	
  environment.	
  A	
  recent	
  
meta-­‐analysis	
  examining	
  the	
  within-­‐person	
  relationship	
  between	
  self-­‐efficacy	
  and	
  
performance	
  shows	
  that	
  although	
  the	
  impact	
  of	
  self-­‐efficacy	
  on	
  subsequent	
  performance	
  is	
  
very	
  small,	
  the	
  influence	
  of	
  performance	
  on	
  subsequent	
  self-­‐efficacy	
  remained	
  at	
  a	
  
moderate	
  level	
  (r=.4;	
  Sitzmann	
  &	
  Yeo,	
  2013).	
  Furthermore,	
  Vancouver	
  and	
  Kendall	
  (2006)	
  
used	
  a	
  cyclical	
  framework	
  to	
  investigate	
  the	
  impact	
  of	
  self-­‐efficacy	
  on	
  performance,	
  and	
  
performance	
  on	
  self-­‐efficacy,	
  and	
  determined	
  that	
  although	
  self-­‐efficacy	
  had	
  a	
  negative	
  
impact	
  on	
  future	
  performance,	
  performance	
  had	
  a	
  positive	
  impact	
  on	
  subsequent	
  self-­‐
efficacy.	
  Therefore,	
  it	
  is	
  expected	
  that	
  the	
  relationship	
  between	
  performance	
  and	
  self-­‐
efficacy	
  will	
  remain	
  positive	
  during	
  the	
  performance	
  process.	
  	
  
Hypothesis	
  17b:	
  In	
  the	
  performance	
  process,	
  performance	
  will	
  have	
  a	
  weak	
  
and	
  positive	
  impact	
  on	
  subsequent	
  self-­‐efficacy.	
  
	
  
Self-­‐efficacy	
  and	
  Goals	
  
	
  

Self-­‐efficacy	
  is	
  the	
  confidence	
  individuals	
  have	
  in	
  their	
  ability	
  to	
  perform	
  a	
  task.	
  This	
  

confidence	
  impacts	
  motivation	
  in	
  two	
  important	
  ways;	
  the	
  first	
  is	
  through	
  impacting	
  goals.	
  

	
  

68	
  

	
  
The	
  self-­‐regulation	
  literature	
  has	
  revealed	
  a	
  complex	
  relationship	
  between	
  self-­‐efficacy	
  and	
  
goals.	
  Bandura	
  (1991)	
  and	
  Zimmerman	
  (1989)	
  suggest	
  that	
  there	
  is	
  a	
  positive	
  association	
  
between	
  these	
  variables	
  as	
  they	
  are	
  a	
  part	
  of	
  a	
  cyclical	
  and	
  internal	
  regulatory	
  system.	
  
However,	
  Vancouver	
  and	
  colleagues	
  (2001)	
  suggest	
  that	
  there	
  is	
  a	
  negative	
  relationship	
  
between	
  these	
  mechanisms	
  over	
  time	
  as	
  individuals	
  become	
  overconfident	
  in	
  their	
  abilities	
  
and	
  set	
  less	
  challenging	
  goals.	
  However,	
  Schmidt	
  and	
  DeShon	
  (2010)	
  suggest	
  that	
  there	
  is	
  a	
  
positive	
  relationship	
  in	
  complex	
  environments	
  where	
  feedback	
  information	
  is	
  given	
  in	
  a	
  
timely	
  manner.	
  Furthermore,	
  in	
  the	
  adaptation	
  literature,	
  Chen	
  et	
  al.	
  (2005)	
  found	
  that	
  the	
  
relationship	
  between	
  self-­‐efficacy	
  and	
  performance	
  is	
  mediated	
  by	
  goal	
  choice	
  and	
  goal	
  
striving	
  activities	
  both	
  at	
  the	
  individual	
  and	
  team	
  levels.	
  Therefore,	
  it	
  is	
  expected	
  that	
  when	
  
individuals	
  are	
  faced	
  with	
  a	
  novel	
  change,	
  self-­‐efficacy	
  and	
  goals	
  will	
  both	
  be	
  negatively	
  
impacted	
  (see	
  Figure	
  2).	
  However,	
  as	
  individuals	
  engage	
  with	
  the	
  task,	
  they	
  will	
  use	
  their	
  
reflection	
  on	
  their	
  efficacy	
  to	
  determine	
  what	
  they	
  deem	
  to	
  be	
  appropriate	
  goals.	
  Thus,	
  self-­‐
efficacy	
  and	
  goals	
  will	
  be	
  strongly	
  and	
  positively	
  related	
  to	
  each	
  other	
  during	
  adaptation	
  
(see	
  Figure	
  4).	
  
Hypothesis	
  18a:	
  In	
  the	
  adaptation	
  process,	
  self-­‐efficacy	
  will	
  have	
  a	
  strong	
  
and	
  positive	
  impact	
  on	
  subsequent	
  goals.	
  
	
  
However,	
  as	
  individuals	
  continue	
  to	
  increase	
  in	
  their	
  performance,	
  both	
  self-­‐efficacy	
  
and	
  goals	
  will	
  reach	
  a	
  point	
  where	
  they	
  will	
  not	
  be	
  fluctuating	
  as	
  much	
  due	
  to	
  individuals	
  
understanding	
  when	
  they	
  have	
  reached	
  their	
  maximum	
  or	
  optimal	
  performance	
  level.	
  As	
  
self-­‐regulation	
  researchers	
  have	
  indicated,	
  achieving	
  the	
  desired	
  performance	
  levels	
  
increases	
  self-­‐efficacy,	
  which	
  then	
  results	
  in	
  creating	
  even	
  higher	
  goals	
  to	
  pursue	
  (Bandura	
  

	
  

69	
  

	
  
&	
  Cervone,	
  1983;	
  Bandura	
  &	
  Locke,	
  2003;	
  Sitzmann	
  &	
  Ely,	
  2011).	
  Given	
  that	
  the	
  present	
  
theory	
  adopts	
  a	
  longitudinal	
  lens	
  where	
  self-­‐efficacy	
  and	
  goals	
  will	
  not	
  be	
  able	
  to	
  increase	
  
continually,	
  it	
  is	
  expected	
  that,	
  although	
  self-­‐efficacy	
  and	
  goals	
  will	
  remain	
  positively	
  
related,	
  they	
  will	
  be	
  so	
  less	
  strongly	
  because	
  they	
  will	
  both	
  likely	
  plateau	
  (see	
  Figure	
  4).	
  
Hypothesis	
  18b:	
  In	
  the	
  performance	
  process,	
  self-­‐efficacy	
  will	
  have	
  a	
  weak	
  
and	
  positive	
  impact	
  on	
  subsequent	
  goals.	
  
	
  
Self-­‐efficacy	
  and	
  Outcome-­‐Oriented	
  Effort	
  
The	
  second	
  way	
  self-­‐efficacy	
  influences	
  motivation	
  is	
  through	
  impacting	
  the	
  effort	
  
individuals	
  devote	
  to	
  a	
  task.	
  The	
  relationship	
  between	
  self-­‐efficacy	
  and	
  effort	
  can	
  be	
  
thought	
  of	
  in	
  two	
  ways:	
  a)	
  individuals	
  are	
  motivated	
  to	
  devote	
  effort	
  toward	
  tasks	
  in	
  which	
  
they	
  are	
  efficacious,	
  or	
  b)	
  individuals	
  become	
  overconfident	
  in	
  tasks	
  in	
  which	
  they	
  feel	
  
confident	
  in	
  their	
  abilities	
  and	
  therefore	
  devote	
  less	
  effort	
  toward	
  them.	
  In	
  early	
  research,	
  
Bandura	
  (1993)	
  suggested	
  that	
  perceptions	
  of	
  high	
  self-­‐efficacy	
  are	
  theoretically	
  associated	
  
with	
  increased	
  motivation	
  to	
  pursue	
  higher	
  goals	
  and	
  putting	
  forth	
  effort	
  to	
  achieve	
  those	
  
goals.	
  This	
  is	
  supported	
  by	
  recent	
  research	
  by	
  Komarraju	
  and	
  Nadler	
  (2013)	
  where	
  
undergraduate	
  students	
  were	
  asked	
  to	
  self-­‐report	
  their	
  self-­‐efficacy,	
  effort,	
  and	
  GPA,	
  and	
  a	
  
positive	
  relationship	
  between	
  self-­‐efficacy	
  and	
  GPA	
  was	
  found	
  with	
  effort	
  serving	
  as	
  the	
  
mediator.	
  Carver	
  &	
  Scheier	
  (1998)	
  also	
  provide	
  some	
  insight	
  into	
  the	
  theoretical	
  dynamics	
  
involved	
  in	
  that	
  they	
  describe	
  effort	
  in	
  the	
  form	
  of	
  behaviors	
  that	
  are	
  performed	
  during	
  a	
  
task.	
  Those	
  behaviors	
  then	
  inform	
  the	
  internal	
  self-­‐regulatory	
  process	
  of	
  the	
  individual	
  (i.e.,	
  
the	
  level	
  of	
  motivation	
  and	
  confidence	
  the	
  individual	
  has),	
  which	
  then	
  informs	
  future	
  effort	
  
behaviors.	
  These	
  studies	
  support	
  the	
  concept	
  that	
  these	
  self-­‐regulatory	
  variables	
  are	
  

	
  

70	
  

	
  
positively	
  related	
  to	
  each	
  other.	
  Therefore,	
  it	
  is	
  expected	
  that,	
  in	
  the	
  adaption	
  process,	
  
there	
  will	
  be	
  a	
  positive	
  relationship	
  between	
  self-­‐efficacy	
  and	
  effort.	
  	
  
Hypothesis	
  19a:	
  In	
  the	
  adaptation	
  process,	
  self-­‐efficacy	
  will	
  have	
  a	
  weak	
  and	
  
positive	
  impact	
  on	
  subsequent	
  outcome-­‐oriented	
  effort	
  behaviors	
  
	
  
The	
  second	
  view	
  of	
  self-­‐efficacy	
  and	
  effort	
  comes	
  into	
  play	
  once	
  individuals	
  have	
  
transitioned	
  to	
  a	
  performance	
  environment.	
  Here,	
  it	
  is	
  expected	
  that	
  the	
  relationship	
  
between	
  these	
  mechanisms	
  would	
  become	
  negative,	
  as	
  individuals	
  are	
  more	
  stagnant	
  in	
  
their	
  efficacy	
  level	
  and	
  possibly	
  overconfident.	
  This	
  is	
  also	
  based	
  on	
  the	
  work	
  of	
  Vancouver	
  
et	
  al.	
  (2001)	
  who	
  found	
  that	
  self-­‐efficacy	
  and	
  performance	
  were	
  negatively	
  related	
  at	
  the	
  
within-­‐person	
  level.	
  He	
  suggests	
  that	
  this	
  is	
  due	
  to	
  individuals	
  being	
  overconfident	
  in	
  their	
  
ability	
  level,	
  resulting	
  in	
  less	
  effort	
  devoted	
  to	
  the	
  goal.	
  Thus,	
  it	
  is	
  expected	
  that	
  higher	
  self-­‐
efficacy	
  would	
  lower	
  effort	
  levels,	
  but	
  lower	
  self-­‐efficacy	
  would	
  result	
  in	
  more	
  effort	
  in	
  a	
  
performance	
  environment	
  (i.e.,	
  when	
  no	
  novel	
  changes	
  are	
  present).	
  
Hypothesis	
  19b:	
  In	
  the	
  performance	
  process,	
  self-­‐efficacy	
  will	
  have	
  a	
  weak	
  
and	
  negative	
  impact	
  on	
  subsequent	
  outcome-­‐oriented	
  effort	
  behaviors

	
  

71	
  

	
  
METHOD	
  
Participants	
  
	
  

The	
  sample	
  consisted	
  of	
  6051	
  undergraduate	
  students	
  from	
  a	
  large	
  midwestern	
  

university.	
  Individuals	
  were	
  recruited	
  from	
  the	
  Psychology	
  Department’s	
  subject	
  pool	
  and	
  
were	
  compensated	
  with	
  credit	
  in	
  their	
  course	
  and	
  the	
  possibility	
  of	
  a	
  reward	
  based	
  on	
  their	
  
performance	
  level.	
  Additional	
  information	
  regarding	
  the	
  reward	
  is	
  provided	
  below.	
  Sixty-­‐
three	
  of	
  these	
  individuals	
  did	
  not	
  return	
  for	
  the	
  second	
  part	
  of	
  the	
  experiment,	
  so	
  they	
  
were	
  eliminated	
  from	
  the	
  final	
  dataset.	
  Eighteen	
  participants	
  were	
  also	
  removed	
  from	
  the	
  
analyses	
  due	
  to	
  their	
  non-­‐compliance.	
  This	
  was	
  seen	
  in	
  either	
  a)	
  the	
  experimenter	
  reported	
  
them	
  as	
  being	
  visibly	
  distracted	
  during	
  the	
  second	
  day	
  of	
  the	
  study	
  (e.g.,	
  listening	
  to	
  music	
  
on	
  headphones	
  while	
  engaging	
  in	
  the	
  task),	
  or	
  b)	
  they	
  did	
  not	
  perform	
  any	
  behaviors	
  
during	
  20%	
  or	
  more	
  of	
  the	
  adaptation	
  trials	
  on	
  Day	
  2	
  (i.e.,	
  they	
  did	
  not	
  click	
  on	
  a	
  single	
  
target	
  or	
  piece	
  of	
  information).	
  This	
  20%	
  cutoff	
  was	
  determined	
  based	
  on	
  a	
  clear	
  difference	
  
in	
  the	
  data	
  from	
  individuals	
  who	
  randomly	
  had	
  one	
  trial	
  of	
  nonresponse	
  versus	
  the	
  other	
  
individuals	
  who	
  fell	
  into	
  the	
  20%	
  category	
  who	
  had	
  multiple	
  nonresponse	
  trials	
  within	
  a	
  
short	
  period	
  of	
  time.	
  Therefore,	
  the	
  final	
  sample	
  consisted	
  of	
  509	
  individuals	
  and	
  16,279	
  
trials.	
  The	
  final	
  sample	
  was	
  comprised	
  of	
  60%	
  males,	
  and	
  spanned	
  the	
  ages	
  of	
  18	
  to	
  24.	
  	
  	
  
A	
  two-­‐part	
  series	
  of	
  simulations	
  was	
  conducted	
  for	
  the	
  longitudinal	
  cross-­‐lag	
  panel	
  
regression	
  model	
  that	
  determined	
  this	
  sample	
  size	
  was	
  appropriate	
  to	
  capture	
  the	
  effects,	
  
given	
  the	
  number	
  of	
  time	
  points	
  in	
  the	
  adaptation	
  and	
  performance	
  environments,	
  and	
  the	
  
anticipated	
  error	
  rates	
  and	
  effect	
  sizes	
  (see	
  Appendix	
  A	
  for	
  a	
  description	
  of	
  the	
  method	
  and	
  
more	
  detailed	
  results	
  from	
  the	
  simulations).	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  
1	
  An	
  additional	
  112	
  individuals	
  participated	
  in	
  a	
  pilot	
  test	
  and	
  the	
  code	
  checked	
  for	
  errors.	
  
Once	
  the	
  task	
  was	
  finalized,	
  data	
  collection	
  began.	
  
	
  

72	
  

	
  
Task	
  
The	
  task	
  used	
  was	
  a	
  computer-­‐based	
  radar-­‐tracking	
  simulation,	
  TANDEM,	
  which	
  is	
  a	
  
decision-­‐making	
  experimental	
  platform	
  that	
  had	
  been	
  utilized	
  in	
  prior	
  research	
  in	
  the	
  
adaptation	
  literature	
  (Kozlowski,	
  Gully	
  et	
  al.,	
  2001;	
  Bell,	
  2002;	
  Bell	
  &	
  Kozlowski,	
  2002a,	
  
2002b;	
  Bell	
  &	
  Kozlowski,	
  2008).	
  Participants	
  are	
  required	
  to	
  view	
  a	
  radar	
  area,	
  detect	
  
targets,	
  identify	
  them	
  as	
  friendly	
  or	
  hostile,	
  and	
  take	
  appropriate	
  actions	
  to	
  protect	
  their	
  
base.	
  As	
  TANDEM	
  is	
  a	
  complex	
  decision-­‐making	
  task,	
  it	
  serves	
  as	
  a	
  platform	
  with	
  high	
  
psychological	
  fidelity	
  to	
  other	
  complex	
  tasks	
  that	
  require	
  information	
  processing.	
  This	
  
experimental	
  task	
  allows	
  the	
  experimenter	
  to	
  control	
  almost	
  every	
  element	
  in	
  the	
  task.	
  One	
  
can	
  manipulate	
  the	
  number	
  of	
  contacts	
  present,	
  the	
  placement	
  of	
  the	
  contacts,	
  the	
  length	
  of	
  
the	
  trial,	
  the	
  information	
  available	
  before	
  and	
  after	
  the	
  trial,	
  and	
  the	
  point	
  allocations.	
  
Figure	
  5	
  shows	
  a	
  representation	
  of	
  this	
  environment.	
  In	
  order	
  to	
  perform	
  well	
  in	
  
this	
  task,	
  participants	
  had	
  to	
  make	
  a	
  series	
  of	
  decisions	
  about	
  the	
  contacts	
  (denoted	
  in	
  
Figure	
  5	
  by	
  the	
  X	
  markings)	
  before	
  making	
  a	
  final	
  decision	
  (i.e.,	
  clear	
  or	
  shoot).	
  The	
  
decision	
  rules	
  were	
  based	
  on	
  three	
  characteristics:	
  type	
  (i.e.,	
  air,	
  surface,	
  or	
  submarine),	
  
class	
  (i.e.,	
  civilian	
  or	
  military),	
  and	
  intent	
  (i.e.,	
  peaceful	
  or	
  hostile).	
  This	
  information	
  was	
  
sought	
  after	
  a	
  contact	
  was	
  selected,	
  or	
  “hooked”.	
  In	
  addition	
  to	
  this	
  set	
  of	
  decisions,	
  the	
  
participants	
  were	
  not	
  allowed	
  to	
  let	
  the	
  contacts	
  penetrate	
  the	
  inner	
  or	
  outer	
  perimeters	
  
(in	
  Figure	
  5,	
  the	
  inner	
  perimeter	
  is	
  indicated	
  with	
  a	
  solid	
  line	
  surrounding	
  the	
  base,	
  while	
  
the	
  invisible	
  outer	
  perimeter	
  is	
  denoted	
  by	
  the	
  dotted	
  line).	
  In	
  order	
  to	
  do	
  this,	
  individuals	
  
utilized	
  the	
  zoom	
  function	
  and	
  prioritized	
  the	
  contacts.	
  Points	
  were	
  gained	
  for	
  correctly	
  
making	
  the	
  four	
  decisions	
  described	
  above,	
  and	
  points	
  were	
  lost	
  for	
  incorrectly	
  making	
  
those	
  decisions	
  and	
  for	
  allowing	
  contacts	
  to	
  cross	
  the	
  perimeters.	
  

	
  

73	
  

	
  
Figure	
  5	
  
Representation	
  of	
  the	
  TANDEM	
  Task	
  Environment	
  

x

x

x

x

x
x

Base

x

x
x

x

x
x

x
x

x
x

x
x
	
  

	
  
Design	
  
	
  

This	
  experiment	
  was	
  a	
  three-­‐condition	
  between-­‐person	
  design	
  using	
  repeated	
  

measures.	
  Individuals	
  were	
  randomly	
  distributed	
  between	
  the	
  control	
  condition	
  (where	
  no	
  
adaptive	
  change	
  was	
  introduced),	
  a	
  component	
  change	
  condition	
  (one	
  type	
  of	
  adaptive	
  
change),	
  or	
  a	
  coordinative	
  change	
  condition	
  (another	
  type	
  of	
  adaptive	
  change).	
  The	
  study	
  
spanned	
  two	
  days.	
  See	
  Appendix	
  B	
  for	
  an	
  overall	
  picture	
  of	
  the	
  flow	
  of	
  the	
  experiment.	
  
	
  
Day	
  1:	
  Training	
  
On	
  Day	
  1,	
  individuals	
  were	
  trained	
  on	
  TANDEM	
  over	
  nine	
  trials.	
  This	
  was	
  done	
  to	
  
ensure	
  that	
  participants	
  in	
  all	
  conditions	
  were	
  trained	
  to	
  proficiency	
  prior	
  to	
  the	
  
introduction	
  of	
  any	
  task	
  changes.	
  The	
  training	
  focused	
  on	
  exploratory	
  learning	
  (with	
  
	
  

74	
  

	
  
general	
  principles	
  guiding	
  learning	
  during	
  the	
  trials)	
  and	
  encouraged	
  errors	
  during	
  the	
  
training	
  period.	
  Given	
  that	
  previous	
  research	
  has	
  consistently	
  shown	
  that	
  these	
  trainings	
  
best	
  prepare	
  individuals	
  to	
  perform	
  effectively	
  when	
  an	
  adaptive	
  change	
  is	
  introduced	
  (see	
  
Bell	
  &	
  Kozlowski,	
  2002;	
  Baard,	
  2013),	
  the	
  training	
  remained	
  the	
  same	
  across	
  conditions.	
  	
  
Familiarization	
  Phase.	
  Upon	
  entering	
  the	
  lab,	
  participants	
  were	
  asked	
  to	
  complete	
  
an	
  informed	
  consent	
  form.	
  Once	
  obtained,	
  a	
  demographics	
  questionnaire	
  was	
  completed.	
  
After	
  all	
  participants	
  were	
  finished,	
  the	
  experimenter	
  gave	
  a	
  demonstration	
  of	
  the	
  task	
  
through	
  a	
  PowerPoint	
  presentation	
  discussing	
  the	
  following	
  topics:	
  how	
  to	
  hook	
  contacts,	
  
engage	
  the	
  zoom	
  function,	
  and	
  determine	
  the	
  sequence	
  in	
  which	
  one	
  makes	
  a	
  decision.	
  
After	
  the	
  demonstration,	
  participants	
  had	
  three	
  minutes	
  to	
  study	
  the	
  manual	
  followed	
  by	
  a	
  
one-­‐minute	
  familiarization	
  trial	
  (from	
  which	
  they	
  would	
  not	
  receive	
  any	
  feedback).	
  The	
  
purpose	
  of	
  that	
  short	
  trial	
  was	
  to	
  expose	
  individuals	
  to	
  the	
  task	
  prior	
  to	
  beginning	
  training.	
  
Training	
  Phase.	
  After	
  the	
  familiarization	
  phase,	
  participants	
  were	
  given	
  
instructions	
  about	
  what	
  they	
  should	
  be	
  learning	
  during	
  the	
  next	
  few	
  trials.	
  This	
  was	
  based	
  
on	
  the	
  work	
  by	
  Bell	
  and	
  Kozlowski	
  (2008)	
  and	
  the	
  same	
  learning	
  objectives	
  and	
  
instructions	
  they	
  established	
  for	
  the	
  TANDEM	
  environment	
  were	
  used	
  (see	
  Appendix	
  C).	
  
The	
  training	
  was	
  given	
  through	
  written	
  direction,	
  separating	
  the	
  nine	
  trials	
  into	
  three	
  
three-­‐trial	
  blocks.	
  In	
  the	
  first	
  block,	
  individuals	
  were	
  instructed	
  to	
  investigate	
  how	
  to	
  
correctly	
  make	
  the	
  four	
  decisions	
  about	
  a	
  contact	
  (type,	
  class,	
  intent,	
  and	
  execution)	
  and	
  
how	
  to	
  navigate	
  the	
  task	
  environment.	
  The	
  second	
  block	
  directed	
  participants	
  to	
  focus	
  on	
  
how	
  to	
  prevent	
  contacts	
  from	
  crossing	
  the	
  perimeters	
  through	
  instruction	
  on	
  how	
  to	
  use	
  
the	
  zoom	
  function,	
  how	
  to	
  identify	
  marker	
  contacts,	
  and	
  how	
  to	
  be	
  prepared	
  for	
  pop-­‐up	
  
contacts	
  that	
  suddenly	
  appear	
  on	
  the	
  screen.	
  Finally,	
  the	
  third	
  block	
  instructed	
  individuals	
  

	
  

75	
  

	
  
in	
  prioritizing	
  contacts	
  and	
  in	
  making	
  tradeoffs	
  between	
  protecting	
  the	
  inner	
  and	
  outer	
  
perimeters.	
  Within	
  each	
  block	
  of	
  three	
  trials,	
  individuals	
  were	
  given	
  two	
  minutes	
  to	
  read	
  
the	
  instructions,	
  then	
  look	
  at	
  the	
  manual	
  (one	
  minute),	
  engage	
  in	
  a	
  trial	
  (four	
  minutes),	
  and	
  
investigate	
  feedback	
  (one	
  minute).	
  This	
  cycle	
  continued	
  until	
  all	
  nine	
  training	
  trials	
  were	
  
completed.	
  At	
  that	
  time	
  participants	
  completed	
  another	
  set	
  of	
  measures	
  containing	
  basic	
  
and	
  strategic	
  knowledge,	
  and	
  state	
  goal	
  orientation.	
  Once	
  those	
  measures	
  were	
  completed,	
  
individuals	
  were	
  dismissed	
  from	
  the	
  experiment	
  for	
  the	
  day.	
  This	
  part	
  of	
  the	
  experiment	
  
took	
  two	
  hours	
  and	
  individuals	
  received	
  course	
  credit	
  for	
  their	
  participation.	
  
	
  
Day	
  2:	
  Performance	
  
On	
  Day	
  2,	
  there	
  was	
  a	
  period	
  for	
  re-­‐learning,	
  followed	
  by	
  five	
  routine	
  performance	
  
trials	
  that	
  were	
  similar	
  to	
  the	
  initial	
  training	
  trials,	
  but	
  the	
  participants	
  were	
  informed	
  that	
  
these	
  were	
  performance	
  trials.	
  After	
  the	
  completion	
  of	
  the	
  routine	
  performance	
  trials,	
  a	
  
series	
  of	
  15	
  “adaptive”	
  trials	
  was	
  introduced.	
  Individuals	
  in	
  the	
  component	
  or	
  coordinative	
  
complexity	
  change	
  conditions	
  were	
  told	
  that	
  “something	
  changed”	
  in	
  their	
  environment	
  
and	
  they	
  would	
  have	
  the	
  remainder	
  of	
  the	
  trials	
  to	
  determine	
  what	
  to	
  do,	
  but	
  the	
  specifics	
  
of	
  the	
  change	
  were	
  not	
  delineated.	
  Individuals	
  in	
  the	
  control	
  condition	
  were	
  told	
  that	
  they	
  
were	
  “entering	
  the	
  second	
  phase	
  of	
  the	
  study”	
  in	
  order	
  to	
  mirror	
  the	
  communication	
  mid-­‐
study,	
  as	
  in	
  the	
  component	
  and	
  coordinative	
  complexity	
  conditions.	
  At	
  this	
  transition	
  point,	
  
individuals	
  were	
  reminded	
  about	
  the	
  reward	
  opportunity	
  for	
  high	
  performance.	
  In	
  
between	
  each	
  trial,	
  both	
  before	
  and	
  after	
  the	
  change,	
  a	
  series	
  of	
  measures	
  was	
  given.	
  
Therefore,	
  for	
  each	
  trial,	
  individuals	
  first	
  had	
  an	
  opportunity	
  to	
  investigate	
  the	
  manual	
  (for	
  
a	
  maximum	
  of	
  one	
  minute),	
  perform	
  the	
  trial	
  (for	
  four	
  minutes),	
  receive	
  feedback	
  on	
  their	
  

	
  

76	
  

	
  
performance	
  (for	
  a	
  maximum	
  of	
  one	
  minute),	
  and	
  respond	
  to	
  the	
  self-­‐report	
  questionnaires	
  
of	
  self-­‐efficacy,	
  goals,	
  effort,	
  and	
  metacognition	
  (for	
  a	
  maximum	
  of	
  two	
  minutes).	
  This	
  was	
  
the	
  cycle	
  for	
  all	
  20	
  trials	
  that	
  the	
  individuals	
  engaged	
  in	
  on	
  Day	
  2.	
  
In	
  the	
  component	
  complexity	
  change	
  condition,	
  when	
  the	
  “adaptive	
  event”	
  
occurred	
  during	
  Trial	
  19,	
  individuals	
  were	
  informed	
  that	
  the	
  task	
  environment	
  changed	
  
and	
  that	
  their	
  goal	
  was	
  to	
  understand	
  the	
  change	
  in	
  order	
  to	
  recover	
  and	
  exceed	
  their	
  pre-­‐
change	
  performance	
  level	
  by	
  the	
  end	
  of	
  the	
  study.	
  These	
  adaptive	
  trials	
  doubled	
  the	
  
number	
  of	
  targets	
  in	
  the	
  scenario.	
  All	
  other	
  features	
  of	
  the	
  task	
  remained	
  constant	
  between	
  
the	
  routine	
  and	
  adaptive	
  trials.	
  
In	
  the	
  coordinative	
  complexity	
  change	
  condition,	
  when	
  the	
  “adaptive	
  event”	
  
occurred	
  during	
  Trial	
  19,	
  individuals	
  were	
  also	
  informed	
  that	
  the	
  task	
  environment	
  
changed	
  and	
  that	
  their	
  goal	
  was	
  to	
  understand	
  the	
  change	
  in	
  order	
  to	
  recover	
  and	
  exceed	
  
their	
  pre-­‐change	
  performance	
  level	
  by	
  the	
  end	
  of	
  the	
  study.	
  In	
  the	
  adaptive	
  trials,	
  
individuals	
  were	
  exposed	
  to	
  a	
  new	
  set	
  of	
  point	
  allocations.	
  The	
  outer	
  perimeter	
  increased	
  
in	
  point	
  value	
  and	
  the	
  number	
  of	
  targets	
  threatening	
  to	
  cross	
  that	
  perimeter	
  increased	
  
(however,	
  the	
  number	
  of	
  targets	
  remained	
  the	
  same).	
  
In	
  the	
  control	
  condition,	
  although	
  no	
  adaptive	
  change	
  was	
  introduced,	
  during	
  Trial	
  
19,	
  individuals	
  were	
  informed	
  that	
  they	
  are	
  entering	
  the	
  final	
  phase	
  of	
  the	
  study	
  (serving	
  as	
  
a	
  placebo	
  set	
  of	
  instructions).	
  All	
  the	
  trials	
  remained	
  the	
  same	
  as	
  the	
  training	
  environment.	
  
Day	
  2	
  took	
  a	
  total	
  of	
  three	
  hours.	
  Individuals	
  received	
  course	
  credit	
  for	
  their	
  
participation	
  and	
  they	
  had	
  an	
  opportunity	
  to	
  win	
  a	
  cash	
  prize	
  for	
  the	
  highest	
  average	
  
performance	
  across	
  all	
  the	
  trials	
  on	
  Day	
  2.	
  	
  As	
  each	
  condition	
  had	
  different	
  trials,	
  the	
  
division	
  of	
  top	
  performers	
  was	
  distributed	
  based	
  on	
  the	
  number	
  of	
  individuals	
  in	
  the	
  

	
  

77	
  

	
  
condition.	
  As	
  the	
  number	
  of	
  individuals	
  assigned	
  to	
  the	
  control	
  condition	
  was	
  half	
  of	
  those	
  
of	
  the	
  other	
  two	
  conditions,	
  the	
  number	
  of	
  individuals	
  receiving	
  prizes	
  was	
  also	
  half	
  of	
  
those	
  in	
  the	
  other	
  two	
  conditions.	
  Thus,	
  the	
  top	
  10	
  individuals	
  (top	
  two	
  of	
  the	
  control	
  
condition,	
  top	
  four	
  of	
  the	
  component	
  condition,	
  and	
  top	
  four	
  of	
  the	
  coordinative	
  condition)	
  
received	
  $50,	
  the	
  following	
  15	
  individuals	
  (three	
  in	
  control,	
  six	
  in	
  component	
  and	
  six	
  in	
  
coordinative)	
  received	
  $20	
  and	
  the	
  next	
  20	
  individuals	
  (four	
  in	
  control,	
  eight	
  in	
  component	
  
and	
  eight	
  in	
  coordinative)	
  received	
  $10.	
  This	
  was	
  intended	
  to	
  provide	
  additional	
  
motivation	
  to	
  remain	
  focused	
  throughout	
  the	
  study.	
  
	
  
Measures	
  
The	
  measures	
  used	
  for	
  each	
  of	
  the	
  key	
  variables	
  involved	
  in	
  the	
  adaptation	
  process,	
  
as	
  well	
  as	
  performance	
  and	
  individual	
  differences,	
  are	
  detailed	
  below.	
  The	
  specific	
  items	
  
can	
  be	
  found	
  in	
  Appendix	
  D,	
  along	
  with	
  additional	
  measures	
  that	
  were	
  used	
  for	
  identifying	
  
when	
  the	
  shift	
  between	
  the	
  adaptation	
  and	
  performance	
  environments	
  occurred.	
  
Metacognition	
  was	
  measured	
  after	
  each	
  performance	
  trial.	
  This	
  measure	
  was	
  
modified	
  from	
  previous	
  research	
  in	
  the	
  TANDEM	
  environment	
  by	
  Bell	
  (2002),	
  which	
  was	
  
adapted	
  from	
  a	
  measure	
  created	
  by	
  Ford	
  et	
  al.	
  (1998).	
  This	
  four-­‐item	
  questionnaire	
  asked	
  
about	
  the	
  strategies	
  used	
  by,	
  and	
  the	
  focus	
  of,	
  individuals	
  in	
  the	
  last	
  scenario.	
  They	
  
responded	
  to	
  the	
  question	
  on	
  a	
  five-­‐point	
  Likert-­‐type	
  scale	
  with	
  1	
  being	
  “never’”	
  and	
  5	
  
being	
  “constantly”.	
  	
  Cronbach’s	
  alpha	
  shows	
  that	
  this	
  measure	
  was	
  reliable	
  over	
  the	
  trials	
  
as	
  they	
  ranged	
  from	
  .864	
  to	
  .960.	
  	
  
Evaluation	
  was	
  measured	
  as	
  a	
  behavioral	
  indicator	
  collected	
  during	
  each	
  of	
  the	
  
trials	
  during	
  the	
  performance	
  trials.	
  This	
  measure	
  assessed	
  the	
  amount	
  of	
  time	
  spent	
  

	
  

78	
  

	
  
viewing	
  feedback	
  about	
  their	
  performance	
  in	
  the	
  last	
  trial.	
  
Learning-­‐oriented	
  effort	
  was	
  measured	
  as	
  a	
  behavioral	
  indicator.	
  It	
  was	
  based	
  on	
  
the	
  effort	
  devoted	
  toward	
  seeking	
  additional	
  information	
  about	
  the	
  task,	
  outside	
  of	
  the	
  trial	
  
itself,	
  through	
  investigating	
  the	
  task	
  manual	
  information	
  available	
  to	
  them.	
  
Self-­‐efficacy	
  was	
  measured	
  after	
  each	
  performance	
  trial	
  with	
  four	
  items	
  adapted	
  
from	
  Ford,	
  et	
  al.	
  (1998).	
  They	
  developed	
  an	
  eight-­‐item	
  self-­‐report	
  measure	
  specifically	
  for	
  
this	
  task	
  paradigm.	
  This	
  measure	
  used	
  a	
  five-­‐point	
  Likert-­‐type	
  scale	
  ranging	
  from	
  “strongly	
  
disagree”	
  (1)	
  to	
  “strongly	
  agree”	
  (5).	
  Cronbach’s	
  alpha	
  shows	
  that	
  this	
  measure	
  was	
  
reliable	
  over	
  the	
  trials	
  as	
  they	
  ranged	
  from	
  .924	
  to	
  .970.	
  
Goals	
  were	
  measured	
  by	
  obtaining	
  information	
  regarding	
  the	
  participants’	
  expected	
  
number	
  of	
  points	
  in	
  the	
  next	
  trial.	
  This	
  was	
  based	
  on	
  previous	
  research	
  in	
  the	
  TANDEM	
  
environment	
  (see	
  Baard,	
  2013;	
  Bell	
  &	
  Kozlowski,	
  2008).	
  	
  
Outcome-­‐oriented	
  effort	
  was	
  captured	
  through	
  the	
  behavioral	
  indicator	
  of	
  the	
  total	
  
amount	
  of	
  effort	
  exerted	
  by	
  the	
  individual	
  (i.e.,	
  how	
  hard	
  an	
  individual	
  is	
  working)	
  through	
  
the	
  number	
  of	
  targets	
  engaged,	
  contacts	
  queried,	
  contacts	
  hooked,	
  zooms,	
  and	
  executions.	
  
There	
  are	
  two	
  components	
  of	
  outcome-­‐oriented	
  effort:	
  strategic	
  effort	
  consisting	
  of	
  
behaviors	
  devoted	
  to	
  the	
  aspects	
  of	
  the	
  task	
  that	
  are	
  relevant	
  to	
  prioritization	
  and	
  resource	
  
allocation	
  (e.g.,	
  zooms	
  and	
  speed	
  queries)	
  and	
  basic	
  effort	
  consisting	
  of	
  behaviors	
  devoted	
  
to	
  the	
  basic	
  principles	
  of	
  the	
  task	
  (e.g.,	
  hooking	
  and	
  executing	
  contacts).	
  The	
  sum	
  of	
  these	
  
behaviors	
  constituted	
  the	
  overall	
  outcome-­‐oriented	
  effort	
  measure.	
  
Performance	
  in	
  TANDEM	
  was	
  dependent	
  on	
  an	
  individual’s	
  ability	
  to	
  complete	
  
several	
  actions:	
  identify	
  contacts	
  within	
  the	
  radar	
  area,	
  make	
  decisions	
  about	
  the	
  type	
  of	
  
contact,	
  and	
  protect	
  the	
  home	
  base	
  by	
  not	
  allowing	
  contacts	
  to	
  cross	
  either	
  the	
  inner	
  or	
  

	
  

79	
  

	
  
outer	
  defensive	
  perimeters.	
  Performance	
  was	
  measured	
  in	
  the	
  same	
  manner	
  as	
  in	
  previous	
  
research	
  using	
  this	
  paradigm	
  (e.g.,	
  Bell	
  &	
  Kozlowski,	
  2008).	
  During	
  training	
  and	
  routine	
  
performance	
  trials,	
  there	
  were	
  30	
  targets,	
  and	
  performance	
  was	
  computed	
  by	
  adding	
  100	
  
points	
  when	
  all	
  four	
  decisions	
  (type,	
  class,	
  intent,	
  final	
  decision)	
  were	
  made	
  correctly,	
  and	
  
by	
  subtracting	
  100	
  points	
  if	
  any	
  one	
  of	
  those	
  decisions	
  was	
  incorrect.	
  Furthermore,	
  10	
  
points	
  were	
  subtracted	
  if	
  a	
  contact	
  crossed	
  the	
  inner	
  or	
  outer	
  defensive	
  perimeter.	
  During	
  
the	
  adaptive	
  trials	
  of	
  the	
  control	
  condition,	
  the	
  details	
  remained	
  the	
  same	
  as	
  the	
  routine	
  
trials.	
  For	
  the	
  component	
  complexity	
  change	
  condition,	
  there	
  were	
  60	
  (as	
  opposed	
  to	
  30)	
  
targets	
  with	
  four	
  threatening	
  to	
  cross	
  the	
  inner	
  perimeter	
  and	
  four	
  the	
  outer,	
  but	
  all	
  the	
  
point	
  distributions	
  remained	
  the	
  same.	
  For	
  the	
  coordinative	
  complexity	
  change	
  
condition,	
  there	
  were	
  30	
  targets,	
  but	
  more	
  targets	
  crossed	
  the	
  perimeters	
  and	
  there	
  was	
  a	
  
shift	
  in	
  the	
  importance	
  of	
  perimeter	
  crossings	
  (175	
  points	
  for	
  visible	
  inner	
  perimeter	
  
intrusions	
  and	
  125	
  points	
  for	
  invisible	
  outer	
  perimeter	
  intrusions).	
  This	
  increase	
  in	
  
complexity	
  was	
  replicated	
  from	
  previous	
  research	
  using	
  this	
  task	
  paradigm	
  in	
  the	
  field	
  of	
  
adaptation.	
  Wood’s	
  (1986)	
  typology	
  of	
  task	
  complexity	
  suggests	
  that	
  the	
  increases	
  in	
  
complexity	
  described	
  above	
  create	
  a	
  novel	
  environment	
  that	
  an	
  individual	
  must	
  adapt	
  to	
  in	
  
order	
  to	
  perform	
  well	
  (Bell	
  &	
  Kozlowski,	
  2008).	
  Table	
  2	
  displays	
  the	
  specifics	
  of	
  the	
  
manipulations	
  for	
  each	
  condition.	
  
TANDEM	
  knowledge	
  tests	
  were	
  given	
  to	
  individuals	
  once	
  the	
  training	
  phase	
  was	
  
completed	
  on	
  Day	
  1.	
  The	
  declarative	
  and	
  procedural	
  knowledge	
  pertaining	
  to	
  the	
  TANDEM	
  
task	
  domain	
  was	
  assessed	
  in	
  order	
  to	
  determine	
  individuals’	
  baseline	
  understandings	
  of	
  
the	
  rules	
  of	
  the	
  task,	
  as	
  well	
  as	
  their	
  objectives	
  or	
  overall	
  goals.	
  Basic	
  knowledge	
  was	
  
measured	
  through	
  the	
  assessment	
  of	
  the	
  basic	
  operating	
  features	
  of	
  the	
  task	
  (e.g.,	
  the	
  

	
  

80	
  

	
  
identification	
  of	
  cues,	
  the	
  decision	
  rules,	
  and	
  other	
  basic	
  operating	
  features).	
  Strategic	
  
knowledge	
  measured	
  individuals’	
  understanding	
  of	
  the	
  resource	
  allocation	
  aspects	
  of	
  the	
  
task	
  (e.g.,	
  prioritization	
  of	
  actions,	
  marker	
  contacts,	
  zooming	
  function).	
  Bell	
  and	
  Kozlowski	
  
(2002b)	
  found	
  that	
  basic	
  and	
  strategic	
  knowledge	
  loaded	
  on	
  two	
  separate	
  factors	
  and	
  the	
  
two-­‐factor	
  representation	
  of	
  knowledge	
  was	
  a	
  better	
  fit	
  to	
  the	
  data	
  than	
  a	
  one	
  factor	
  model.	
  
This	
  suggests	
  that	
  there	
  are	
  two	
  distinct	
  knowledge	
  domains	
  that	
  need	
  to	
  be	
  evaluated	
  in	
  
this	
  test	
  of	
  knowledge	
  acquisition.	
  
Demographics	
  and	
  individual	
  differences	
  information	
  was	
  collected	
  immediately	
  
upon	
  individuals’	
  entering	
  the	
  lab	
  on	
  Day	
  1,	
  and	
  contained	
  other	
  individual	
  difference	
  items	
  
such	
  as	
  year	
  in	
  school,	
  undergraduate	
  major,	
  gender,	
  age,	
  cognitive	
  ability,	
  individual	
  
adaptability,	
  and	
  trait	
  and	
  state	
  goal	
  orientation.	
  These	
  data	
  were	
  gathered	
  in	
  the	
  event	
  
that	
  the	
  measures	
  could	
  provide	
  insight	
  into	
  any	
  odd	
  findings	
  or	
  provide	
  insight	
  into	
  
significant	
  differences	
  between	
  groups;	
  however,	
  as	
  this	
  did	
  not	
  occur,	
  these	
  measures	
  
were	
  not	
  used	
  in	
  the	
  data	
  analyses,	
  but	
  the	
  measurement	
  origins	
  and	
  statistics	
  can	
  be	
  found	
  
at	
  the	
  end	
  of	
  Appendix	
  D.	
  
	
  
Statistical	
  Models	
  
In	
  order	
  to	
  test	
  the	
  hypotheses,	
  I	
  used	
  a	
  variety	
  of	
  analyses.	
  For	
  Hypotheses	
  1,	
  2,	
  and	
  
3,	
  which	
  focused	
  on	
  determining	
  the	
  transition	
  point	
  of	
  the	
  adaptation	
  and	
  performance	
  
phases	
  that	
  occurred	
  after	
  the	
  change	
  was	
  introduced,	
  I	
  employed	
  discontinuous	
  growth	
  
curve	
  models.	
  This	
  model	
  has	
  been	
  used	
  in	
  adaptation	
  research	
  in	
  the	
  investigation	
  of	
  
differences	
  in	
  performance	
  trajectories	
  in	
  pre-­‐	
  and	
  post-­‐change	
  environments	
  (Lang	
  &	
  
Bliese,	
  2009).	
  As	
  the	
  utility	
  of	
  this	
  model	
  is	
  in	
  its	
  ability	
  to	
  provide	
  empirical	
  specification	
  of	
  

	
  

81	
  

	
  
the	
  differences	
  in	
  trajectories	
  in	
  two	
  environments,	
  the	
  extension	
  of	
  this	
  model	
  to	
  the	
  
adaptation	
  and	
  post-­‐change	
  performance	
  environment	
  seems	
  to	
  be	
  within	
  the	
  bounds	
  of	
  
the	
  previous	
  use	
  of	
  this	
  analysis.	
  
Each	
  trajectory	
  specified	
  in	
  Hypotheses	
  4,	
  5,	
  6,	
  and	
  7,	
  which	
  focus	
  on	
  the	
  trajectory	
  
change	
  in	
  the	
  self-­‐regulatory	
  variables	
  between	
  the	
  post-­‐change	
  adaptation	
  and	
  
performance	
  environments,	
  was	
  tested	
  with	
  separate	
  latent	
  growth	
  curve	
  models.	
  There	
  
are	
  two	
  key	
  aspects	
  of	
  latent	
  growth	
  curve	
  modeling:	
  the	
  intercept	
  and	
  the	
  slope.	
  The	
  
intercept	
  indicates	
  the	
  level	
  of	
  the	
  initial	
  score	
  of	
  the	
  average	
  individual	
  on	
  the	
  variable	
  
being	
  tested	
  (e.g.,	
  self-­‐efficacy).	
  The	
  slope	
  indicates	
  the	
  average	
  growth	
  of	
  the	
  average	
  of	
  
individuals	
  from	
  the	
  start	
  to	
  the	
  end	
  of	
  that	
  period	
  of	
  time	
  (e.g.,	
  a	
  growth	
  curve	
  will	
  be	
  
estimated	
  for	
  all	
  trials	
  in	
  the	
  adaptation	
  environment	
  separately	
  from	
  those	
  in	
  the	
  
performance	
  environment).	
  	
  
Hypotheses	
  8	
  through	
  19	
  focused	
  on	
  the	
  relationships	
  between	
  the	
  self-­‐regulatory	
  
variables	
  in	
  the	
  adaptation	
  and	
  performance	
  processes,	
  and	
  were	
  tested	
  with	
  one	
  
longitudinal	
  cross-­‐lag	
  panel	
  regression	
  model.	
  Cross-­‐lag	
  models	
  allow	
  for	
  the	
  investigation	
  
of	
  reciprocal	
  relationships	
  and	
  the	
  specification	
  of	
  different	
  relationships	
  over	
  time.	
  
Therefore,	
  this	
  analysis	
  provides	
  the	
  most	
  complete	
  model	
  for	
  investigating	
  a	
  longitudinal	
  
process	
  as	
  is	
  available	
  in	
  our	
  literature.	
  Cross-­‐lag	
  models	
  have	
  also	
  been	
  referred	
  to	
  as	
  
multivariate	
  autoregressive	
  models,	
  autoregressive	
  cross-­‐lag	
  models,	
  and	
  cross-­‐lagged	
  
panel	
  models	
  (Selig	
  &	
  Little,	
  2012).	
  Below	
  is	
  the	
  general	
  equation	
  for	
  a	
  two	
  variable	
  model:	
  

Xt,n	
  =	
  β11Xt-­‐1,n	
  	
  γ12Yt-­‐1,n	
  +	
  εxt,n	
  
Yt,n	
  =	
  β22Yt-­‐1,n	
  	
  γ21Xt-­‐1,n	
  +	
  εyt,n	
  
	
  

	
  

82	
  

	
  
Figure	
  6	
  shows	
  the	
  base	
  cross-­‐lag	
  model	
  with	
  two	
  variables	
  over	
  two	
  time	
  points.	
  
This	
  model	
  specifies	
  an	
  autoregressive	
  effect	
  of	
  the	
  variable	
  on	
  itself	
  over	
  time,	
  and	
  a	
  cross-­‐
lag	
  relationship	
  of	
  X	
  on	
  Y	
  as	
  well	
  as	
  Y	
  on	
  X.	
  This	
  reciprocal	
  relationship	
  estimation	
  is	
  critical	
  
for	
  investigating	
  processes	
  over	
  time.	
  Figure	
  7	
  presents	
  part	
  of	
  the	
  model	
  that	
  was	
  used	
  to	
  
test	
  this	
  set	
  of	
  hypotheses.	
  The	
  first	
  few	
  trials	
  in	
  the	
  figure	
  represent	
  the	
  adaptation	
  trials	
  
and	
  the	
  final	
  trials	
  representing	
  the	
  performance	
  trials;	
  note	
  the	
  slight	
  change	
  in	
  the	
  
thickness	
  of	
  the	
  lines	
  connecting	
  the	
  trials,	
  indicating	
  these	
  relationships	
  will	
  be	
  allowed	
  to	
  
differ	
  in	
  their	
  estimations	
  across	
  the	
  adaptation	
  and	
  performance	
  processes.	
  
	
  
Table	
  2	
  
Specific	
  Adaptive	
  Manipulations	
  for	
  Each	
  Experimental	
  Condition	
  
	
  

	
  

Total	
  #	
   #	
  Cross	
  
Points:	
  
Points:	
  	
  
Targets	
   Inner/Outer	
   Perimeter	
   Target	
  
Perimeters	
   Crosses	
  
Decisions	
  
30	
  
2	
  inner	
  	
  
-­‐10	
  inner	
  	
  
+100	
  correct	
  	
  
Training	
  
All	
  
2	
  outer	
  
-­‐10	
  outer	
  
-­‐100	
  incorrect	
  
Routine	
  
30	
  
2	
  inner	
  	
  
-­‐10	
  inner	
  	
  
+100	
  correct	
  	
  
All	
  
Performance	
  
2	
  outer	
  
-­‐10	
  outer	
  
-­‐100	
  incorrect	
  
Control	
  
30	
  
2	
  inner	
  	
  
-­‐10	
  inner	
  	
  
+100	
  correct	
  	
  
2	
  outer	
  
-­‐10	
  outer	
  
-­‐100	
  incorrect	
  
Adaptive	
  
Component	
   60	
  
4	
  inner	
  
-­‐10	
  inner	
  	
  
+100	
  correct	
  	
  
Performance	
   Change	
  
4	
  outer	
  
-­‐10	
  outer	
  
-­‐100	
  incorrect	
  
Coordinative	
   30	
  
2	
  inner	
  
-­‐175	
  inner	
   +100	
  correct	
  	
  
Change	
  
6	
  outer	
  
-­‐125	
  outer	
   -­‐100	
  incorrect	
  
	
  
Figure	
  6	
  
Example	
  of	
  a	
  Fully	
  Crossed	
  Longitudinal	
  Cross-­‐Lag	
  Panel	
  Model	
  With	
  Two	
  Variables	
  

Trial"1
X"

"

"
β11""
γ12""

""Trial"2
X"

γ21""
Y"

	
  

β22""

"

"
β11""
γ12""

"Trial"3"
X"

γ21""
Y"

83	
  

β22""

Y"
	
  

	
  
Figure	
  7	
  
Example	
  of	
  the	
  Hypothesized	
  Relationships	
  in	
  the	
  Longitudinal	
  Cross-­‐Lag	
  Panel	
  Regression	
  Model	
  
	
  
	
  	
  	
  	
  	
  	
  Adaptation	
  Process	
  
	
  
	
  
	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
   	
  	
  	
  	
  	
  	
  	
  	
  Performance	
  Process	
  
Trial*1

	
  

*

*

*

*

**Trial*2

*

Performance*

Performance*

Evalua/on*

Evalua/on*

Learning*Effort*

Learning*Effort*

Metacogni/on*

Metacogni/on*

Self7efficacy*

Self7efficacy*

Goals*

Goals*

Outcome*Effort*

Outcome*Effort*

*

*

*

*****Trial*3*
**** *

*

…	
  
	
  
	
  
…	
  
Evalua/on*
	
  
	
  
Learning*Effort* …	
  
	
  
	
  
Metacogni/on* …	
  
	
  
	
  
…	
  
Self7efficacy*
	
  
	
  
…	
  
Goals*
	
  
	
  
Outcome*Effort* …	
  
	
  	
  	
  	
  	
  	
  	
  
	
  
	
  
Performance*

84	
  

*

*

*

**Trial*14 *

*

*

*

*****Trial*15*

Performance*

Performance*

Evalua/on*

Evalua/on*

Learning*Effort*

Learning*Effort*

Metacogni/on*

Metacogni/on*

Self7efficacy*

Self7efficacy*

Goals*

Goals*

Outcome*Effort*

Outcome*Effort*

	
  

	
  
RESULTS	
  
Pre-­‐Hypothesis	
  Testing	
  
Variable	
  Information	
  and	
  Data	
  Cleaning	
  
As	
  many	
  of	
  the	
  variables	
  in	
  this	
  dataset	
  were	
  collected	
  20	
  to	
  32	
  times,	
  presenting	
  a	
  
correlation	
  matrix	
  in	
  the	
  body	
  of	
  this	
  paper	
  is	
  not	
  feasible.	
  However,	
  as	
  the	
  cross-­‐lag	
  
models	
  are	
  primarily	
  based	
  on	
  the	
  differences	
  between	
  two	
  processes	
  occurring	
  after	
  a	
  
change	
  was	
  introduced	
  in	
  the	
  component	
  and	
  coordinative	
  complexity	
  conditions;	
  
therefore,	
  the	
  means,	
  standard	
  deviations,	
  and	
  correlations	
  are	
  provided	
  based	
  on	
  the	
  
average	
  of	
  the	
  trials	
  in	
  these	
  two	
  processes	
  (see	
  Tables	
  3	
  and	
  4).	
  Only	
  one	
  variable	
  had	
  an	
  
outlier	
  situation	
  based	
  on	
  non-­‐compliance	
  on	
  the	
  part	
  of	
  the	
  participant.	
  This	
  occurred	
  in	
  
the	
  goal	
  variable	
  where	
  four	
  individuals	
  reported	
  impossible	
  goals	
  for	
  a	
  total	
  of	
  five	
  trials	
  
that	
  were	
  far	
  outside	
  the	
  range	
  of	
  the	
  other	
  responses	
  (i.e.,	
  “1000000”,	
  “10000000”	
  and	
  
“2147483647”).	
  Given	
  that	
  it	
  was	
  not	
  my	
  intention	
  to	
  remove	
  this	
  information	
  (and	
  
therefore	
  lose	
  their	
  entire	
  trial	
  data)	
  or	
  substitute	
  it	
  to	
  zero	
  (thereby	
  suggesting	
  that	
  they	
  
had	
  no	
  goal	
  when	
  they	
  reported	
  a	
  very	
  high	
  one),	
  I	
  replaced	
  their	
  responses	
  with	
  the	
  
highest	
  response	
  below	
  one	
  million	
  (i.e.,	
  40,000).	
  This	
  procedure	
  removed	
  the	
  very	
  small	
  
number	
  of	
  outliers	
  while	
  maintaining	
  the	
  integrity	
  of	
  the	
  data	
  suggesting	
  these	
  four	
  
individuals	
  were	
  pursuing	
  a	
  high	
  goal	
  in	
  those	
  five	
  trials.	
  
Metric	
  invariance	
  tests	
  were	
  also	
  conducted	
  on	
  the	
  self-­‐report	
  measures	
  of	
  self-­‐
efficacy	
  and	
  metacognition,	
  given	
  that	
  these	
  measures	
  were	
  collected	
  after	
  every	
  trial	
  of	
  the	
  
study.	
  Therefore,	
  it	
  was	
  necessary	
  to	
  show	
  that	
  the	
  measure	
  did	
  not	
  change	
  over	
  the	
  course	
  
of	
  the	
  22	
  trials.	
  These	
  longitudinal	
  metric	
  invariance	
  tests	
  indicated	
  that	
  self-­‐efficacy	
  and	
  
metacognition	
  did	
  not	
  change	
  over	
  the	
  course	
  of	
  the	
  study.	
  	
  

	
  

85	
  

	
  
Table 3
Descriptives and Correlations: Adaptation Environment
LearningPerformance

Means
Standard Deviations
Performance Correlation

344.433
427.623
1

Sig.
N

LearningCorrelation
Oriented Effort Sig.
N

Metacognition Correlation
Sig.
N

Evaluation

Correlation
Sig.
N

Goals

Correlation
Sig.
N

OutcomeCorrelation
Oriented Effort Sig.
N

Self-efficacy

Correlation
Sig.
N

	
  

413
.011
.821
413
-.029
.555
413
.099*
.045
413
.343**
.000
413
.523**
.000
413
.245**
.000
413

Oriented Effort

Metacognition

11.938
13.654
.011
.821
413
1

3.269
.897
-.029
.555
413
.139**
.005
413
1

413
.139**
.005
413
.217**
.000
413
-.053
.283
413
.152**
.002
413
-.026
.592
413

413
.105*
.034
413
.044
.373
413
-.031
.525
413
.423**
.000
413
86	
  

Evaluation

Goals

20.506 754.322
7.945 767.743
.099*
.343**
.045
.000
413
413
**
.217
-.053
.000
.283
413
413
*
.105
.044
.034
.373
413
413
1
.005
.913
413
413
.005
1
.913
413
413
**
.127
.199**
.010
.000
413
413
.043
.227**
.383
.000
413
413

Outcome-

Self-

Oriented Effort

efficacy

93.605
29.924
.523**
.000
413
.152**
.002
413
-.031
.525
413
.127**
.010
413
.199**
.000
413
1
413
.199**
.000
413

3.798
.726
.245**
.000
413
-.026
.592
413
.423**
.000
413
.043
.383
413
.227**
.000
413
.199**
.000
413
1
413

	
  
Table 4
Descriptives and Correlations: Performance Environment
LearningPerformance

Means
Standard Deviations
Performance
Correlation

426.424
470.365
1

Sig.
N

LearningCorrelation
Oriented Effort Sig.
N

Metacognition

Correlation
Sig.
N

Evaluation

Correlation
Sig.
N

Goals

Correlation
Sig.
N

OutcomeCorrelation
Oriented Effort Sig.
N

Self-efficacy

Correlation
Sig.
N

	
  

413
.115*
.020
413
.041
.409
413
.131**
.008
413
.324**
.000
413
.629**
.000
413
.322**
.000
413

Oriented Effort

Metacognition

4.382
8.955
.115*
.020
413
1

3.095
1.032
.041
.409
413
.086
.080
413
1

413
.086
.080
413
.251**
.000
413
.012
.802
413
.223**
.000
413
.083
.093
413

413
.050
.311
413
.126*
.010
413
-.007
.879
413
.491**
.000
413
87	
  

Evaluation

Goals

13.567 684.312
4.922 808.400
.131**
.324**
.008
.000
413
413
**
.251
.012
.000
.802
413
413
.050
.126*
.311
.010
413
413
1
.013
.790
413
413
.013
1
.790
413
413
*
.115
.186**
.019
.000
413
413
.063
.217**
.201
.000
413
413

Outcome-

Self-

Oriented Effort

efficacy

90.141
32.869
.629**
.000
413
.223**
.000
413
-.007
.879
413
.115*
.019
413
.186**
.000
413
1
413
.257**
.000
413

4.609
.855
.322**
.000
413
.083
.093
413
.491**
.000
413
.063
.201
413
.217**
.000
413
.257**
.000
413
1
413

	
  
First,	
  while	
  investigating	
  configural	
  invariance,	
  tests	
  indicated	
  that	
  the	
  pattern	
  
loadings	
  of	
  both	
  measures	
  remained	
  the	
  same	
  across	
  time	
  (self	
  efficacy:	
  	
  χ2(2829)	
  =	
  
6209.839,	
  p<.000,	
  RMSEA	
  =	
  .048,	
  CFI	
  =	
  .950,	
  SRMR	
  =	
  .025;	
  metacognition:	
  χ2(2829)	
  =	
  
5378.830,	
  p<.000,	
  RMSEA	
  =	
  .042,	
  CFI	
  =	
  .960,	
  SRMR	
  =	
  .034).	
  Next,	
  weak	
  invariance	
  was	
  
tested	
  and	
  although	
  the	
  chi-­‐square	
  difference	
  tests	
  revealed	
  significantly	
  more	
  misfit,	
  the	
  
fit	
  statistics	
  were	
  within	
  acceptable	
  parameters	
  according	
  to	
  Hu	
  and	
  Bentler	
  (1999),	
  thus	
  it	
  
was	
  concluded	
  that	
  the	
  measures	
  have	
  the	
  same	
  unit	
  over	
  time	
  (χ2(2895)	
  =	
  6331.698,	
  
p<.000,	
  RMSEA	
  =	
  .048,	
  CFI	
  =	
  .950,	
  SRMR	
  =	
  .031;	
  metacognition:	
  χ2(2895)	
  =	
  5467.934,	
  
p<.000,	
  RMSEA	
  =	
  .041,	
  CFI	
  =	
  .959,	
  SRMR	
  =	
  .033).	
  Strong	
  metric	
  invariance	
  also	
  showed	
  
significantly	
  more	
  misfit	
  with	
  the	
  chi-­‐square	
  different	
  tests;	
  however,	
  the	
  fit	
  statics	
  
remained	
  adequate	
  and	
  so	
  the	
  measures	
  were	
  considered	
  to	
  have	
  the	
  same	
  origins	
  over	
  
time	
  (self-­‐efficacy:	
  χ2(2895)	
  =	
  5467.934,	
  p<.000,	
  RMSEA	
  =	
  .041,	
  CFI	
  =	
  .959,	
  SRMR	
  =	
  .033;	
  
metacognition:	
  χ2(2983)	
  =	
  5787.318,	
  p<.000,	
  RMSEA	
  =	
  .043,	
  CFI	
  =	
  .956,	
  SRMR	
  =	
  .038).	
  
Finally,	
  the	
  test	
  of	
  strict	
  invariance,	
  investigating	
  whether	
  the	
  measures	
  had	
  the	
  same	
  
variance	
  over	
  time,	
  was	
  also	
  considered	
  as	
  having	
  significantly	
  more	
  misfit	
  but	
  still	
  within	
  
acceptable	
  parameters	
  (self-­‐efficacy:	
  χ2(3071)	
  =	
  7205.827,	
  p<.000,	
  RMSEA	
  =	
  .051,	
  CFI	
  =	
  
.939,	
  SRMR	
  =	
  .044;	
  metacognition:	
  χ2(3071)	
  =	
  6292.876,	
  p<.000,	
  RMSEA	
  =	
  .045,	
  CFI	
  =	
  .949,	
  
SRMR	
  =	
  .044).	
  	
  
These	
  tests	
  show	
  support	
  that	
  a	
  single	
  latent	
  factor	
  of	
  both	
  self-­‐efficacy	
  and	
  
metacognition	
  can	
  be	
  used	
  to	
  examine	
  the	
  relationships	
  between	
  these	
  variables	
  and	
  
others	
  over	
  the	
  course	
  of	
  the	
  study.	
  Therefore,	
  the	
  average	
  of	
  the	
  items	
  of	
  these	
  scales	
  will	
  
be	
  used	
  for	
  the	
  remainder	
  of	
  the	
  analyses.	
  	
  
	
  

88	
  

	
  
Test	
  of	
  Training	
  Effectiveness	
  
	
  

As	
  described	
  in	
  the	
  methodology,	
  training	
  and	
  initial	
  performance	
  were	
  

administered	
  on	
  Day	
  1,	
  while	
  performance	
  in	
  the	
  adaptive	
  environment	
  occurred	
  on	
  Day	
  2.	
  
This	
  research	
  design	
  was	
  chosen,	
  as	
  the	
  length	
  of	
  the	
  study	
  would	
  be	
  such	
  that	
  participants	
  
would	
  not	
  have	
  been	
  likely	
  to	
  remain	
  engaged	
  for	
  the	
  duration	
  of	
  the	
  study	
  had	
  both	
  days	
  
been	
  combined.	
  However,	
  this	
  separation	
  merits	
  a	
  brief	
  investigation	
  of	
  1)	
  whether	
  there	
  
was	
  any	
  decay	
  of	
  knowledge	
  between	
  the	
  two	
  days	
  (there	
  was	
  a	
  48-­‐hour	
  period	
  of	
  
separation	
  between	
  the	
  two	
  parts	
  of	
  the	
  experiment),	
  and	
  2)	
  whether	
  performance	
  
plateaued	
  before	
  individuals	
  were	
  faced	
  with	
  the	
  change.	
  
	
  

In	
  addressing	
  the	
  first	
  concern,	
  mean	
  differences	
  were	
  examined	
  to	
  see	
  if	
  knowledge	
  

on	
  Day	
  2	
  was	
  different	
  from	
  Day	
  1.	
  All	
  three	
  conditions	
  were	
  combined	
  for	
  these	
  analyses	
  
given	
  that	
  training	
  and	
  initial	
  performance	
  trials	
  were	
  identical.	
  Knowledge	
  actually	
  
increased	
  from	
  Day	
  1	
  to	
  Day	
  2	
  (t	
  =	
  -­‐4.504;	
  df	
  =	
  1016;	
  p	
  <.	
  000;	
  mean	
  score	
  Day	
  1	
  =	
  14.130;	
  
mean	
  score	
  Day	
  2	
  =	
  15.183).	
  This	
  suggests	
  that,	
  although	
  there	
  may	
  have	
  been	
  a	
  test-­‐retest	
  
effect,	
  they	
  remembered	
  the	
  information	
  they	
  learned	
  during	
  training.	
  Therefore,	
  I	
  
concluded	
  that	
  the	
  training	
  from	
  Day	
  1	
  was	
  valid	
  on	
  Day	
  2.	
  
	
  

To	
  investigate	
  the	
  second	
  concern	
  regarding	
  the	
  plateauing	
  of	
  performance	
  prior	
  to	
  

the	
  adaptive	
  change	
  on	
  Day	
  2,	
  a	
  discontinuous	
  growth	
  curve	
  model	
  was	
  fitted	
  to	
  the	
  data,	
  
separating	
  Day	
  1	
  training	
  and	
  performance	
  trials	
  from	
  Day	
  2	
  non-­‐adaptive	
  performance	
  
trials.	
  The	
  results	
  of	
  this	
  model	
  show	
  that	
  there	
  was	
  a	
  significantly	
  positive	
  growth	
  curve	
  
during	
  Day	
  1	
  training	
  and	
  performance	
  (t(8317)=37.882,	
  p<.000;	
  see	
  Table	
  5	
  for	
  the	
  
specific	
  parameter	
  estimates),	
  while	
  there	
  was	
  a	
  significant	
  decrease	
  in	
  performance	
  on	
  
Day	
  2	
  (t(8317)=-­‐3.130,	
  p=.002).	
  This	
  latter	
  trajectory	
  is	
  slightly	
  confusing	
  as	
  it	
  was	
  not	
  	
  

	
  

89	
  

	
  
Table	
  5	
  
Discontinuous	
  Growth	
  Curve	
  Analysis	
  of	
  Training	
  and	
  Performance	
  
	
  
Condition	
  
Day	
  1	
  Training	
  
Transition	
  to	
  
Day	
  2	
  Routine	
  
Day	
  2	
  
Performance	
  
Value:	
  	
  	
  	
  	
  	
  	
  	
  	
  42.132	
   Value:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  13.694	
   Value:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  -­‐13.617	
  
Std.error:	
  	
  	
  1.112	
  
Std.error:	
  	
  	
  13.704	
   Std.error:	
  	
  	
  4.250	
  
ALL	
  conditions	
  
DF:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  8137	
  
DF:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  8137	
  
DF:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  8137	
  
	
  
t-­‐value:	
  	
  	
  	
  	
  	
  	
  37.882	
   t-­‐value:	
  	
  	
  	
  	
  	
  	
  .872	
  
t-­‐value:	
  	
  	
  	
  	
  	
  	
  -­‐3.130	
  
p-­‐value:	
  	
  	
  	
  	
  	
  .000	
  
p-­‐value:	
  	
  	
  	
  	
  	
  .382	
  
p-­‐value:	
  	
  	
  	
  	
  	
  .002	
  
Value:	
  	
  	
  	
  	
  	
  	
  	
  	
  42.712	
   Value:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  2.654	
   Value:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  -­‐8.972	
  
Std.error:	
  	
  	
  2.602	
  
Std.error:	
  	
  	
  36.750	
   Std.error:	
  	
  	
  10.178	
  
ONLY	
  Control	
  
DF:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  1533	
  
DF:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  1533	
  
DF:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  1533	
  
	
  
t-­‐value:	
  	
  	
  	
  	
  	
  	
  .072	
  
t-­‐value:	
  	
  	
  	
  	
  	
  	
  .072	
  
t-­‐value:	
  	
  	
  	
  	
  	
  	
  -­‐.881	
  
p-­‐value:	
  	
  	
  	
  	
  	
  .000	
  
p-­‐value:	
  	
  	
  	
  	
  	
  .942	
  
p-­‐value:	
  	
  	
  	
  	
  	
  .3782	
  
Value:	
  	
  	
  	
  	
  	
  	
  	
  	
  45.	
  920	
   Value:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  3.124	
   Value:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  -­‐19.924	
  
ONLY	
  
Std.error:	
  	
  	
  1.670	
  
Std.error:	
  	
  	
  23.670	
   Std.error:	
  	
  	
  6.555	
  
Component	
  
DF:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  3468	
  
DF:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  3468	
  
DF:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  3468	
  
Complexity	
  
t-­‐value:	
  	
  	
  	
  	
  	
  	
  27.398	
   t-­‐value:	
  	
  	
  	
  	
  	
  	
  .132	
  
t-­‐value:	
  	
  	
  	
  	
  	
  	
  -­‐3.040	
  
p-­‐value:	
  	
  	
  	
  	
  	
  .000	
  
p-­‐value:	
  	
  	
  	
  	
  	
  .895	
  
p-­‐value:	
  	
  	
  	
  	
  	
  .002	
  
Value:	
  	
  	
  	
  	
  	
  	
  	
  	
  37.648	
   Value:	
  	
  	
  	
  	
  	
  	
  	
  	
  30.845	
   Value:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  -­‐8.897	
  
ONLY	
  
Std.error:	
  	
  	
  1.808	
  
Std.error:	
  	
  	
  25.518	
   Std.error:	
  	
  	
  7.072	
  
Coordinative	
  
DF:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  3130	
  
DF:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  3130	
  
DF:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  3130	
  
Complexity	
  
t-­‐value:	
  	
  	
  	
  	
  	
  	
  20.824	
   t-­‐value:	
  	
  	
  	
  	
  	
  	
  1.209	
   t-­‐value:	
  	
  	
  	
  	
  	
  	
  -­‐1.258	
  
	
  
p-­‐value:	
  	
  	
  	
  	
  	
  .000	
  
p-­‐value:	
  	
  	
  	
  	
  	
  .227	
  
p-­‐value:	
  	
  	
  	
  	
  	
  .208	
  
	
  
expected	
  to	
  decrease	
  over	
  the	
  second	
  day;	
  however,	
  when	
  investigating	
  the	
  differences	
  
between	
  the	
  conditions,	
  it	
  was	
  found	
  that	
  the	
  control	
  and	
  coordinative	
  complexity	
  
conditions	
  had	
  non-­‐significant	
  slopes	
  (Control:	
  t(1533)=-­‐.881,	
  p=.378;	
  Coordinative:	
  
t(3130)=-­‐1.258,	
  p=.208),	
  while	
  the	
  component	
  complexity	
  condition	
  had	
  a	
  significantly	
  
negative	
  slope	
  (t(3468)=-­‐3.040,	
  p=.002).	
  Although	
  this	
  is	
  not	
  ideal,	
  it	
  still	
  provides	
  support	
  
that	
  individuals	
  completed	
  the	
  learning	
  process	
  and	
  reached	
  asymptote	
  in	
  performance	
  
prior	
  to	
  the	
  adaptive	
  change	
  being	
  introduced.	
  This	
  is	
  supported	
  by	
  all	
  three	
  conditions	
  
reaching	
  similar	
  levels	
  both	
  directly	
  after	
  training	
  on	
  Day	
  1	
  (F(1,506)=	
  2.039,	
  p=.154),	
  at	
  
the	
  end	
  of	
  performance	
  on	
  Day	
  1	
  (F(1,506)=	
  .336,	
  p=.563),	
  at	
  the	
  beginning	
  of	
  Day	
  2	
  
(F(1,507)=	
  1.216,	
  p=.271),	
  and	
  just	
  before	
  the	
  change	
  was	
  introduced	
  on	
  Day	
  2	
  (F(1,507)=	
  

	
  

90	
  

	
  
1.477,	
  p=.225).	
  Thus,	
  all	
  conditions	
  appeared	
  to	
  follow	
  the	
  same	
  pattern	
  of	
  learning	
  and	
  
performance,	
  and	
  asymptoting	
  prior	
  to	
  any	
  change	
  being	
  introduced.	
  
	
  
Hypothesis	
  Testing	
  
Identifying	
  the	
  Transition	
  Point	
  	
  
	
  

In	
  order	
  to	
  test	
  Hypothesis	
  1,	
  I	
  examined	
  several	
  methodologies	
  before	
  settling	
  on	
  

conducting	
  a	
  series	
  of	
  discontinuous	
  growth	
  curve	
  models	
  to	
  detect	
  a	
  shift	
  in	
  performance	
  
after	
  the	
  change	
  was	
  introduced.	
  The	
  pattern	
  described	
  in	
  the	
  presentation	
  of	
  the	
  theory	
  
suggested	
  that	
  there	
  would	
  be	
  a	
  strong	
  positive	
  increase	
  in	
  performance	
  during	
  the	
  initial	
  
adaptation	
  process,	
  followed	
  by	
  a	
  weak	
  (or	
  non-­‐significant)	
  slope	
  in	
  the	
  performance	
  
process.	
  First,	
  I	
  attempted	
  to	
  determine	
  this	
  through	
  the	
  self-­‐report	
  questionnaire	
  that	
  
asked	
  individuals	
  to	
  indicate	
  when	
  they	
  comprehended	
  what	
  changed	
  in	
  the	
  environment.	
  
Unfortunately,	
  this	
  question	
  was	
  not	
  helpful	
  as	
  it	
  was	
  consistently	
  misinterpreted	
  as	
  
indicate	
  what	
  changed	
  in	
  the	
  environment	
  since	
  the	
  last	
  trial	
  (rather	
  than	
  from	
  when	
  the	
  
only	
  change	
  was	
  introduced).	
  I	
  then	
  investigated	
  whether	
  the	
  correlations	
  or	
  mean	
  
difference	
  tests	
  between	
  the	
  trials	
  would	
  be	
  insightful;	
  however,	
  they	
  too	
  were	
  
uninformative	
  as	
  all	
  the	
  results	
  were	
  significant.	
  Therefore,	
  as	
  discontinuous	
  growth	
  curve	
  
models	
  can	
  empirically	
  test	
  differences	
  in	
  trajectories,	
  this	
  method	
  was	
  chosen	
  as	
  the	
  best	
  
alternative	
  to	
  determine	
  when	
  a	
  shift	
  occurred	
  from	
  the	
  adaptive	
  to	
  the	
  performance	
  
process,	
  based	
  on	
  the	
  pattern	
  described	
  in	
  the	
  theory.	
  I	
  systematically	
  went	
  through	
  a	
  
series	
  of	
  discontinuous	
  growth	
  curve	
  analyses,	
  modifying	
  only	
  which	
  trial	
  was	
  the	
  
transition	
  trial.	
  As	
  it	
  was	
  expected	
  that	
  the	
  two	
  conditions	
  would	
  have	
  different	
  transition	
  
points	
  (Hypothesis	
  3),	
  the	
  component	
  and	
  complexity	
  conditions	
  were	
  tested	
  separately.	
  	
  

	
  

91	
  

	
  
Based	
  on	
  the	
  expected	
  pattern	
  described	
  in	
  the	
  theory,	
  in	
  the	
  component	
  complexity	
  
condition	
  individuals	
  transitioned	
  from	
  the	
  adaptation	
  process	
  to	
  the	
  performance	
  process	
  
at	
  Trial	
  25.	
  This	
  was	
  based	
  on	
  the	
  first	
  time	
  the	
  following	
  pattern	
  was	
  observed:	
  there	
  was	
  
a	
  positive	
  and	
  significant	
  slope	
  for	
  the	
  trials	
  directly	
  after	
  the	
  change	
  (Trials	
  19	
  through	
  24;	
  
t(3032)=3.776,	
  p<.000),	
  then	
  a	
  non-­‐significant	
  slope	
  for	
  the	
  remaining	
  trials	
  (Trials	
  25	
  
through	
  33;	
  t(3032)=-­‐1.734,	
  p=.083),	
  with	
  a	
  non-­‐significant	
  transition	
  trial	
  (Trial	
  25;	
  
t(3032)=1.624,	
  p=.105).	
  A	
  significant	
  transition	
  trial	
  would	
  indicate	
  that	
  during	
  this	
  trial	
  
there	
  was	
  either	
  a	
  significant	
  event	
  or	
  a	
  significant	
  reaction	
  that	
  resulted	
  in	
  a	
  large	
  increase	
  
or	
  decrease	
  in	
  the	
  variable	
  of	
  interest	
  at	
  that	
  specific	
  point.	
  For	
  example,	
  in	
  Lang	
  and	
  Bliese	
  
(2009)	
  they	
  labeled	
  the	
  transition	
  trial	
  as	
  the	
  moment	
  an	
  adaptive	
  change	
  was	
  introduced	
  
so	
  that	
  the	
  pre-­‐change	
  performance	
  trajectory	
  was	
  compared	
  to	
  the	
  post-­‐change	
  
performance	
  trajectory.	
  Here,	
  however,	
  a	
  non-­‐significant	
  transition	
  trial	
  was	
  desirable	
  
given	
  that	
  there	
  was	
  no	
  reason	
  any	
  particular	
  trial	
  should	
  have	
  significant	
  meaning,	
  since	
  
both	
  of	
  the	
  growth	
  curves	
  being	
  estimated	
  occurred	
  after	
  the	
  change	
  was	
  introduced.	
  The	
  
slopes	
  before	
  and	
  after	
  should	
  show	
  the	
  pattern	
  of	
  significantly	
  positive	
  to	
  non-­‐significant,	
  
revealing	
  that	
  performance	
  plateaued,	
  reaching	
  a	
  somewhat	
  stable	
  equilibrium.	
  Therefore,	
  
both	
  the	
  intercepts	
  and	
  slopes	
  for	
  Hypothesis	
  1a	
  and	
  1b	
  were	
  supported	
  by	
  the	
  component	
  
complexity	
  condition	
  (see	
  Figure	
  8	
  for	
  a	
  depiction	
  of	
  the	
  slope	
  changes	
  and	
  Table	
  6	
  for	
  the	
  
specific	
  parameter	
  estimates).	
  
The	
  coordinative	
  complexity	
  condition,	
  however,	
  did	
  not	
  follow	
  the	
  expected	
  
pattern	
  of	
  performance	
  change	
  and	
  instead	
  showed	
  consistently	
  low	
  performance	
  followed	
  
by	
  a	
  lagged	
  increase	
  in	
  performance.	
  Therefore,	
  in	
  order	
  to	
  appropriately	
  detect	
  the	
  
transition	
  point	
  between	
  the	
  adaptation	
  and	
  performance	
  processes,	
  I	
  applied	
  the	
  same	
  

	
  

92	
  

	
  
logic	
  of	
  a	
  significant	
  slope	
  at	
  one	
  point	
  with	
  a	
  non-­‐significant	
  transition	
  trial	
  and	
  a	
  non-­‐
significant	
  slope	
  at	
  the	
  other	
  point.	
  In	
  the	
  component	
  complexity	
  condition	
  the	
  pattern	
  was	
  
such	
  that	
  the	
  significantly	
  increasing	
  slope	
  was	
  first	
  (in	
  the	
  adaptation	
  environment);	
  
however,	
  the	
  coordinative	
  complexity	
  condition	
  showed	
  that	
  the	
  significantly	
  increasing	
  
slope	
  was	
  in	
  the	
  second	
  environment	
  (i.e.,	
  the	
  performance	
  environment;	
  t(2739)=2.757,	
  
p=.006),	
  while	
  the	
  non-­‐significant	
  slope	
  was	
  in	
  the	
  adaptation	
  environment	
  (t(2739)=-­‐
1.325,	
  p=.185).	
  Similar	
  to	
  the	
  component	
  complexity	
  condition,	
  the	
  coordinative	
  
complexity	
  condition	
  also	
  had	
  a	
  non-­‐significant	
  transition	
  trial	
  (t(2739)=-­‐.359,	
  p=.720).	
  
The	
  analysis	
  also	
  revealed	
  that	
  the	
  transition	
  between	
  these	
  two	
  processes	
  occurred	
  one	
  
trial	
  later	
  than	
  the	
  component	
  complexity	
  condition	
  (see	
  Table	
  6	
  and	
  Figure	
  8).	
  Therefore,	
  
although	
  the	
  intercepts	
  of	
  the	
  trajectories	
  for	
  Hypothesis	
  1a	
  and	
  1b	
  were	
  supported	
  for	
  the	
  
coordinative	
  complexity	
  condition,	
  the	
  slopes	
  were	
  not.	
  Given	
  the	
  difference	
  in	
  transition	
  
points	
  between	
  the	
  two	
  conditions	
  (component	
  at	
  Trial	
  25,	
  coordinative	
  at	
  Trial	
  26),	
  
Hypothesis	
  3	
  was	
  supported.	
  
This	
  led	
  to	
  a	
  question	
  of	
  whether	
  the	
  coordinative	
  complexity	
  condition	
  had	
  a	
  lagged	
  
adaptation	
  process	
  in	
  which	
  case	
  the	
  performance	
  process	
  would	
  not	
  have	
  been	
  measured.	
  
In	
  order	
  to	
  determine	
  this,	
  an	
  additional	
  discontinuous	
  growth	
  model	
  was	
  fitted	
  with	
  self-­‐
reported	
  focus	
  of	
  the	
  type	
  of	
  goal	
  being	
  pursued	
  as	
  the	
  variable	
  of	
  interest.	
  It	
  is	
  logical	
  that,	
  
when	
  individuals	
  shift	
  their	
  focus	
  from	
  learning	
  as	
  much	
  as	
  they	
  could	
  about	
  the	
  task	
  to	
  
determining	
  what	
  changed	
  and	
  how	
  to	
  deal	
  with	
  it	
  (i.e.,	
  exploration	
  behaviors)	
  to	
  focusing	
  
on	
  executing	
  as	
  much	
  as	
  possible	
  to	
  increase	
  performance	
  effectiveness	
  (i.e.,	
  exploitation	
  
behaviors),	
  we	
  then	
  can	
  conclude	
  that	
  there	
  was	
  a	
  transition	
  from	
  the	
  adaptation	
  process	
  
to	
  the	
  performance	
  process,	
  given	
  the	
  theory	
  discussed	
  earlier.	
  In	
  both	
  conditions	
  (with	
  

	
  

93	
  

	
  
Table	
  6	
  

	
  

	
  
	
  

	
  

Results	
  of	
  Discontinuous	
  Growth	
  Curve	
  Analyses	
  for	
  Performance	
  
	
  
	
  
Condition	
  
Adaptation	
  Process	
  
Transition	
  Trial	
  
Performance	
  Process	
  
Parameter:	
  	
  18.857	
   Parameter:	
  	
  34.828	
  
Parameter:	
  	
  -­‐9.844	
  
Component	
  
Std.error:	
  	
  	
  	
  	
  4.993	
  
Std.error:	
  	
  	
  	
  21.451	
  
Std.error:	
  	
  	
  	
  	
  5.676	
  
Complexity	
  
DF:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  3032	
  
DF:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  3032	
  
DF:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  3032	
  
Performance	
   t-­‐value:	
  	
  	
  	
  	
  	
  	
  	
  	
  3.776	
  
t-­‐value:	
  	
  	
  	
  	
  	
  	
  	
  1.624	
  
t-­‐value:	
  	
  	
  	
  	
  	
  	
  	
  	
  -­‐1.734	
  
p-­‐value:	
  	
  	
  	
  	
  	
  	
  	
  .000	
  
p-­‐value:	
  	
  	
  	
  	
  	
  	
  	
  .105	
  
p-­‐value:	
  	
  	
  	
  	
  	
  	
  	
  .083	
  
Parameter:	
  	
  -­‐5.938	
  
Parameter:	
  	
  -­‐8.818	
  
Parameter:	
  	
  15.965	
  
Coordinative	
   Std.error:	
  	
  	
  	
  	
  4.483	
  
Std.error:	
  	
  	
  	
  	
  24.556	
  
Std.error:	
  	
  	
  	
  	
  5.790	
  
Complexity	
  
DF:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  2739	
  
DF:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  2739	
  
DF:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  2739	
  
Performance	
   t-­‐value:	
  	
  	
  	
  	
  	
  	
  	
  	
  -­‐1.325	
  
t-­‐value:	
  	
  	
  	
  	
  	
  	
  	
  	
  -­‐.359	
  
t-­‐value:	
  	
  	
  	
  	
  	
  	
  	
  	
  2.757	
  
p-­‐value:	
  	
  	
  	
  	
  	
  	
  	
  .185	
  
p-­‐value:	
  	
  	
  	
  	
  	
  	
  	
  .720	
  
p-­‐value:	
  	
  	
  	
  	
  	
  	
  	
  .006	
  
	
  	
  
	
  Figure	
  8	
  
Discontinuous	
  Growth	
  Curves	
  for	
  Performance	
  
	
  

Adaptation	
  
Environment	
  

	
  

Performance	
  
Environment	
  

	
  
	
  

Adaptation	
  
Environment	
  

Performance	
  
Environment	
  

	
  

	
  

94	
  

	
  
Table	
  7	
  
Results	
  of	
  Discontinuous	
  Growth	
  Curve	
  Analyses	
  for	
  Type	
  of	
  Effort	
  
	
  
Condition	
  
Adaptation	
  Process	
  
Transition	
  Trial	
  
Performance	
  Process	
  
	
  
Parameter:	
  	
  .081	
  
Parameter:	
  	
  -­‐.184	
  
Parameter:	
  	
  -­‐.141	
  
Component	
  
Std.error:	
  	
  	
  	
  	
  .030	
  
Std.error:	
  	
  	
  	
  .131	
  
Std.error:	
  	
  	
  	
  	
  .035	
  
Complexity	
  
DF:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  3032	
  
DF:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  3032	
  
DF:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  3032	
  
Learning	
  Effort	
   t-­‐value:	
  	
  	
  	
  	
  	
  	
  	
  	
  2.673	
  
t-­‐value:	
  	
  	
  	
  	
  	
  	
  	
  -­‐1.403	
  
t-­‐value:	
  	
  	
  	
  	
  	
  	
  	
  	
  -­‐4.051	
  
	
  
p-­‐value:	
  	
  	
  	
  	
  	
  	
  	
  .008	
  
p-­‐value:	
  	
  	
  	
  	
  	
  	
  	
  .161	
  
p-­‐value:	
  	
  	
  	
  	
  	
  	
  	
  .000	
  
	
  
Parameter	
  	
  .097	
  
Parameter	
  	
  	
  -­‐.290	
  
Parameter	
  	
  	
  -­‐.161	
  
Coordinative	
   Std.error:	
  	
  	
  	
  .0241	
  
Std.error:	
  	
  	
  	
  	
  .132	
  
Std.error:	
  	
  	
  	
  	
  .031	
  
Complexity	
  
DF:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  2739	
  
DF:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  2739	
  
DF:	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  2739	
  
Learning	
  Effort	
   t-­‐value:	
  	
  	
  	
  	
  	
  	
  	
  4.028	
  
t-­‐value:	
  	
  	
  	
  	
  	
  	
  	
  	
  -­‐2.192	
  
t-­‐value:	
  	
  	
  	
  	
  	
  	
  	
  	
  -­‐5.177	
  
	
  
p-­‐value:	
  	
  	
  	
  	
  	
  	
  .000	
  
p-­‐value:	
  	
  	
  	
  	
  	
  	
  	
  .028	
  
p-­‐value:	
  	
  	
  	
  	
  	
  	
  	
  .000	
  
	
  	
  	
  
	
  
Figure	
  9	
  
Discontinuous	
  Growth	
  Curves	
  for	
  Type	
  of	
  Effort	
  

n	
  
io nt	
  
t
ta e
ap o n m
d
A vi r
En

Per
E n f o rm
vi r
onmance	
  
ent
	
  

	
  

n	
   	
  
tio ent
a
t
ap o n m
d
A vi r
En

	
  

	
  

	
  

Per
E n f o rm
vi r
onmance	
  
ent
	
  

95	
  

	
  
their	
  respective	
  transition	
  points,	
  see	
  Figure	
  9	
  and	
  Table	
  7),	
  individuals	
  focused	
  
significantly	
  more	
  on	
  exploration	
  behaviors	
  to	
  learn	
  about	
  the	
  task	
  (the	
  upper	
  end	
  of	
  the	
  
scale)	
  over	
  the	
  course	
  of	
  the	
  adaptation	
  process	
  (component	
  complexity:	
  t(3032)=2.673,	
  
p=.008;	
  coordinative	
  complexity:	
  t(2739)=4.028,	
  p<.000),	
  and	
  then	
  shifted	
  focus	
  onto	
  
exploitation	
  behaviors	
  (the	
  lower	
  end	
  of	
  the	
  scale)	
  to	
  maximize	
  performance	
  output	
  in	
  the	
  
performance	
  process	
  (component	
  complexity:	
  t(3032)=-­‐4.051,	
  p<.000;	
  coordinative	
  
complexity:	
  t(2739)=-­‐5.177,	
  p<.000).	
  This	
  change	
  in	
  strategy,	
  which	
  supports	
  Hypothesis	
  2,	
  
also	
  supports	
  the	
  conclusion	
  that	
  both	
  the	
  component	
  and	
  coordinative	
  complexity	
  
conditions	
  first	
  engaged	
  in	
  an	
  adaptation	
  process	
  (even	
  though	
  performance	
  trajectories	
  
appeared	
  slightly	
  different),	
  and	
  then	
  they	
  transitioned	
  to	
  a	
  performance	
  process.	
  
Therefore,	
  for	
  the	
  remainder	
  of	
  the	
  analyses,	
  the	
  two	
  conditions	
  will	
  be	
  combined,	
  
removing	
  Trial	
  25	
  in	
  order	
  to	
  maintain	
  the	
  consistency	
  of	
  the	
  transition	
  points.	
  
	
  
Examining	
  Trajectory	
  Changes	
  	
  
	
  

In	
  order	
  to	
  test	
  Hypotheses	
  4	
  through	
  7,	
  which	
  focused	
  on	
  the	
  trajectories	
  of	
  the	
  

self-­‐regulatory	
  variables	
  in	
  the	
  adaptation	
  and	
  performance	
  environments,	
  twelve	
  latent	
  
growth	
  curve	
  analyses	
  were	
  conducted	
  to	
  determine	
  the	
  significance	
  of	
  the	
  levels	
  and	
  
slopes	
  of	
  the	
  variables.	
  Both	
  the	
  component	
  and	
  coordinative	
  complexity	
  conditions	
  
showed	
  very	
  similar	
  trajectories	
  and,	
  therefore,	
  the	
  analyses	
  reported	
  are	
  from	
  the	
  
combined	
  data.	
  All	
  significant	
  differences	
  are	
  noted	
  below.	
  Figures	
  10	
  and	
  11	
  display	
  the	
  
trajectories	
  of	
  the	
  combined,	
  component	
  complexity	
  and	
  coordinative	
  complexity	
  
conditions	
  data	
  separately	
  in	
  order	
  to	
  further	
  show	
  the	
  similarities	
  between	
  these	
  
conditions	
  in	
  the	
  self-­‐regulatory	
  variables.	
  Note	
  that	
  the	
  trial	
  numbers	
  refer	
  only	
  to	
  the	
  

	
  

96	
  

	
  
post-­‐change	
  trials;	
  therefore,	
  the	
  transition	
  between	
  the	
  adaptation	
  and	
  performance	
  
environments	
  occurred	
  at	
  Trial	
  7	
  in	
  the	
  component	
  condition	
  and	
  at	
  Trial	
  8	
  in	
  the	
  
coordinative	
  condition.	
  Table	
  8	
  provides	
  a	
  summary	
  of	
  the	
  hypotheses,	
  analyses	
  used	
  for	
  
testing	
  and	
  conclusions	
  based	
  on	
  the	
  results.	
  These	
  conclusions	
  were	
  based	
  the	
  following	
  
standards.	
  Predicted	
  high	
  and	
  mid-­‐level	
  intercepts	
  were	
  considered	
  supported	
  if	
  they	
  were	
  
significantly	
  different	
  from	
  zero.	
  Low-­‐level	
  intercepts	
  were	
  supported	
  by	
  non-­‐significant	
  
intercept	
  estimates.	
  Similarly,	
  predicted	
  strong	
  slopes	
  were	
  supported	
  by	
  significant	
  
estimates;	
  predicted	
  weak	
  slopes	
  were	
  supported	
  by	
  non-­‐significant	
  estimates.	
  	
  
	
  

With	
  regard	
  to	
  the	
  cognitive	
  variables	
  in	
  the	
  adaptation	
  environment,	
  metacognition	
  

was	
  elevated	
  after	
  the	
  change	
  was	
  introduced	
  and	
  did	
  not	
  significantly	
  decrease	
  
throughout	
  the	
  adaptation	
  environment,	
  supporting	
  Hypothesis	
  4a	
  with	
  a	
  high	
  intercept	
  
and	
  a	
  non-­‐significantly	
  negative	
  slope	
  (see	
  Table	
  9	
  for	
  the	
  specific	
  estimates	
  and	
  
significance	
  levels	
  for	
  all	
  cognitive	
  variables).	
  Evaluation	
  and	
  learning-­‐oriented	
  effort	
  both	
  
were	
  initially	
  elevated,	
  but	
  significantly	
  decreased	
  over	
  the	
  adaptation	
  environment.	
  It	
  
should	
  be	
  noted	
  that	
  the	
  decrease	
  was	
  non-­‐significant	
  in	
  the	
  component	
  complexity	
  
condition.	
  Therefore,	
  Hypotheses	
  4b	
  and	
  4c	
  were	
  not	
  supported,	
  however,	
  the	
  intercepts	
  
were	
  high	
  (as	
  expected)	
  but	
  the	
  slopes	
  were	
  in	
  the	
  expected	
  direction	
  (i.e.,	
  negative),	
  
though	
  they	
  were	
  significant	
  as	
  opposed	
  to	
  the	
  predicted	
  non-­‐significant	
  slopes,	
  thus,	
  the	
  
results	
  suggest	
  that	
  the	
  finding	
  is	
  in	
  the	
  appropriate	
  direction.	
  This	
  finding	
  may	
  indicate	
  
that	
  individuals	
  did	
  not	
  feel	
  the	
  need	
  to	
  engage	
  in	
  as	
  much	
  metacognitive	
  or	
  learning-­‐
oriented	
  behaviors	
  in	
  the	
  adaptation	
  environment.	
  

	
  

97	
  

	
  
Table	
  8	
  
Summary	
  of	
  Trajectory	
  Hypotheses,	
  Analyses,	
  and	
  Results	
  
	
  
	
  

Description	
  

#	
  

Performance	
  
Trajectory	
  

1a	
  

	
  

Reason	
  for	
  
Shift	
  in	
  
Environments	
  

1b	
  
2	
  
3	
  
4a	
  
4b	
  

Trajectory	
  
Changes	
  of	
  
Cognitive	
  
Cycle	
  
Mechanisms	
  

4c	
  
5a	
  
5b	
  
5c	
  
6a	
  
6b	
  

Trajectory	
  
Changes	
  of	
  
Motivational	
  
Cycle	
  
Mechanisms	
  

6c	
  
7a	
  
7b	
  
7c	
  

	
  

Hypothesis	
  

Environment	
  

Performance:	
  
Adaptation	
  
	
  Low	
  Intercept,	
  Strong	
  +	
  Slope	
  
Performance:	
  	
  
Performance	
  
High	
  Intercept,	
  Weak	
  +	
  Slope	
  
The	
  environment	
  shift	
  occurs	
  when	
  a	
  strategy	
  is	
  
chosen	
  
Component	
  complexity	
  condition	
  will	
  shift	
  more	
  
quickly	
  
Metacognition:	
  	
  
Adaptation	
  
High	
  Intercept	
  Weak	
  -­‐	
  Slope	
  
Evaluation:	
  	
  
Adaptation	
  
High	
  Intercept	
  Weak	
  –	
  Slope	
  
Learning	
  Effort:	
  	
  
Adaptation	
  
High	
  Intercept,	
  Weak	
  -­‐	
  Slope	
  
Metacognition:	
  	
  
Performance	
  
Mid	
  Intercept,	
  Strong	
  -­‐	
  Slope	
  
Evaluation:	
  	
  
Performance	
  
Mid	
  Intercept,	
  Strong	
  -­‐	
  Slope	
  
Learning	
  Effort:	
  	
  
Performance	
  
Mid	
  Intercept,	
  Strong	
  -­‐Slope	
  
Goals:	
  
Adaptation	
  
Low	
  Intercept,	
  Weak	
  +	
  Slope	
  
Self-­‐efficacy:	
  	
  
Adaptation	
  
Low	
  Intercept,	
  Weak	
  +	
  Slope	
  
Outcome	
  Effort:	
  	
  
Adaptation	
  
Low	
  Intercept,	
  Weak	
  +	
  Slope	
  
Goals:	
  	
  
Performance	
  
Mid	
  Intercept,	
  Strong	
  +	
  Slope	
  
Self-­‐efficacy:	
  	
  
Performance	
  
Mid	
  Intercept,	
  Strong	
  +	
  Slope	
  
Outcome	
  Effort:	
  	
  
Performance	
  
Mid	
  Intercept,	
  Strong	
  +Slope	
  

Analysis	
  

Summary	
  of	
  Results	
  

Component	
  Complexity:	
  Supported	
  
Coordinative	
  Complexity:	
  Not	
  Supported	
  
Component	
  Complexity:	
  Supported	
  
Discontinuous	
   Coordinative	
  Complexity:	
  Not	
  Supported	
  

Growth	
  Curve	
  
Supported	
  
(2	
  models)	
  
Supported	
  

Intercept:	
  Supported
Slope:	
  Supported	
  
Intercept:	
  Supported
Slope:	
  Not	
  Supported	
  
Intercept:	
  Supported
Slope:	
  Not	
  Supported	
  
Intercept:	
  Supported
Slope:	
  Supported	
  
Intercept:	
  Supported
Slope:	
  Supported	
  
Intercept:	
  Supported
Latent	
  Growth	
  
Slope:	
  Supported	
  
Curve	
  
Intercept:	
  Not	
  Supported
(12	
  models)	
  
Slope:	
  Not	
  Supported	
  
Intercept:	
  Not	
  Supported
Slope:	
  Not	
  Supported	
  
Intercept:	
  Not	
  Supported
Slope:	
  Not	
  Supported	
  
Intercept:	
  Supported
Slope:	
  Not	
  Supported	
  
Intercept:	
  Supported
Slope:	
  Not	
  Supported	
  
Intercept:	
  Supported
Slope:	
  Not	
  Supported	
  

98	
  

	
  
Shifting	
  to	
  the	
  performance	
  environment,	
  all	
  cognitive	
  variables	
  (i.e.,	
  metacognition,	
  
evaluation,	
  and	
  learning-­‐oriented	
  effort	
  behaviors)	
  performed	
  as	
  expected.	
  They	
  each	
  had	
  
significantly	
  high	
  initial	
  intercepts,	
  and	
  the	
  trajectories	
  significantly	
  decreased	
  over	
  time.	
  
These	
  findings	
  were	
  consistent	
  across	
  the	
  two	
  conditions	
  and	
  fully	
  support	
  Hypotheses	
  5a,	
  
5b,	
  and	
  5c	
  (see	
  Table	
  9	
  and	
  Figure	
  10).	
  
	
  

With	
  regard	
  to	
  the	
  motivation	
  variables,	
  the	
  trajectories	
  in	
  both	
  environments	
  were	
  

not	
  as	
  expected.	
  Instead	
  of	
  the	
  variables	
  being	
  very	
  low,	
  directly	
  after	
  the	
  change	
  (i.e.,	
  
having	
  an	
  intercept	
  not	
  significantly	
  different	
  from	
  zero),	
  individuals	
  had	
  elevated	
  levels	
  of	
  
self-­‐efficacy	
  and	
  goals	
  (see	
  Table	
  10	
  for	
  the	
  specific	
  estimates	
  and	
  significance	
  levels	
  for	
  all	
  
motivational	
  variables).	
  This	
  was	
  unexpected	
  as	
  it	
  was	
  anticipated	
  that,	
  when	
  faced	
  with	
  a	
  
change,	
  self-­‐efficacy	
  would	
  be	
  low,	
  as	
  confusion	
  and	
  lack	
  of	
  confidence	
  in	
  one’s	
  ability	
  
would	
  set	
  in	
  and	
  goals	
  would	
  follow	
  suit.	
  However,	
  instead	
  of	
  an	
  abrupt	
  drop	
  (as	
  was	
  seen	
  
in	
  performance),	
  there	
  was	
  a	
  significantly	
  negative	
  decrease	
  in	
  these	
  variables	
  throughout	
  
the	
  adaptation	
  environment.	
  Although	
  positive	
  slopes	
  were	
  predicted	
  for	
  self-­‐efficacy	
  and	
  
goals,	
  given	
  the	
  elevated	
  initial	
  levels,	
  significantly	
  negative	
  slopes	
  were	
  reasonable.	
  
Therefore,	
  although	
  Hypotheses	
  6a	
  and	
  6b	
  were	
  not	
  supported,	
  it	
  appears	
  that	
  this	
  may	
  
have	
  been	
  due	
  to	
  initial	
  overconfidence	
  in	
  their	
  ability	
  to	
  adapt,	
  which	
  subsequently	
  
resulted	
  in	
  inflated	
  goals.	
  Outcome-­‐oriented	
  effort	
  behaviors	
  also	
  followed	
  an	
  odd	
  
trajectory	
  in	
  the	
  adaptation	
  environment	
  as	
  they	
  also	
  started	
  significantly	
  elevated,	
  
suggesting	
  that	
  individuals	
  did	
  not	
  pause	
  to	
  learn	
  how	
  to	
  adapt	
  but	
  continued	
  to	
  work	
  hard	
  
despite	
  their	
  effort	
  likely	
  being	
  incorrectly	
  placed.	
  It	
  is	
  possible	
  that	
  this	
  was	
  due	
  to	
  
overconfidence	
  in	
  their	
  ability.	
  Furthermore,	
  the	
  slope	
  of	
  outcome-­‐oriented	
  effort	
  
behaviors	
  in	
  the	
  adaptation	
  environment	
  decreased,	
  not	
  supporting	
  Hypothesis	
  6c,	
  which	
  

	
  

99	
  

	
  
proposed	
  that	
  there	
  would	
  be	
  a	
  low	
  intercept	
  and	
  positive	
  slope,	
  as	
  was	
  found	
  with	
  self-­‐
efficacy	
  and	
  goals.	
  However,	
  it	
  should	
  be	
  noted	
  that	
  the	
  component	
  complexity	
  condition	
  
did	
  not	
  have	
  a	
  significantly	
  decreasing	
  slope,	
  showing	
  that	
  individuals	
  in	
  that	
  condition	
  
(where	
  the	
  number	
  of	
  targets	
  increased	
  twofold)	
  correctly	
  assessed	
  the	
  situation	
  and	
  
recognized	
  that	
  effort	
  had	
  to	
  remain	
  elevated.	
  	
  
	
  

Given	
  the	
  differences	
  in	
  the	
  trajectories	
  in	
  the	
  motivation	
  variables	
  in	
  the	
  adaptation	
  

environment,	
  it	
  is	
  not	
  surprising	
  that	
  they	
  should	
  be	
  different	
  in	
  the	
  performance	
  
environment	
  as	
  well.	
  Hypotheses	
  7a,	
  7b,	
  and	
  7c	
  suggested	
  that	
  self-­‐efficacy,	
  goals,	
  and	
  
outcome-­‐oriented	
  effort	
  would	
  all	
  follow	
  the	
  pattern	
  of	
  a	
  significantly	
  high	
  intercept	
  with	
  a	
  
significantly	
  positive	
  slope.	
  Instead	
  of	
  self-­‐efficacy,	
  goals,	
  and	
  outcome-­‐oriented	
  effort	
  
strongly	
  increasing	
  over	
  that	
  period	
  of	
  time,	
  they	
  stabilized	
  at	
  a	
  lower	
  level	
  for	
  the	
  duration	
  
of	
  the	
  performance	
  environment	
  (see	
  Table	
  10	
  and	
  Figure	
  11).	
  This	
  suggests	
  that	
  although	
  
individuals	
  faced	
  the	
  adaptation	
  trials	
  with	
  overconfidence,	
  they	
  eventually	
  stabilized	
  their	
  
self-­‐efficacy,	
  goals,	
  and	
  effort.	
  Self-­‐efficacy,	
  goals,	
  and	
  effort	
  had	
  levels	
  significantly	
  higher	
  
than	
  zero	
  at	
  the	
  beginning	
  of	
  the	
  performance	
  environment,	
  thereby	
  not	
  supporting	
  
Hypotheses	
  7a,	
  7b,	
  and	
  7c,	
  but	
  providing	
  some	
  promise	
  that	
  the	
  hypotheses	
  may	
  hold	
  true	
  
in	
  other	
  experiments	
  in	
  that	
  the	
  intercept	
  was	
  as	
  expected,	
  while	
  the	
  slopes	
  were	
  not.	
  It	
  
should	
  be	
  noted	
  that	
  self-­‐efficacy	
  and	
  outcome-­‐oriented	
  effort	
  behaviors	
  in	
  the	
  
coordinative	
  complexity	
  condition	
  did	
  significantly	
  decrease	
  in	
  the	
  performance	
  
environment,	
  most	
  likely	
  due	
  to	
  the	
  increased	
  difficulty	
  of	
  the	
  task,	
  and	
  participants	
  being	
  
encouraged	
  to	
  work	
  smarter,	
  not	
  harder.	
  Thus,	
  effort	
  decreasing	
  is	
  not	
  surprising,	
  as	
  in	
  that	
  
condition	
  (where	
  the	
  strategy	
  was	
  shifted)	
  fewer,	
  but	
  more	
  targeted	
  behaviors,	
  led	
  to	
  more	
  
effective	
  outcomes,	
  but	
  these	
  behaviors	
  were	
  more	
  challenging	
  to	
  identify	
  and	
  execute.	
  

	
  

100	
  

	
  
Table	
  9	
  
Latent	
  Growth	
  Curve	
  Analyses	
  of	
  Cognitively	
  Focused	
  Variables	
  
Environment	
  

Adaptation	
  

Performance	
  

Variable	
  
Metacognition	
  
(Hypothesis	
  4a)	
  
Evaluation	
  
(Hypothesis	
  4b)	
  
Learning	
  Effort	
  
(Hypothesis	
  4c)	
  
Metacognition	
  
(Hypothesis	
  5a)	
  
Evaluation	
  
(Hypothesis	
  5b)	
  
Learning	
  Effort	
  
(Hypothesis	
  5c)	
  

Element	
  
Intercept	
  
Slope	
  
Intercept	
  
Slope	
  
Intercept	
  
Slope	
  
Intercept	
  
Slope	
  
Intercept	
  
Slope	
  
Intercept	
  
Slope	
  

Parameter	
  
3.339	
  
-­‐.003	
  
50.286	
  
-­‐1.385	
  
23.158	
  
-­‐.522	
  
3.522	
  
-­‐.014	
  
38.185	
  
-­‐.835	
  
30.303	
  
-­‐.879	
  
	
  
	
  

Std.	
  Er	
  
.161	
  
.008	
  
2.785	
  
.125	
  
4.367	
  
.201	
  
.164	
  
.006	
  
1.930	
  
.062	
  
3.446	
  
.108	
  

DF	
  
2062	
  
2062	
  
2062	
  
2062	
  
2062	
  
2062	
  
2889	
  
2889	
  
2889	
  
2889	
  
2889	
  
2889	
  

t-­‐value	
  
20.705	
  
-­‐.0427	
  
18.059	
  
-­‐11.044	
  
5.302	
  
-­‐2.598	
  
21.421	
  
-­‐2.617	
  
19.789	
  
-­‐13.493	
  
8.794	
  
-­‐8.139	
  

p-­‐value	
  
.000	
  
.669	
  
.000	
  
.000	
  
.000	
  
.010	
  
.000	
  
.001	
  
.000	
  
.000	
  
.000	
  
.000	
  

Figure	
  10	
  
Trajectories	
  of	
  Cognitively	
  Focused	
  Variables	
  
	
  

–	
  Combined	
  	
  
–	
  Component	
  
–	
  Coordinative	
  	
  
	
  

–	
  Combined	
  	
  
–	
  Component	
  
–	
  Coordinative	
  	
  
	
  

–	
  Combined	
  	
  
–	
  Component	
  
–	
  Coordinative	
  	
  
	
  

	
  

101	
  

	
  
Table	
  10	
  
Latent	
  Growth	
  Curve	
  Analysis	
  of	
  Motivationally	
  Focused	
  Variables	
  
Environment	
  

Adaptation	
  

Performance	
  

Variable	
  
Goals	
  
(Hypothesis	
  6a)	
  
Self-­‐efficacy	
  
(Hypothesis	
  6b)	
  
Outcome	
  effort	
  
(Hypothesis	
  6c)	
  
Goals	
  
(Hypothesis	
  7a)	
  
Self-­‐efficacy	
  
(Hypothesis	
  7b)	
  
Outcome	
  effort	
  
(Hypothesis	
  7c)	
  

Element	
  
Intercept	
  

Parameter	
  
2296.626	
  

Std.	
  Er	
  
133.913	
  

DF	
  
2062	
  

t-­‐value	
  
17.150	
  

p-­‐value	
  
.000	
  

Slope	
  
Intercept	
  
Slope	
  
Intercept	
  
Slope	
  
Intercept	
  
Slope	
  
Intercept	
  
Slope	
  
Intercept	
  
Slope	
  

-­‐71.733	
  
5.302	
  
-­‐.070	
  
106.472	
  
-­‐.598	
  
931.561	
  
-­‐8.376	
  
3.770	
  
-­‐.005	
  
96.503	
  
-­‐.216	
  
	
  
	
  

5.700	
  
.134	
  
.007	
  
3.709	
  
.164	
  
365.522	
  
11.870	
  
.153	
  
.005	
  
4.228	
  
.141	
  

2062	
  
2062	
  
2062	
  
2062	
  
2062	
  
2889	
  
2889	
  
2889	
  
2889	
  
2889	
  
2889	
  

-­‐12.584	
  
39.695	
  
-­‐10.573	
  
28.708	
  
-­‐3.644	
  
2.549	
  
-­‐.706	
  
24.684	
  
-­‐1.079	
  
22.825	
  
-­‐1.530	
  

.000	
  
.000	
  
.000	
  
.000	
  
.000	
  
.011	
  
.481	
  
.000	
  
.281	
  
.000	
  
.126	
  

Figure	
  11	
  
Trajectories	
  of	
  Motivationally	
  Focused	
  Variables	
  

–	
  Combined	
  	
  
–	
  Component	
  
–	
  Coordinative	
  	
  
	
  

–	
  Combined	
  	
  
–	
  Component	
  
–	
  Coordinative	
  	
  
	
  

–	
  Combined	
  	
  
–	
  Component	
  
–	
  Coordinative	
  	
  
	
  

	
  
	
  

102	
  

	
  
Understanding	
  Relationship	
  Changes	
  
	
  

Hypotheses	
  8	
  through	
  19	
  focused	
  on	
  the	
  relationships	
  among	
  the	
  self-­‐regulatory	
  

variables	
  in	
  the	
  adaptation	
  and	
  performance	
  processes.	
  All	
  relationships	
  in	
  both	
  processes	
  
were	
  tested	
  in	
  one	
  partially	
  crossed,	
  partially	
  time-­‐varying,	
  cross-­‐lag	
  panel	
  regression	
  
model	
  with	
  seven	
  variables.	
  Just	
  as	
  the	
  two	
  adaptive	
  conditions	
  were	
  combined	
  in	
  the	
  
trajectory	
  analyses,	
  so	
  they	
  were	
  combined	
  in	
  this	
  model,	
  removing	
  Trial	
  25	
  to	
  ensure	
  
consistency	
  of	
  the	
  two	
  processes.	
  This	
  removal	
  was	
  due	
  to	
  the	
  difference	
  in	
  the	
  transition	
  
between	
  the	
  performance	
  and	
  adaptation	
  environments	
  (which	
  put	
  Trial	
  25	
  in	
  the	
  
performance	
  environment	
  for	
  the	
  component	
  condition	
  but	
  in	
  adaptation	
  for	
  the	
  
coordinative	
  condition).	
  Therefore,	
  for	
  consistency	
  in	
  the	
  description	
  of	
  the	
  processes,	
  this	
  
trial	
  had	
  to	
  be	
  removed	
  from	
  this	
  analysis	
  in	
  order	
  to	
  combine	
  the	
  two	
  conditions.	
  
As	
  the	
  variables	
  were	
  of	
  different	
  scales	
  (e.g.,	
  self-­‐efficacy	
  on	
  a	
  five-­‐point	
  Likert	
  scale	
  
versus	
  evaluation	
  of	
  feedback	
  information	
  measured	
  in	
  zero	
  to	
  60	
  seconds	
  versus	
  
performance	
  ranging	
  from	
  about	
  -­‐2,000	
  to	
  +2,000),	
  in	
  order	
  for	
  the	
  maximum	
  likelihood	
  
estimation	
  to	
  work	
  properly,	
  all	
  variables	
  were	
  standardized.	
  The	
  model	
  fit	
  was	
  not	
  
excellent,	
  but	
  will	
  be	
  retained	
  for	
  the	
  purposes	
  of	
  analyzing	
  the	
  theoretical	
  model	
  as	
  well	
  as	
  
possible,	
  particularly	
  since	
  research	
  have	
  not	
  investigated	
  what	
  the	
  norm	
  of	
  cross	
  lag	
  
model	
  fit	
  statistics	
  should	
  be	
  (X2(5408)	
  =	
  19983.202,	
  p<.000,	
  RMSEA	
  =	
  .081,	
  CFI	
  =	
  .691).	
  	
  
Before	
  continuing	
  to	
  the	
  analysis	
  of	
  the	
  hypothesized	
  relationships	
  in	
  the	
  adaptation	
  
and	
  performance	
  processes,	
  I	
  would	
  like	
  to	
  discuss	
  one	
  important	
  element	
  of	
  the	
  cross-­‐lag	
  
model:	
  the	
  autoregressive	
  components.	
  Although	
  not	
  hypothesized,	
  the	
  autoregressions	
  
were	
  all	
  expected	
  to	
  be	
  significantly	
  and	
  positively	
  related	
  to	
  each	
  other,	
  showing	
  that	
  
previous	
  levels	
  of	
  the	
  variables	
  were	
  associated	
  with	
  subsequent	
  levels	
  of	
  the	
  variables.	
  

	
  

103	
  

	
  
The	
  reason	
  these	
  relationships	
  were	
  not	
  hypothesized	
  is	
  that	
  it	
  is	
  understood	
  that	
  from	
  one	
  
time	
  point	
  to	
  the	
  next,	
  variables	
  are	
  more	
  similar	
  to	
  each	
  other	
  than	
  not	
  (DeShon,	
  2012).	
  
The	
  results	
  of	
  the	
  cross-­‐lag	
  model	
  show	
  that,	
  across	
  both	
  the	
  adaptation	
  and	
  performance	
  
environments,	
  the	
  autoregressive	
  estimates	
  of	
  performance	
  and	
  all	
  the	
  self-­‐regulatory	
  
variables	
  were	
  positive	
  and	
  significant	
  (see	
  Table	
  11).	
  The	
  positive	
  autoregression	
  
estimates	
  are	
  critical	
  to	
  consider,	
  given	
  that	
  some	
  of	
  the	
  findings	
  among	
  the	
  relationships	
  
appear	
  to	
  be	
  confusing	
  based	
  on	
  their	
  trajectory	
  differences.	
  However,	
  when	
  removing	
  the	
  
variance	
  due	
  to	
  the	
  previous	
  level	
  of	
  a	
  variable,	
  it	
  opens	
  up	
  the	
  possibility	
  that	
  other	
  
variables	
  may	
  have	
  an	
  opposite	
  relationship,	
  but	
  that	
  the	
  positive	
  relationship	
  between	
  the	
  
previous	
  and	
  subsequent	
  levels	
  of	
  that	
  variable	
  was	
  so	
  strong	
  that	
  the	
  trajectory	
  continued	
  
in	
  the	
  same	
  direction	
  as	
  earlier.	
  For	
  example,	
  performance	
  increased	
  over	
  time	
  and	
  the	
  
autoregressive	
  component	
  suggests	
  that	
  the	
  level	
  in	
  the	
  prior	
  trial	
  was	
  strongly	
  associated	
  
with	
  an	
  increase	
  in	
  the	
  subsequent	
  trial;	
  contrarily,	
  metacognitive	
  behaviors	
  decreased	
  
over	
  time	
  and	
  the	
  autoregressive	
  component	
  suggested	
  that	
  a	
  decrease	
  in	
  the	
  last	
  trial	
  led	
  
to	
  a	
  decrease	
  in	
  the	
  next.	
  When	
  considering	
  the	
  predictive	
  relationship	
  between	
  
metacognition	
  and	
  performance,	
  it	
  is	
  important	
  to	
  consider	
  that	
  previous	
  performance	
  as	
  
well	
  as	
  previous	
  metacognition	
  is	
  being	
  used	
  to	
  predict	
  subsequent	
  performance.	
  
Therefore,	
  the	
  relationship	
  could	
  be	
  positive	
  or	
  negative,	
  based	
  on	
  fluctuations	
  in	
  the	
  
variables	
  between	
  time	
  points,	
  as	
  opposed	
  to	
  the	
  overall	
  trajectories	
  of	
  the	
  variables.	
  	
  
The	
  following	
  paragraphs	
  will	
  take	
  each	
  hypothesis	
  in	
  the	
  order	
  it	
  was	
  initially	
  
presented,	
  beginning	
  with	
  the	
  cognitive	
  cycle.	
  See	
  Table	
  12	
  for	
  the	
  parameter	
  estimates,	
  
summary	
  of	
  the	
  findings,	
  and	
  support	
  for	
  the	
  hypotheses	
  and	
  Figure	
  12	
  for	
  a	
  visual	
  
representation	
  of	
  the	
  theoretical	
  and	
  empirical	
  models	
  for	
  both	
  processes.	
  

	
  

104	
  

	
  
Table	
  11	
  
Autoregressive	
  Results	
  from	
  the	
  Cross-­‐Lag	
  Model	
  
Cognitive	
  Variables	
  
AR	
  
Std.	
  
Variable	
  
estimate	
  
Error	
  
Learning	
  Effort	
  
.424	
  
.011	
  

p-­‐value	
  

Variable	
  

.000	
  

Goals	
  
Outcome	
  
Effort	
  
Self-­‐efficacy	
  
	
  

Metacognition	
  

.847	
  

.007	
  

.000	
  

Evaluation	
  
Performance	
  

.313	
  
.556	
  

.012	
  
.010	
  

.000	
  
.000	
  

Motivation	
  Variables	
  
AR	
  
Std.	
  
estimate	
  
Error	
  
.697	
  
.004	
  

p-­‐value	
  
.000	
  

.884	
  

.006	
  

.000	
  

.849	
  
	
  

.007	
  
	
  

.000	
  
	
  

	
  
Table	
  12	
  
Summary	
  of	
  Hypotheses	
  and	
  Results	
  from	
  the	
  Cross-­‐Lag	
  Model	
  
Process	
  

CL	
  
Estimate	
  

Std.	
  
Error	
  

p-­‐
value	
  

Adaptation	
  

-­‐.032	
  

.022	
  

.145	
  

Performance	
  

-­‐.003	
  

.010	
  

.724	
  

9a	
   Learning	
  Effortà	
  Metacognition:	
  Strong	
  +	
  
9b	
   Learning	
  Effortà	
  Metacognition:	
  Weak	
  +	
  
10a	
   Metacognitionà	
  Performance:	
  Strong	
  +	
  

Adaptation	
  
Performance	
  
Adaptation	
  

.039	
  
.007	
  
.040	
  

.010	
  
.011	
  
.015	
  

.000	
  
538	
  
.007	
  

10b	
   Metacognitionà	
  Performance:	
  Weak	
  -­‐	
  

Performance	
  

.054	
  

.011	
  

.000	
  

Adaptation	
  

.041	
  

.022	
  

.063	
  

Performance	
  

.051	
  

.011	
  

.000	
  

Adaptation	
  

.063	
  

.020	
  

.002	
  

12b	
   EvaluationàLearning	
  Effort:	
  Weak	
  +	
  

Performance	
  

.083	
  

.013	
  

.000	
  

13a	
   EvaluationàMetacognition:	
  Strong	
  +	
  

Adaptation	
  

.014	
  

.012	
  

.256	
  

13b	
   EvaluationàMetacognition:	
  Weak	
  +	
  

Performance	
  

-­‐.008	
  

.010	
  

.444	
  

Adaptation	
  

.083	
  

.008	
  

.000	
  

Performance	
  
Adaptation	
  
Performance	
  

.067	
  
.016	
  
.020	
  

.006	
  
.015	
  
.006	
  

.000	
  
.284	
  
.001	
  

Adaptation	
  

.236	
  

.017	
  

.000	
  

16b	
   Outcome	
  Effortà	
  Performance:	
  Strong	
  +	
  
17a	
   PerformanceàSelf-­‐efficacy:	
  Strong	
  +	
  

Performance	
  
Adaptation	
  

.275	
  
.087	
  

.012	
  
.011	
  

.000	
  
.000	
  

17b	
   PerformanceàSelf-­‐efficacy:	
  Weak	
  +	
  

Performance	
  

.094	
  

.009	
  

.000	
  

18a	
   Self-­‐efficacyàGoals:	
  Strong	
  +	
  

Adaptation	
  

.049	
  

.008	
  

.000	
  

18b	
   Self-­‐efficacyàGoals:	
  Weak	
  +	
  

Performance	
  

.029	
  

.006	
  

.000	
  

19a	
   Self-­‐efficacyà	
  Outcome	
  Effort:	
  Weak	
  +	
  

Adaptation	
  

.016	
  

.010	
  

.108	
  

19b	
   Self-­‐efficacyà	
  Outcome	
  Effort:	
  Weak	
  -­‐	
  

Performance	
  

.030	
  

.008	
  

.000	
  

	
  

105	
  

#	
  

Hypothesis	
  

8a	
  

Performanceà	
  Learning	
  Effort:	
  Weak	
  -­‐	
  

8b	
  

Performanceà	
  Learning	
  Effort:	
  Strong	
  -­‐	
  

11a	
   Performanceà	
  Evaluation:	
  Weak	
  -­‐	
  
11b	
   Performanceà	
  Evaluation:	
  Strong	
  -­‐	
  
12a	
   Evaluationà	
  Learning	
  Effort:	
  Strong	
  +	
  

14a	
   Performanceà	
  Goals:	
  Weak	
  +	
  
14b	
   PerformanceàGoals:	
  Strong	
  +	
  
15a	
   GoalsàOutcome	
  Effort:	
  Weak	
  +	
  
15b	
   GoalsàOutcome	
  Effort:	
  Strong	
  +	
  
16a	
   Outcome	
  Effortà	
  Performance:	
  Weak	
  +	
  

Summary	
  
of	
  Results	
  
Supported	
  
Not	
  
Supported	
  
Supported	
  
Supported	
  
Supported	
  
Not	
  
Supported	
  
Not	
  
Supported	
  
Not	
  
Supported	
  
Supported	
  
Not	
  
Supported	
  
Not	
  
Supported	
  
Not	
  
Supported	
  
Not	
  
Supported	
  
Supported	
  
Supported	
  
Supported	
  
Not	
  
Supported	
  
Supported	
  
Supported	
  
Not	
  
Supported	
  
Supported	
  
Not	
  
Supported	
  
Supported	
  
Not	
  
Supported	
  

	
  
Figure	
  12	
  
Comparison	
  of	
  the	
  Theoretical	
  Model	
  and	
  Empirical	
  Findings

!Theoretical!Models! !

!

!

!

!

!

!!!!!!Empirical!Findings!

Performance*

Performance*

Adaptation!!
Process!

Evalua:on*
Learning*
Effort*

Learning*
Effort*

Metacogni:on*
Outcome*
Effort*
*
*

Metacogni:on*
*
Outcome*
*
Effort*

Self/*
efficacy*

Self/*
efficacy*

Goals*

Goals*

Performance*

Performance*

Evalua:on*
Learning*
Effort*

	
  

Evalua:on*

Evalua:on*

Performance!
Process!
Learning*
Effort*

Metacogni:on*
Outcome*
Effort*
*
*

Metacogni:on*
*
Outcome*
*
Effort*

Self/*
efficacy*

Self/*
efficacy*

Goals*

Goals*

106	
  

	
  

	
  
Results	
  of	
  the	
  Cognitive	
  Cycle.	
  Hypothesis	
  8a	
  suggested	
  that	
  performance	
  would	
  
have	
  a	
  weak	
  and	
  negative	
  cross-­‐lag	
  relationship	
  during	
  the	
  adaptation	
  process	
  given	
  that,	
  
initially,	
  performance	
  would	
  be	
  low	
  but	
  growing,	
  while	
  learning-­‐oriented	
  effort	
  would	
  be	
  
high	
  and	
  remain	
  so.	
  This	
  finding	
  was	
  consistent	
  with	
  the	
  hypothesis.	
  As	
  individuals	
  shift	
  to	
  
the	
  performance	
  process,	
  Hypothesis	
  8b	
  suggested	
  that	
  these	
  variables	
  would	
  have	
  a	
  
strong	
  negative	
  relationship;	
  however,	
  similar	
  to	
  the	
  adaptation	
  process,	
  the	
  variables	
  had	
  
a	
  weak	
  negative	
  cross-­‐lag	
  relationship,	
  showing	
  only	
  partial	
  support	
  for	
  the	
  hypothesis.	
  
This	
  suggests	
  that	
  when	
  performance	
  increased,	
  less	
  learning-­‐oriented	
  effort	
  was	
  required.	
  
The	
  weakness	
  of	
  this	
  relationship	
  may	
  be	
  partially	
  due	
  to	
  the	
  relative	
  stability	
  of	
  both	
  
performance	
  and	
  learning-­‐oriented	
  effort,	
  causing	
  a	
  lack	
  of	
  variance	
  from	
  which	
  strong	
  
predictions	
  can	
  be	
  made.	
  
Hypothesis	
  9a	
  stated	
  that	
  learning-­‐oriented	
  effort	
  would	
  have	
  a	
  strong	
  and	
  positive	
  
impact	
  on	
  subsequent	
  metacognition	
  behaviors	
  in	
  the	
  adaptation	
  process	
  as	
  investigating	
  
information	
  outside	
  of	
  the	
  task	
  would	
  lead	
  to	
  the	
  testing	
  of	
  new	
  strategies	
  during	
  the	
  task.	
  
The	
  findings	
  support	
  this	
  expectation.	
  Unlike	
  the	
  strong	
  positive	
  relationship	
  between	
  
learning-­‐oriented	
  effort	
  and	
  metacognition	
  in	
  adaptation,	
  Hypothesis	
  9b	
  suggested	
  that	
  this	
  
relationship	
  would	
  become	
  weak	
  after	
  the	
  shift	
  to	
  the	
  performance	
  process,	
  given	
  that	
  both	
  
of	
  these	
  variables	
  would	
  decrease	
  over	
  time	
  and	
  the	
  lack	
  of	
  variance	
  would	
  result	
  in	
  a	
  non-­‐
significant	
  relationship.	
  This	
  was	
  also	
  supported	
  by	
  the	
  findings.	
  
Hypothesis	
  10a	
  presented	
  the	
  expectation	
  that	
  metacognition	
  and	
  performance	
  
would	
  have	
  a	
  strong	
  and	
  positive	
  cross-­‐lagged	
  relationship	
  in	
  the	
  adaptation	
  process,	
  as	
  it	
  
was	
  posited	
  that	
  increased	
  focus	
  on	
  strategy	
  development	
  during	
  the	
  prior	
  trial	
  would	
  
result	
  in	
  higher	
  performance	
  in	
  the	
  next.	
  The	
  results	
  supported	
  this	
  hypothesis.	
  However,	
  

	
  

107	
  

	
  
Hypothesis	
  10b	
  postulates	
  that	
  as	
  individuals	
  shift	
  to	
  the	
  performance	
  process,	
  the	
  
relationship	
  would	
  become	
  weak	
  and	
  negative,	
  as	
  it	
  was	
  anticipated	
  that	
  when	
  more	
  effort	
  
was	
  devoted	
  to	
  strategy	
  development	
  and	
  testing	
  during	
  a	
  more	
  routine	
  environment,	
  then	
  
less	
  effort	
  could	
  be	
  devoted	
  to	
  the	
  execution	
  of	
  the	
  task.	
  The	
  results	
  did	
  not	
  support	
  this	
  
expectation	
  as	
  it	
  was	
  found	
  that	
  these	
  variables	
  had	
  a	
  positive	
  cross-­‐lagged	
  relationship,	
  
suggesting	
  that	
  despite	
  the	
  decrease	
  in	
  metacognitive	
  behaviors	
  over	
  the	
  course	
  of	
  the	
  
performance	
  environment,	
  metacognitive	
  increases	
  or	
  decreases	
  were	
  associated	
  with	
  
similar	
  patterns	
  in	
  performance.	
  This	
  may	
  have	
  been	
  due	
  to	
  the	
  sawtooth	
  pattern	
  that	
  both	
  
of	
  these	
  variables	
  followed,	
  resulting	
  in	
  a	
  positive	
  relationship	
  despite	
  differences	
  in	
  the	
  
trajectories.	
  In	
  other	
  words,	
  if	
  both	
  metacognition	
  and	
  performance	
  followed	
  the	
  same	
  
pattern	
  (i.e.,	
  increasing,	
  then	
  decreasing),	
  then	
  the	
  consistent	
  fluctuations,	
  though	
  small,	
  
can	
  show	
  a	
  strong	
  positive	
  relationship.	
  Although	
  this	
  finding	
  was	
  unexpected,	
  it	
  may	
  speak	
  
to	
  a	
  critical	
  difference	
  between	
  standard	
  trajectory	
  analyses	
  and	
  cross-­‐lag	
  analyses	
  in	
  their	
  
ability	
  to	
  detect	
  such	
  interesting	
  nuances.	
  This	
  will	
  be	
  revisited	
  in	
  the	
  discussion	
  section.	
  
In	
  Hypothesis	
  11a,	
  performance	
  was	
  expected	
  to	
  have	
  a	
  weak	
  and	
  negative	
  impact	
  
on	
  subsequent	
  evaluation	
  behaviors	
  in	
  the	
  adaptation	
  process,	
  suggesting	
  that	
  when	
  
individuals	
  had	
  poor	
  performance	
  in	
  one	
  trial,	
  there	
  would	
  be	
  a	
  greater	
  need	
  to	
  perform	
  
evaluation	
  activities	
  in	
  the	
  subsequent	
  trial.	
  The	
  results	
  show	
  partial	
  support	
  for	
  this	
  
hypothesis	
  in	
  that	
  the	
  relationship	
  was	
  weak,	
  though	
  positive.	
  Similarly,	
  when	
  transitioning	
  
the	
  performance	
  process,	
  Hypothesis	
  11b	
  suggested	
  that	
  the	
  relationship	
  between	
  
previous	
  performance	
  and	
  evaluation	
  behaviors	
  would	
  be	
  negative	
  and	
  stronger,	
  as	
  it	
  was	
  
expected	
  that	
  individuals	
  would	
  understand	
  the	
  change	
  in	
  the	
  environment	
  and	
  thus	
  not	
  
need	
  to	
  investigate	
  feedback	
  information	
  to	
  continue	
  to	
  perform	
  effectively.	
  However,	
  the	
  

	
  

108	
  

	
  
results	
  did	
  not	
  support	
  this	
  hypothesis	
  in	
  that	
  there	
  was	
  a	
  strong	
  positive	
  relationship.	
  The	
  
results	
  of	
  the	
  test	
  of	
  these	
  hypotheses	
  (11a	
  and	
  11b)	
  may	
  have	
  been	
  influenced	
  by	
  the	
  
sawtooth	
  pattern	
  of	
  performance	
  and	
  evaluation,	
  similar	
  to	
  the	
  impact	
  on	
  metacognition	
  
and	
  performance,	
  as	
  discussed	
  in	
  relation	
  to	
  Hypothesis	
  10b.	
  In	
  that	
  case,	
  these	
  
relationships	
  pointed	
  to	
  an	
  interesting	
  outcome	
  of	
  conducting	
  a	
  cross-­‐lag	
  analysis	
  rather	
  
than	
  just	
  showing	
  trajectory	
  changes.	
  These	
  results	
  show	
  that	
  despite	
  the	
  trajectories	
  of	
  
performance	
  increasing	
  and	
  evaluation	
  behaviors	
  decreasing	
  over	
  time	
  (Hypotheses	
  1a,	
  1b,	
  
4b,	
  and	
  5b),	
  investigating	
  the	
  relationships	
  between	
  the	
  variables	
  presents	
  a	
  different	
  look	
  
into	
  the	
  process.	
  The	
  positive	
  relationship	
  may	
  suggest	
  that	
  as	
  performance	
  decreased,	
  
individuals	
  became	
  apathetic	
  to	
  the	
  outcome,	
  and	
  therefore	
  did	
  not	
  investigate	
  feedback	
  to	
  
attempt	
  to	
  determine	
  why	
  that	
  was.	
  On	
  the	
  other	
  hand,	
  it	
  may	
  suggest	
  that	
  when	
  
performance	
  increased,	
  after	
  having	
  decreased	
  in	
  the	
  previous	
  trial,	
  subsequent	
  feedback	
  
evaluation	
  increased,	
  possibly	
  to	
  ascertain	
  what	
  was	
  done	
  correctly	
  in	
  order	
  to	
  attempt	
  to	
  
replicate	
  those	
  behaviors	
  in	
  the	
  next	
  trial.	
  
Hypothesis	
  12a	
  proposed	
  that	
  evaluation	
  behaviors	
  would	
  have	
  a	
  strong	
  positive	
  
relationship	
  on	
  subsequent	
  learning-­‐oriented	
  effort	
  behaviors.	
  This	
  finding	
  was	
  consistent	
  
with	
  the	
  hypothesis,	
  showing	
  that,	
  if	
  individuals	
  increased	
  in	
  evaluation	
  behaviors	
  in	
  the	
  
prior	
  trial,	
  they	
  decreased	
  in	
  learning-­‐oriented	
  effort	
  behaviors	
  in	
  the	
  next	
  trial.	
  Similarly,	
  
Hypothesis	
  12b,	
  suggested	
  that	
  the	
  lagged	
  relationship	
  between	
  evaluation	
  and	
  learning-­‐
oriented	
  effort	
  would	
  also	
  be	
  positive,	
  but	
  would	
  become	
  weaker	
  as	
  individuals	
  
transitioned	
  to	
  the	
  performance	
  process,	
  given	
  that	
  there	
  was	
  expected	
  to	
  be	
  less	
  variance	
  
available	
  to	
  be	
  explained	
  in	
  these	
  variables.	
  This	
  prediction	
  was	
  partially	
  supported	
  in	
  that	
  
there	
  was	
  a	
  positive	
  relationship,	
  but	
  it	
  was	
  strong.	
  This	
  may	
  have	
  been	
  due	
  to	
  the	
  fact	
  that,	
  

	
  

109	
  

	
  
although	
  both	
  of	
  these	
  variables	
  decreased	
  over	
  time,	
  any	
  fluctuations	
  in	
  these	
  variables	
  
may	
  have	
  caused	
  a	
  strong	
  positive	
  prediction.	
  
The	
  last	
  of	
  the	
  relationships	
  in	
  the	
  cognitive	
  cycle	
  was	
  that	
  of	
  evaluation	
  and	
  
metacognition,	
  and	
  Hypothesis	
  13a	
  explained	
  that	
  there	
  should	
  have	
  been	
  a	
  strong	
  positive	
  
relationship	
  in	
  the	
  adaptation	
  process,	
  as	
  increases	
  in	
  feedback	
  seeking	
  (i.e.,	
  evaluation)	
  
were	
  expected	
  to	
  result	
  in	
  more	
  metacognitive	
  activities	
  during	
  the	
  task.	
  The	
  results	
  of	
  the	
  
analysis	
  partially	
  supported	
  this	
  hypothesis,	
  in	
  that	
  there	
  was	
  a	
  positive	
  relationship,	
  albeit	
  
a	
  weak	
  one,	
  when	
  a	
  strong	
  positive	
  relationship	
  was	
  predicted.	
  This	
  may	
  have	
  been	
  due	
  to	
  
evaluation	
  following	
  a	
  sawtooth	
  pattern	
  but	
  metacognition	
  having	
  a	
  relatively	
  stable	
  
trajectory	
  during	
  adaptation.	
  Contrary	
  to	
  Hypothesis	
  13a,	
  13b	
  suggested	
  that	
  the	
  lagged	
  
relationship	
  should	
  become	
  weak,	
  though	
  still	
  positive,	
  as	
  individuals	
  transitioned	
  to	
  the	
  
performance	
  process,	
  since	
  the	
  expected	
  low	
  levels	
  of	
  metacognition	
  and	
  evaluation	
  were	
  
expected	
  to	
  result	
  in	
  less	
  variance	
  and,	
  therefore,	
  in	
  a	
  weaker	
  relationship.	
  This	
  hypothesis	
  
was	
  not	
  supported	
  by	
  the	
  data,	
  in	
  that	
  the	
  relationship	
  was	
  weak,	
  but	
  negative.	
  It	
  is	
  
possible	
  that	
  this	
  negative	
  relationship	
  was	
  due	
  to	
  the	
  sawtooth	
  trajectories	
  of	
  the	
  
variables.	
  The	
  negative	
  relationship	
  suggests	
  that,	
  as	
  previous	
  evaluation	
  behaviors	
  
increased,	
  metacognitive	
  behaviors	
  decreased.	
  Given	
  the	
  interpretation	
  of	
  Hypothesis	
  11b	
  
regarding	
  the	
  positive	
  relationship	
  between	
  previous	
  performance	
  and	
  subsequent	
  
evaluation	
  behaviors,	
  it	
  may	
  have	
  been	
  the	
  case	
  that	
  as	
  individuals	
  investigated	
  their	
  
feedback	
  following	
  increased	
  performance	
  and	
  determined	
  what	
  elements	
  were	
  correct,	
  
subsequently	
  less	
  metacognitive	
  effort	
  was	
  required.	
  Similarly,	
  when	
  fewer	
  feedback	
  
behaviors	
  were	
  employed,	
  individuals	
  may	
  have	
  decided	
  to	
  devote	
  their	
  effort	
  to	
  changing	
  
metacognitive	
  strategies	
  instead.	
  These	
  series	
  of	
  interesting	
  relationships	
  among	
  

	
  

110	
  

	
  
performance,	
  evaluation,	
  metacognition,	
  and	
  subsequent	
  performance	
  will	
  be	
  reviewed	
  in	
  
more	
  detail	
  in	
  the	
  discussion	
  section.	
  
Results	
  of	
  the	
  Motivational	
  Cycle.	
  As	
  Hypothesis	
  14a	
  stated,	
  performance	
  was	
  
expected	
  to	
  have	
  a	
  weak	
  and	
  positive	
  lagged	
  impact	
  on	
  goals	
  in	
  the	
  adaptation	
  process,	
  
given	
  that	
  performance	
  would	
  not	
  likely	
  be	
  calibrated	
  to	
  goals	
  initially,	
  as	
  the	
  change	
  was	
  
in	
  the	
  process	
  of	
  being	
  understood.	
  The	
  results	
  indicate	
  a	
  strong	
  positive	
  relationship,	
  not	
  
supporting	
  Hypothesis	
  14a,	
  though	
  it	
  shows	
  that	
  performance	
  maintained	
  its	
  positive	
  
impact	
  on	
  goals	
  despite	
  the	
  differences	
  in	
  trajectories,	
  likely	
  due	
  to	
  the	
  strong	
  sawtooth	
  
pattern	
  in	
  performance	
  and	
  the	
  weak	
  one	
  in	
  goals.	
  Hypothesis	
  14b	
  suggested	
  that,	
  as	
  
individuals	
  transitioned	
  to	
  the	
  performance	
  process,	
  the	
  relationship	
  between	
  performance	
  
and	
  goals	
  was	
  expected	
  to	
  be	
  even	
  stronger,	
  as	
  goals	
  were	
  anticipated	
  to	
  be	
  better	
  
calibrated	
  as	
  the	
  change	
  was	
  understood.	
  This	
  hypothesis	
  was	
  supported	
  by	
  the	
  data.	
  
The	
  next	
  link	
  in	
  the	
  motivational	
  cycle	
  was	
  the	
  relationship	
  between	
  goals	
  and	
  
outcome-­‐oriented	
  effort.	
  Hypothesis	
  15a	
  predicted	
  that	
  there	
  would	
  be	
  a	
  weak	
  positive	
  
cross-­‐lag	
  relationship,	
  given	
  that	
  goals	
  would	
  not	
  likely	
  be	
  well-­‐calibrated	
  during	
  the	
  
adaptation	
  process.	
  Results	
  support	
  this	
  prediction.	
  Similar	
  to	
  the	
  change	
  in	
  the	
  
relationship	
  between	
  performance	
  and	
  goals,	
  Hypothesis	
  15b	
  suggested	
  that,	
  in	
  the	
  
performance	
  process,	
  the	
  relationship	
  between	
  previous	
  goals	
  and	
  subsequent	
  outcome-­‐
oriented	
  effort	
  would	
  transition	
  from	
  weak	
  to	
  strong,	
  given	
  that	
  goals	
  were	
  expected	
  to	
  be	
  
better	
  calibrated	
  in	
  this	
  process.	
  The	
  finding	
  supported	
  the	
  hypothesis.	
  
Hypothesis	
  16a	
  proposed	
  that,	
  in	
  the	
  adaptation	
  process,	
  outcome-­‐oriented	
  effort	
  
behaviors	
  would	
  have	
  a	
  weak	
  and	
  positive	
  impact	
  on	
  subsequent	
  performance,	
  indicating,	
  
as	
  more	
  effort	
  is	
  devoted	
  to	
  the	
  task,	
  higher	
  performance	
  would	
  result.	
  However,	
  this	
  

	
  

111	
  

	
  
would	
  be	
  weakened	
  by	
  the	
  fact	
  that,	
  in	
  the	
  adaptation	
  phase,	
  effort	
  would	
  not	
  be	
  well-­‐	
  
calibrated	
  to	
  understand	
  what	
  aspects	
  of	
  the	
  task	
  had	
  changed	
  and	
  how	
  that	
  impacted	
  
performance.	
  Results	
  show	
  partial	
  support	
  for	
  the	
  hypothesis	
  in	
  that	
  the	
  relationship	
  was	
  
positive	
  but	
  strong,	
  instead	
  of	
  weak.	
  This	
  suggests	
  that	
  individuals	
  were	
  able	
  to	
  more	
  
quickly	
  calibrate	
  their	
  effort	
  to	
  their	
  performance,	
  even	
  directly	
  after	
  the	
  change.	
  This	
  may	
  
have	
  been	
  due	
  to	
  the	
  fact	
  that,	
  in	
  the	
  component	
  condition	
  in	
  particular,	
  increasing	
  effort	
  
was	
  related	
  to	
  better	
  performance,	
  thus,	
  even	
  slight	
  increases	
  in	
  effort	
  would	
  result	
  in	
  
better	
  performance.	
  When	
  transitioning	
  to	
  the	
  performance	
  process,	
  Hypothesis	
  16b	
  
suggested	
  that,	
  similar	
  to	
  the	
  previous	
  motivational	
  relationships	
  discussed,	
  the	
  lagged	
  
relationship	
  between	
  outcome-­‐oriented	
  effort	
  and	
  performance	
  would	
  transition	
  to	
  have	
  a	
  
strong	
  and	
  positive	
  relationship	
  in	
  the	
  performance	
  phase,	
  as	
  effort	
  would	
  be	
  better-­‐
calibrated	
  in	
  this	
  phase.	
  The	
  findings	
  were	
  consistent	
  with	
  this	
  hypothesis.	
  
Hypothesis	
  17a	
  proposed	
  that	
  performance	
  would	
  have	
  a	
  strong	
  and	
  positive	
  
impact	
  on	
  subsequent	
  self-­‐efficacy	
  in	
  the	
  adaptation	
  phase,	
  as	
  increases	
  in	
  performance,	
  
despite	
  the	
  initial	
  decline,	
  would	
  be	
  associated	
  with	
  later	
  increases	
  in	
  confidence	
  in	
  their	
  
ability	
  to	
  handle	
  the	
  new	
  environment.	
  The	
  results	
  show	
  support	
  for	
  this	
  hypothesis.	
  
Hypothesis	
  17b	
  also	
  proposed	
  that	
  the	
  lagged	
  relationship	
  would	
  be	
  positive	
  as	
  individuals	
  
transitioned	
  to	
  the	
  performance	
  process,	
  but	
  would	
  become	
  weak,	
  given	
  the	
  expectation	
  
that	
  self-­‐efficacy	
  would	
  stabilize,	
  resulting	
  in	
  less	
  variability	
  for	
  prediction.	
  Results	
  from	
  
the	
  analysis	
  showed	
  partial	
  support	
  for	
  the	
  hypothesis	
  in	
  that	
  the	
  relationship	
  was	
  positive,	
  
but	
  strong.	
  This	
  suggests	
  that,	
  even	
  in	
  the	
  performance	
  phase,	
  performing	
  well	
  or	
  poorly	
  
continued	
  to	
  impact	
  self-­‐efficacy,	
  showing	
  that	
  individuals	
  may	
  have	
  thought	
  the	
  task	
  was	
  
continuing	
  to	
  change.	
  This	
  conclusion	
  is	
  partially	
  based	
  on	
  the	
  misinterpretation	
  apparent	
  

	
  

112	
  

	
  
in	
  the	
  responses	
  to	
  the	
  self-­‐report	
  variable	
  that	
  asked	
  participants	
  to	
  explicate	
  what	
  
changed	
  in	
  the	
  task	
  itself,	
  during	
  the	
  current	
  trial	
  (when	
  the	
  change	
  was	
  introduced),	
  about	
  
which	
  they	
  instead	
  reported	
  what	
  they	
  thought	
  changed	
  from	
  the	
  last	
  trial.	
  
Hypothesis	
  18a	
  anticipated	
  that	
  in	
  the	
  adaptation	
  process,	
  self-­‐efficacy	
  would	
  have	
  a	
  
strong	
  and	
  positive	
  impact	
  on	
  subsequent	
  goals,	
  given	
  that	
  belief	
  in	
  ability	
  has	
  been	
  shown	
  
to	
  have	
  a	
  strong	
  impact	
  on	
  goals	
  created.	
  This	
  was	
  supported	
  by	
  the	
  results.	
  However,	
  
when	
  shifting	
  to	
  the	
  performance	
  process,	
  it	
  was	
  expected	
  that	
  this	
  positive	
  relationship	
  
would	
  have	
  a	
  weaker	
  effect,	
  given	
  that	
  both	
  variables	
  were	
  expected	
  to	
  have	
  less	
  
variability.	
  Thus,	
  the	
  result	
  of	
  the	
  analysis	
  did	
  not	
  show	
  support	
  for	
  Hypothesis	
  18b	
  in	
  that	
  
the	
  relationship	
  was	
  positive;	
  instead,	
  the	
  relationship	
  was	
  stronger	
  than	
  anticipated.	
  This	
  
suggests	
  that	
  even	
  small	
  fluctuations	
  in	
  self-­‐efficacy	
  predicted	
  fluctuations	
  in	
  goal	
  levels.	
  
For	
  the	
  final	
  relationship	
  of	
  the	
  motivational	
  cycle,	
  Hypothesis	
  19a	
  stated	
  that	
  self-­‐
efficacy	
  was	
  expected	
  to	
  have	
  a	
  weak	
  and	
  positive	
  impact	
  on	
  subsequent	
  outcome-­‐oriented	
  
effort	
  behaviors,	
  given	
  that	
  expectations	
  of	
  ability	
  in	
  a	
  new	
  environment	
  were	
  anticipated	
  
to	
  help	
  individuals	
  persist	
  in	
  a	
  difficult	
  situation,	
  despite	
  the	
  fact	
  that	
  individuals	
  may	
  not	
  
have	
  their	
  effort	
  well-­‐calibrated	
  initially.	
  Results	
  show	
  support	
  for	
  this	
  hypothesis.	
  Unlike	
  
in	
  the	
  adaptation	
  process,	
  Hypothesis	
  19b	
  proposed	
  that,	
  in	
  the	
  performance	
  process,	
  self-­‐
efficacy	
  was	
  expected	
  to	
  have	
  a	
  weak	
  and	
  negative	
  impact	
  on	
  subsequent	
  outcome-­‐oriented	
  
effort	
  behaviors	
  given	
  that	
  research	
  suggests	
  that	
  overconfidence	
  can	
  lead	
  to	
  less	
  effort	
  
being	
  devoted	
  to	
  the	
  task.	
  Results	
  did	
  not	
  support	
  this	
  hypothesis,	
  instead	
  showing	
  that	
  
there	
  continued	
  to	
  be	
  a	
  positive	
  relationship,	
  and	
  a	
  strong	
  one	
  at	
  that.	
  Given	
  that	
  self-­‐
efficacy	
  stabilized	
  at	
  a	
  relatively	
  low	
  level,	
  this	
  may	
  have	
  accounted	
  for	
  the	
  maintenance	
  of	
  
effort	
  despite	
  transitioning	
  to	
  the	
  performance	
  process.	
  

	
  

113	
  

	
  
DISCUSSION	
  
	
  
The	
  purpose	
  of	
  the	
  study	
  was	
  to	
  present	
  a	
  theory	
  of	
  the	
  process	
  of	
  adaptation,	
  
specifying	
  the	
  dynamics	
  involved,	
  theoretically	
  distinguishing	
  adaptation	
  from	
  routine	
  
performance,	
  and	
  empirically	
  presenting	
  evidence	
  for	
  such	
  a	
  distinction.	
  This	
  research	
  was	
  
necessary	
  given	
  that	
  the	
  literature	
  has	
  not	
  yet	
  empirically	
  investigated	
  adaptation	
  as	
  a	
  
process	
  (Baard	
  et	
  al.,	
  2014),	
  and	
  the	
  self-­‐regulatory	
  processes,	
  typically	
  associated	
  with	
  the	
  
conceptualizations	
  of	
  the	
  process,	
  have	
  not	
  been	
  examined	
  longitudinally	
  (e.g.,	
  Kozlowski	
  
et	
  al.,	
  1996;	
  Burke	
  et	
  al.,	
  2006).	
  However,	
  the	
  workplace	
  continues	
  to	
  require	
  individuals,	
  
teams,	
  and	
  organizations	
  to	
  adapt	
  to	
  new	
  situations.	
  Given	
  this	
  need	
  and	
  the	
  fact	
  that	
  many	
  
of	
  the	
  proposed	
  hypotheses	
  were	
  supported,	
  the	
  present	
  research	
  has	
  the	
  ability	
  to	
  inform	
  
individuals	
  and	
  organizations	
  in	
  how	
  they	
  can	
  better	
  prepare	
  for	
  adaptive	
  situations.	
  	
  
In	
  this	
  last	
  section,	
  I	
  will	
  structure	
  my	
  thoughts	
  in	
  the	
  following	
  way.	
  First,	
  I	
  will	
  
discuss	
  the	
  interesting	
  elements	
  discovered	
  concerning	
  the	
  trajectory	
  changes	
  in	
  
performance	
  and	
  the	
  self-­‐regulatory	
  variables,	
  presenting	
  both	
  what	
  was	
  expected	
  and	
  
what	
  was	
  not.	
  Then,	
  I	
  will	
  move	
  to	
  describe	
  the	
  relationship	
  changes	
  in	
  both	
  the	
  adaptation	
  
and	
  performance	
  processes,	
  focusing	
  mostly	
  on	
  what	
  was	
  not	
  expected,	
  since	
  the	
  results	
  
show	
  support	
  for	
  many	
  of	
  the	
  hypotheses.	
  Therefore,	
  an	
  exhaustive	
  discussion	
  on	
  them	
  
would	
  be	
  redundant	
  to	
  the	
  theoretical	
  framing	
  previously	
  presented.	
  Next,	
  I	
  will	
  
summarize	
  the	
  findings	
  of	
  the	
  overall	
  process,	
  as	
  well	
  as	
  some	
  higher-­‐level	
  observations	
  
about	
  the	
  results.	
  Finally,	
  I	
  will	
  move	
  onto	
  the	
  practical	
  implications,	
  limitations,	
  and	
  
opportunities	
  for	
  future	
  research	
  prior	
  to	
  my	
  concluding	
  remarks.	
  
	
  

	
  

114	
  

	
  
Discussion	
  of	
  the	
  results	
  
1st	
  Order	
  Changes:	
  Transition	
  and	
  Trajectories	
  
	
  

The	
  first	
  step	
  in	
  understanding	
  when	
  the	
  adaptation	
  process	
  began	
  and	
  ended	
  was	
  

through	
  identifying	
  the	
  transition	
  point	
  that	
  separated	
  the	
  adaptation	
  and	
  performance	
  
processes	
  in	
  the	
  conditions.	
  The	
  results	
  demonstrated	
  that	
  the	
  coordinative	
  complexity	
  
condition	
  had	
  a	
  later	
  transition	
  point,	
  suggesting	
  that	
  that	
  type	
  of	
  change	
  is	
  more	
  difficult	
  
to	
  adapt	
  to	
  than	
  a	
  component	
  complexity	
  change,	
  supporting	
  Wood’s	
  (1986)	
  
conceptualization	
  and	
  the	
  continuum	
  of	
  change	
  types	
  proposed	
  in	
  Figure	
  1	
  earlier.	
  This	
  
finding	
  extends	
  the	
  research	
  by	
  Lang	
  and	
  Bliese	
  (2009)	
  in	
  applying	
  the	
  analysis	
  of	
  
discontinuous	
  growth	
  curves	
  to	
  not	
  only	
  separate	
  performance	
  occurring	
  after	
  a	
  change	
  
(rather	
  than	
  just	
  the	
  transition	
  point	
  to	
  identify	
  when	
  adaptation	
  began)	
  but	
  also	
  in	
  
applying	
  it	
  to	
  a	
  non-­‐performance	
  variable	
  (i.e.,	
  focus	
  of	
  effort)	
  to	
  provide	
  additional	
  
support	
  for	
  the	
  transition	
  point.	
  This	
  suggests,	
  even	
  though	
  the	
  performance	
  trajectories	
  
were	
  slightly	
  different	
  between	
  the	
  conditions	
  (i.e.,	
  the	
  trajectory	
  of	
  performance	
  in	
  the	
  
coordinative	
  condition	
  was	
  non-­‐significant	
  during	
  the	
  adaptation	
  process	
  and	
  increased	
  in	
  
the	
  performance	
  process,	
  while	
  the	
  component	
  complexity	
  condition	
  had	
  the	
  opposite	
  
finding),	
  this	
  analysis	
  can	
  be	
  used	
  to	
  examine	
  variables	
  other	
  than	
  performance	
  to	
  
understand	
  changes	
  over	
  time.	
  Furthermore,	
  the	
  results	
  on	
  the	
  trajectory	
  differences	
  
present	
  an	
  interesting	
  situation.	
  It	
  would	
  appear	
  the	
  coordinative	
  complexity	
  condition	
  did	
  
not	
  adapt	
  if	
  the	
  adaptation	
  environment	
  was	
  only	
  defined	
  by	
  performance	
  increasing	
  and	
  if	
  
the	
  performance	
  environment	
  was	
  only	
  defined	
  by	
  performance	
  stabilizing	
  during	
  this	
  
period.	
  However,	
  using	
  the	
  self-­‐regulatory	
  variable	
  that	
  measured	
  the	
  focus	
  on	
  learning	
  
versus	
  execution	
  allowed	
  for	
  the	
  process	
  to	
  be	
  the	
  focus,	
  not	
  just	
  performance.	
  These	
  

	
  

115	
  

	
  
results	
  may	
  suggest	
  that	
  although	
  performance	
  trajectories	
  may	
  look	
  slightly	
  different	
  
based	
  on	
  the	
  level	
  of	
  difficulty	
  of	
  the	
  change,	
  the	
  process	
  duration	
  may	
  be	
  more	
  similar.	
  In	
  
other	
  words,	
  more	
  difficult	
  changes	
  will	
  result	
  in	
  a	
  lag	
  in	
  performance	
  such	
  that,	
  even	
  
though	
  the	
  adaptation	
  process	
  is	
  completed	
  (i.e.,	
  individuals	
  have	
  all	
  the	
  information	
  they	
  
need	
  to	
  understand	
  what	
  they	
  need	
  to	
  do),	
  performance	
  will	
  take	
  longer	
  to	
  be	
  corrected	
  
than	
  if	
  the	
  change	
  was	
  easier	
  to	
  identify	
  and	
  understand.	
  It	
  would	
  be	
  interesting	
  for	
  a	
  
conversation	
  to	
  begin	
  in	
  the	
  literature	
  regarding	
  whether	
  performance	
  should	
  be	
  the	
  
primary	
  way	
  in	
  which	
  adaptation	
  is	
  diagnosed	
  (as	
  is	
  the	
  favored	
  approach	
  in	
  the	
  current	
  
adaptation	
  literature;	
  e.g.,	
  Bell	
  &	
  Kozlowski,	
  2008;	
  Dorman	
  &	
  Frese,	
  1994;	
  Keith	
  &	
  Frese,	
  
2005;	
  Lang	
  &	
  Bliese,	
  2009;	
  Mathieu	
  et	
  al.,	
  2000),	
  or	
  whether	
  this	
  should	
  be	
  accomplished	
  
with	
  some	
  other	
  factor	
  (e.g.,	
  a	
  focus	
  on	
  learning	
  versus	
  a	
  focus	
  on	
  execution)	
  that	
  should	
  
drive	
  our	
  understanding	
  of	
  whether	
  the	
  adaptation	
  process	
  is	
  complete	
  or	
  not.	
  
	
  

The	
  majority	
  of	
  the	
  trajectory	
  changes	
  of	
  the	
  cognitive	
  variables	
  in	
  both	
  the	
  

adaptation	
  and	
  performance	
  environments	
  were	
  as	
  expected.	
  This	
  provides	
  additional	
  
evidence	
  to	
  the	
  literature	
  that	
  has	
  long	
  suggested	
  that	
  there	
  is	
  a	
  high	
  need	
  to	
  cognitively	
  
evaluate	
  changes	
  that	
  occur	
  in	
  the	
  environment	
  and	
  the	
  ramifications	
  of	
  various	
  possible	
  
actions.	
  	
  
The	
  motivational	
  variables,	
  however,	
  did	
  not	
  follow	
  the	
  expected	
  trajectories.	
  
Instead	
  of	
  self-­‐efficacy	
  and	
  goals	
  dropping	
  suddenly	
  after	
  the	
  adaptive	
  change	
  was	
  
introduced,	
  individuals	
  appeared	
  to	
  have	
  overinflated	
  levels	
  of	
  these	
  regulatory	
  variables.	
  
This	
  may	
  have	
  been	
  caused	
  by	
  the	
  artificial	
  setting,	
  with	
  little	
  at	
  stake,	
  but	
  may	
  also	
  be	
  
attributable	
  to	
  individuals	
  not	
  being	
  aware	
  of	
  their	
  own	
  ability	
  to	
  adapt,	
  showing	
  a	
  lack	
  of	
  
self-­‐awareness.	
  This	
  speculation	
  is	
  supported	
  by	
  the	
  positively	
  skewed	
  responses	
  to	
  the	
  

	
  

116	
  

	
  
IADAPT	
  questionnaire	
  in	
  which	
  individuals	
  had	
  higher	
  responses	
  than	
  a	
  normal	
  
distribution	
  would	
  dictate.	
  The	
  Shapiro-­‐Wilk	
  test	
  for	
  normality	
  indicated	
  that	
  the	
  IADAPT	
  
measure	
  was	
  non-­‐normal	
  (W=.992,	
  p=.006).	
  Further	
  investigation	
  indicated	
  that	
  this	
  non-­‐
normality	
  was	
  positively	
  valenced,	
  with	
  the	
  mean	
  of	
  the	
  adaptability	
  measure	
  being	
  3.642	
  
and	
  the	
  standard	
  deviation	
  equaling	
  .313	
  (while	
  the	
  scale	
  only	
  ranged	
  from	
  one	
  to	
  five).	
  
Furthermore,	
  there	
  was	
  range	
  restriction	
  of	
  this	
  measure,	
  with	
  the	
  minimum	
  being	
  2.728;	
  
and	
  the	
  maximum,	
  4.8.	
  This	
  suggests	
  that	
  not	
  only	
  in	
  this	
  task,	
  but	
  in	
  general,	
  individuals	
  
are	
  overconfident	
  in	
  their	
  ability	
  to	
  deal	
  with	
  adaptive	
  situations.	
  There	
  are	
  pros	
  and	
  cons	
  
to	
  this	
  finding.	
  On	
  the	
  positive	
  side,	
  maintaining	
  high	
  levels	
  of	
  self-­‐efficacy	
  and	
  goals	
  may	
  
show	
  that	
  individuals	
  felt	
  well-­‐prepared	
  to	
  handle	
  the	
  challenge	
  of	
  the	
  change.	
  However,	
  
the	
  quick	
  decrease	
  in	
  self-­‐efficacy	
  and	
  goals	
  over	
  the	
  next	
  few	
  trials	
  suggests	
  that	
  
individuals	
  lost	
  that	
  confidence	
  as	
  the	
  task	
  continued	
  to	
  be	
  challenging	
  and	
  performance	
  
was	
  not	
  increasing	
  rapidly.	
  On	
  the	
  other	
  hand,	
  overconfidence	
  has	
  the	
  ability	
  to	
  distort	
  
one’s	
  perspective	
  on	
  the	
  task,	
  resulting	
  in	
  less	
  motivation,	
  if	
  performance	
  does	
  not	
  
immediately	
  increase,	
  or	
  in	
  a	
  lack	
  of	
  adaptation,	
  if	
  the	
  performance	
  decrease	
  is	
  not	
  taken	
  
seriously.	
  Thus,	
  it	
  appears	
  that	
  the	
  general	
  overconfidence	
  individuals	
  have	
  in	
  their	
  
adaptability	
  will	
  likely	
  have	
  a	
  negative	
  impact	
  on	
  their	
  ability	
  to	
  adapt.	
  Perhaps	
  more	
  
training	
  on	
  effective	
  adaptive	
  techniques	
  (e.g.,	
  where	
  to	
  get	
  information	
  if	
  a	
  change	
  should	
  
occur,	
  or	
  how	
  to	
  develop	
  different	
  strategies)	
  may	
  help	
  individuals	
  better	
  align	
  their	
  self-­‐
efficacy	
  and	
  goals	
  in	
  adaptive	
  situations.	
  
Finally,	
  the	
  similarity	
  in	
  these	
  trajectories	
  across	
  the	
  two	
  conditions	
  (component	
  
and	
  coordinative	
  complexity	
  changes)	
  lends	
  support	
  to	
  the	
  conclusion	
  that,	
  although	
  
changes	
  may	
  be	
  different	
  in	
  difficulty	
  and	
  type,	
  individuals	
  engage	
  in	
  the	
  same	
  adaptation	
  

	
  

117	
  

	
  
process.	
  This	
  has	
  large	
  ramifications	
  for	
  both	
  research	
  and	
  practice,	
  as	
  theories	
  can	
  be	
  
developed	
  with	
  multiple	
  types	
  of	
  changes	
  in	
  mind,	
  and	
  organizations	
  can	
  implement	
  
trainings	
  or	
  learning	
  initiatives	
  that	
  are	
  not	
  job	
  specific	
  but	
  rather	
  focused	
  on	
  the	
  general	
  
principles	
  presented	
  in	
  this	
  study.	
  
	
  
2nd	
  Order	
  Changes:	
  Relationships	
  	
  
	
  

Although	
  the	
  trajectory	
  changes	
  of	
  the	
  cognitive	
  variables	
  were	
  as	
  expected,	
  not	
  all	
  

of	
  the	
  relationship	
  changes	
  were.	
  On	
  the	
  other	
  hand,	
  	
  the	
  motivational	
  variables	
  had	
  very	
  
unexpected	
  trajectories,	
  yet	
  their	
  relationships	
  were	
  mostly	
  as	
  expected.	
  This	
  presents	
  an	
  
interesting	
  difference	
  in	
  the	
  ways	
  in	
  which	
  dynamics	
  are	
  investigated.	
  Unlike	
  single	
  
variable,	
  single	
  process	
  growth	
  curve	
  analyses,	
  the	
  partially	
  crossed,	
  partially	
  time	
  varying	
  
cross-­‐lag	
  panel	
  regression	
  analysis	
  incorporates	
  all	
  variables	
  involved	
  in	
  the	
  processes	
  into	
  
one	
  analysis,	
  providing	
  a	
  very	
  holistic	
  approach	
  to	
  the	
  examination	
  of	
  the	
  changes	
  in	
  the	
  
processes.	
  Furthermore,	
  the	
  cross-­‐lag	
  analysis	
  estimates	
  an	
  autoregressive	
  term	
  which	
  
accounts	
  for	
  the	
  variance	
  associated	
  with	
  previous	
  levels	
  of	
  that	
  same	
  variable,	
  which	
  
appears	
  to	
  have	
  had	
  a	
  large	
  impact	
  on	
  the	
  relationships	
  estimated.	
  This	
  will	
  be	
  discussed	
  in	
  
more	
  detail	
  as	
  I	
  focus	
  on	
  the	
  cognitive	
  and	
  motivational	
  sub-­‐cycles.	
  
Cognitive	
  Cycle	
  Observations.	
  Of	
  particular	
  interest	
  in	
  this	
  sub-­‐cycle	
  were	
  the	
  
lagged	
  relationships	
  between	
  initial	
  performance,	
  evaluation,	
  metacognition,	
  and	
  
subsequent	
  performance.	
  For	
  one,	
  all	
  the	
  paths	
  in	
  the	
  adaptation	
  process	
  were	
  positive	
  
although	
  performance	
  was	
  increasing	
  while	
  metacognition	
  was	
  somewhat	
  stable	
  and	
  
evaluation	
  was	
  decreasing.	
  The	
  positive	
  cross-­‐lag	
  relationship	
  between	
  performance	
  and	
  
evaluation	
  may	
  have	
  been	
  due	
  to	
  individuals	
  desiring	
  the	
  continuance	
  of	
  increasing	
  

	
  

118	
  

	
  
performance	
  by	
  seeking	
  more	
  feedback,	
  even	
  when	
  performance	
  was	
  increasing	
  (rather	
  
than	
  decreasing).	
  This	
  appeared	
  to	
  lead	
  to	
  somewhat	
  more	
  metacognitive	
  behaviors	
  
(checking	
  on	
  strategy	
  effectiveness),	
  which	
  lead	
  to	
  even	
  better	
  performance	
  in	
  the	
  next	
  
trial.	
  Individuals	
  may	
  also	
  have	
  not	
  wanted	
  to	
  focus	
  on	
  their	
  failures	
  and	
  instead	
  were	
  more	
  
motivated	
  to	
  look	
  at	
  feedback	
  when	
  their	
  performance	
  was	
  higher.	
  Finally,	
  the	
  positive	
  
relationship	
  between	
  metacognition	
  and	
  performance	
  may	
  have	
  been	
  due	
  to	
  the	
  similar	
  
sawtooth	
  pattern	
  that	
  developed	
  for	
  all	
  of	
  these	
  variables.	
  This	
  suggests	
  that	
  even	
  though	
  
the	
  trajectories	
  were	
  in	
  the	
  opposite	
  direction	
  (with	
  metacognition	
  being	
  slightly	
  negative	
  
and	
  performance	
  being	
  positive),	
  accounting	
  for	
  the	
  previous	
  levels	
  of	
  these	
  variables	
  (i.e.,	
  
the	
  autoregressive	
  component)	
  allowed	
  for	
  the	
  remaining	
  variance	
  to	
  be	
  distributed	
  
differently	
  such	
  that	
  instead	
  of	
  an	
  expected	
  negative	
  relationship,	
  the	
  result	
  show	
  a	
  positive	
  
relationship	
  due	
  to	
  the	
  way	
  in	
  which	
  they	
  fluctuated	
  and	
  not	
  their	
  overall	
  trajectories.	
  This	
  
presents	
  an	
  interesting	
  conclusion	
  –	
  trajectories	
  and	
  relationships	
  tell	
  different	
  parts	
  of	
  the	
  
story	
  when	
  it	
  comes	
  to	
  unpacking	
  a	
  dynamic	
  process.	
  
In	
  the	
  performance	
  process,	
  the	
  three	
  pathways	
  discussed	
  above	
  were	
  in	
  the	
  
opposite	
  direction	
  as	
  predicted.	
  Similar	
  to	
  the	
  above,	
  the	
  relationship	
  between	
  
performance	
  and	
  evaluation	
  being	
  positive	
  may	
  have	
  to	
  do	
  with	
  a	
  desire	
  to	
  continue	
  to	
  
increase	
  performance	
  (and	
  is	
  likely	
  driven	
  by	
  the	
  coordinative	
  complexity	
  condition	
  where	
  
performance	
  did	
  not	
  stabilize	
  in	
  the	
  performance	
  environment).	
  However,	
  that	
  positive	
  
relationship	
  may	
  be	
  associated	
  with	
  the	
  fatigue	
  individuals	
  faced	
  as	
  the	
  experiment	
  
continued,	
  resulting	
  in	
  evaluation	
  behaviors	
  being	
  less	
  frequent.	
  Considering	
  there	
  was	
  a	
  
strong	
  relationship	
  between	
  performance	
  and	
  evaluation	
  in	
  the	
  performance	
  process,	
  it	
  
makes	
  sense	
  that	
  if	
  more	
  evaluation	
  were	
  conducted	
  in	
  the	
  previous	
  trial,	
  this	
  would	
  lead	
  

	
  

119	
  

	
  
to	
  less	
  subsequent	
  metacognitive	
  behavior.	
  For	
  instance,	
  if	
  more	
  participants	
  spent	
  time	
  
evaluating	
  performance	
  in	
  a	
  routine	
  environment,	
  and	
  receiving	
  feedback	
  that	
  effective	
  
performance	
  levels	
  were	
  achieved,	
  it	
  is	
  less	
  likely	
  that	
  more	
  effort	
  would	
  need	
  to	
  be	
  
devoted	
  to	
  thinking	
  about	
  different	
  strategies	
  to	
  enhance	
  performance.	
  However,	
  it	
  is	
  still	
  
logical	
  to	
  assert	
  that	
  thinking	
  about	
  more	
  effective	
  strategies,	
  even	
  if	
  the	
  thinking	
  is	
  not	
  
entirely	
  necessary,	
  would	
  lead	
  to	
  enhanced	
  performance	
  given	
  that	
  more	
  efficient	
  
strategies	
  to	
  deal	
  with	
  the	
  task	
  may	
  be	
  discovered.	
  This	
  logic	
  supports	
  the	
  positive	
  
relationship	
  between	
  metacognition	
  and	
  performance,	
  although	
  that	
  relationship	
  was	
  not	
  
anticipated.	
  It	
  is	
  likely	
  that,	
  with	
  an	
  extended	
  investigation	
  of	
  the	
  performance	
  process,	
  this	
  
may	
  change	
  when	
  it	
  is	
  clear	
  to	
  the	
  individual	
  that	
  the	
  task	
  is	
  no	
  longer	
  changing	
  and	
  that	
  
spending	
  time	
  strategizing	
  may	
  not	
  be	
  beneficial.	
  This	
  positive	
  relationship	
  may	
  have	
  been	
  
enhanced	
  given	
  that	
  individuals	
  may	
  not	
  have	
  recognized	
  that	
  the	
  task	
  ceased	
  to	
  change,	
  
particularly	
  in	
  the	
  coordinative	
  condition.	
  
Motivational	
  Cycle	
  Observations.	
  Of	
  particular	
  interest	
  in	
  the	
  motivational	
  sub-­‐
cycle	
  was	
  the	
  consistently	
  positive	
  impact	
  of	
  motivation.	
  As	
  expected	
  in	
  the	
  adaptation	
  
process,	
  the	
  relationship	
  between	
  performance	
  and	
  self-­‐efficacy	
  was	
  high,	
  even	
  with	
  the	
  
sawtooth	
  pattern	
  of	
  performance,	
  the	
  relationship	
  between	
  self-­‐efficacy	
  and	
  goals	
  was	
  
strong,	
  and	
  the	
  relationships	
  between	
  self-­‐efficacy	
  and	
  effort,	
  and	
  goals	
  and	
  effort,	
  were	
  
weak.	
  This	
  is	
  understandable	
  given	
  that,	
  even	
  though	
  one	
  may	
  feel	
  more	
  confident	
  or	
  set	
  
higher	
  goals,	
  effort	
  may	
  not	
  follow	
  that	
  increase	
  either	
  because	
  they	
  are	
  already	
  
performing	
  at	
  a	
  maximal	
  level,	
  or	
  because	
  effort	
  may	
  not	
  have	
  been	
  well-­‐calibrated	
  in	
  an	
  
adaptive	
  environment	
  where	
  the	
  change	
  is	
  still	
  not	
  understood.	
  It	
  was,	
  however,	
  surprising	
  
that	
  performance	
  would	
  have	
  a	
  strong	
  positive	
  impact	
  on	
  goals	
  in	
  the	
  adaptation	
  process,	
  

	
  

120	
  

	
  
rather	
  than	
  the	
  expected	
  weak	
  positive	
  relationship,	
  but	
  this	
  may	
  have	
  been	
  an	
  artifact	
  of	
  
goals	
  being	
  measured	
  on	
  the	
  same	
  scale	
  and	
  very	
  soon	
  after	
  performance	
  feedback	
  was	
  
available.	
  This	
  may	
  have	
  resulted	
  in	
  the	
  two	
  scores	
  being	
  more	
  related	
  than	
  if	
  these	
  two	
  
factors	
  were	
  different.	
  It	
  appears	
  that	
  when	
  performance	
  was	
  poor,	
  individuals	
  set	
  a	
  lower	
  
goal;	
  however,	
  when	
  performance	
  was	
  high,	
  they	
  set	
  a	
  higher	
  goal,	
  despite	
  their	
  not	
  fully	
  
understanding	
  why.	
  Also	
  somewhat	
  unexpected	
  was	
  the	
  strong	
  positive	
  relationship	
  
between	
  outcome	
  effort	
  and	
  performance.	
  The	
  trajectories	
  of	
  the	
  two	
  conditions	
  show	
  that	
  
this	
  relationship	
  may	
  have	
  been	
  primarily	
  influenced	
  by	
  the	
  component	
  condition,	
  as	
  they	
  
maintained	
  effort	
  levels	
  over	
  time	
  given	
  that	
  the	
  nature	
  of	
  the	
  adaptive	
  change	
  required	
  a	
  
high	
  level	
  of	
  effort	
  behaviors.	
  
	
  

The	
  performance	
  process	
  also	
  correctly	
  anticipated	
  that	
  there	
  would	
  be	
  strong	
  

positive	
  relationships	
  between	
  performance,	
  goals,	
  outcome	
  effort,	
  and	
  performance.	
  
However,	
  it	
  was	
  unexpected	
  that	
  there	
  would	
  be	
  a	
  strong	
  relationship	
  between	
  
performance	
  and	
  self-­‐efficacy.	
  This	
  may	
  have	
  been	
  due	
  to	
  self-­‐efficacy	
  stabilizing	
  at	
  a	
  level	
  
lower	
  than	
  expected,	
  perhaps	
  showing	
  a	
  need	
  to	
  extend	
  the	
  measurements	
  of	
  the	
  
performance	
  environment	
  to	
  capture	
  the	
  process	
  for	
  longer	
  period	
  of	
  time	
  to	
  see	
  if	
  self-­‐
efficacy	
  would	
  recover.	
  There	
  was	
  also	
  a	
  strong	
  relationship	
  between	
  self-­‐efficacy	
  and	
  
goals,	
  even	
  though	
  only	
  a	
  weak	
  one	
  was	
  anticipated.	
  This	
  may	
  have	
  been	
  due	
  to	
  both	
  
variables	
  being	
  low	
  in	
  level,	
  but	
  still	
  following	
  a	
  small	
  sawtooth	
  pattern,	
  possibly	
  resulting	
  
in	
  even	
  small	
  changes	
  in	
  the	
  variables	
  but	
  in	
  the	
  same	
  direction,	
  which	
  is	
  likely	
  the	
  
contributing	
  factor	
  for	
  the	
  strong	
  relationship.	
  Most	
  surprisingly,	
  was	
  the	
  positive	
  
relationship	
  between	
  self-­‐efficacy	
  and	
  outcome	
  effort.	
  The	
  literature	
  suggests	
  that	
  there	
  
would	
  be	
  a	
  negative	
  relationship,	
  since	
  overestimation	
  of	
  ability	
  would	
  lead	
  to	
  less	
  effort	
  

	
  

121	
  

	
  
being	
  devoted	
  to	
  the	
  task.	
  The	
  findings	
  suggest	
  that	
  there	
  may	
  have	
  been	
  insufficient	
  time	
  
to	
  capture	
  the	
  performance	
  process	
  or	
  it	
  may	
  have	
  been	
  an	
  artifact	
  of	
  both	
  variables	
  having	
  
less	
  variance	
  over	
  time.	
  However,	
  given	
  the	
  sawtooth	
  trajectories,	
  the	
  small	
  fluctuations	
  in	
  
the	
  same	
  direction	
  may	
  have	
  caused	
  the	
  resultant	
  significant	
  relationship.	
  
	
  

Although	
  much	
  of	
  the	
  explication	
  above	
  has	
  been	
  focused	
  on	
  the	
  interesting,	
  

unexpected	
  elements	
  of	
  the	
  results,	
  the	
  preponderance	
  of	
  the	
  evidence	
  suggests	
  that	
  the	
  
findings	
  are	
  consistent	
  with	
  many	
  aspects	
  of	
  the	
  theory.	
  This	
  is	
  evident	
  in	
  that	
  although	
  
several	
  hypotheses	
  were	
  not	
  supported,	
  they	
  were	
  in	
  a	
  promising	
  direction	
  where	
  the	
  
relationship	
  was	
  strong	
  instead	
  of	
  weak	
  (or	
  vice	
  versa,	
  though	
  that	
  was	
  much	
  less	
  
common)	
  or	
  when	
  the	
  weak	
  relationship	
  (i.e.,	
  non-­‐significant)	
  was	
  positive	
  rather	
  than	
  
negative.	
  This	
  is	
  promising	
  given	
  that	
  a	
  null	
  relationship	
  was	
  expected	
  in	
  any	
  event,	
  and	
  
also	
  since	
  this	
  was	
  the	
  first	
  attempt	
  to	
  specify	
  the	
  relationship	
  changes	
  between	
  processes	
  
in	
  as	
  detailed	
  a	
  way	
  as	
  to	
  present	
  weak	
  and	
  strong	
  relationships.	
  The	
  next	
  section	
  will	
  take	
  
a	
  more	
  macro	
  view	
  of	
  the	
  findings	
  to	
  discuss	
  the	
  ramifications	
  of	
  the	
  results.	
  
	
  
Overall	
  Observations	
  of	
  the	
  Results	
  
	
  

First,	
  motivation	
  was	
  more	
  important	
  in	
  the	
  adaptation	
  process	
  than	
  expected.	
  This	
  

suggests	
  that	
  more	
  motivated	
  individuals,	
  regardless	
  of	
  their	
  overconfident	
  state,	
  were	
  able	
  
to	
  increase	
  their	
  performance	
  in	
  an	
  adaptive	
  environment.	
  Thus,	
  even	
  in	
  the	
  face	
  of	
  a	
  
change,	
  maintaining	
  confidence	
  in	
  one’s	
  ability,	
  pursuing	
  high	
  goals,	
  and	
  putting	
  forth	
  effort	
  
remained	
  just	
  as	
  valuable	
  in	
  the	
  adaptation	
  process	
  as	
  it	
  was	
  in	
  the	
  performance	
  process,	
  
despite	
  participants’	
  not	
  fully	
  understanding	
  the	
  impact	
  of	
  their	
  behaviors,	
  as	
  they	
  would	
  in	
  
a	
  more	
  typical	
  performance	
  environment.	
  

	
  

122	
  

	
  
	
  

Second,	
  cognition	
  was	
  more	
  important	
  in	
  the	
  performance	
  process	
  than	
  expected.	
  

This	
  suggests	
  that	
  maintaining	
  a	
  level	
  of	
  understanding	
  of	
  the	
  impact	
  of	
  one’s	
  actions	
  on	
  
performance	
  outcomes	
  was	
  still	
  beneficial	
  in	
  a	
  performance	
  environment.	
  This	
  may	
  also	
  
indicate	
  that	
  there	
  was	
  insufficient	
  time	
  to	
  capture	
  the	
  performance	
  process,	
  particularly	
  
since	
  performance	
  followed	
  a	
  sawtooth	
  pattern.	
  It	
  is	
  possible	
  that	
  this	
  situation	
  may	
  have	
  
confused	
  the	
  participants,	
  causing	
  them	
  to	
  believe	
  that	
  they	
  were	
  not	
  operating	
  in	
  a	
  truly	
  
routine	
  performance	
  environment,	
  but	
  rather	
  one	
  that	
  was	
  more	
  dynamic	
  than	
  it	
  was	
  
before	
  the	
  change.	
  Although	
  this	
  was	
  not	
  the	
  case	
  and	
  the	
  task	
  scenarios	
  fluctuated	
  just	
  as	
  
they	
  had	
  before	
  the	
  change,	
  the	
  additional	
  complexity	
  of	
  the	
  task,	
  coupled	
  with	
  small	
  task	
  
variations	
  (e.g.,	
  positioning	
  of	
  the	
  targets,	
  changing	
  the	
  types	
  of	
  targets),	
  may	
  have	
  given	
  
the	
  impression	
  that	
  the	
  task	
  had	
  not	
  yet	
  stabilized	
  and	
  changes	
  were	
  consistently	
  being	
  
introduced.	
  
	
  

Third,	
  although	
  the	
  processes	
  appear	
  similar	
  at	
  first	
  glance	
  from	
  the	
  results	
  presented	
  

here,	
  it	
  is	
  likely	
  that	
  they	
  are	
  still	
  distinct.	
  The	
  number	
  of	
  trials	
  given	
  in	
  this	
  experimental	
  
design	
  were	
  likely	
  insufficient	
  to	
  capture	
  the	
  performance	
  process,	
  particularly	
  given	
  that	
  
performance	
  in	
  the	
  coordinative	
  complexity	
  condition	
  did	
  not	
  stabilize.	
  This	
  is	
  a	
  limitation	
  
of	
  a	
  laboratory	
  design,	
  as	
  fatigue	
  was	
  already	
  evident	
  and	
  extended	
  time	
  would	
  likely	
  not	
  
be	
  possible.	
  Therefore,	
  a	
  field	
  setting	
  would	
  provide	
  insight	
  into	
  what	
  the	
  processes	
  may	
  
look	
  like	
  without	
  the	
  fatigue	
  factor,	
  which	
  may	
  have	
  caused	
  self-­‐efficacy	
  and	
  goals	
  to	
  
stabilize	
  at	
  such	
  a	
  low	
  level	
  during	
  this	
  study.	
  Although	
  these	
  items	
  would	
  not	
  be	
  able	
  to	
  be	
  
measured	
  in	
  the	
  same	
  way,	
  it	
  is	
  likely	
  that	
  variables,	
  such	
  as	
  self-­‐efficacy,	
  would	
  rebound	
  
slightly,	
  or	
  individuals	
  would	
  voluntarily	
  (or	
  involuntarily)	
  leave	
  the	
  organization.	
  
Alternatively,	
  the	
  possibility	
  that	
  participants	
  may	
  have	
  felt	
  as	
  if	
  they	
  were	
  not	
  in	
  a	
  routine	
  

	
  

123	
  

	
  
performance	
  environment	
  (given	
  the	
  sawtooth	
  trajectory	
  of	
  performance),	
  as	
  discussed	
  
earlier,	
  could	
  have	
  resulted	
  in	
  the	
  variables	
  having	
  more	
  fluctuations	
  in	
  the	
  performance	
  
process	
  than	
  expected,	
  resulting	
  in	
  stronger	
  (and	
  possibly	
  more	
  positive)	
  relationships	
  in	
  
both	
  processes.	
  This	
  may	
  have	
  contributed	
  to	
  the	
  similarities	
  evident	
  in	
  the	
  results.	
  
	
  

Finally,	
  and	
  most	
  critically,	
  the	
  differences	
  in	
  the	
  conclusions	
  of	
  the	
  trajectory	
  changes	
  

as	
  compared	
  to	
  the	
  relationship	
  changes	
  strongly	
  indicate	
  that	
  it	
  is	
  necessary	
  for	
  research	
  
to	
  be	
  conducted	
  not	
  only	
  beyond	
  investigating	
  performance	
  trajectories	
  but	
  also	
  expanded	
  
beyond	
  exploring	
  the	
  trajectories	
  of	
  the	
  self-­‐regulatory	
  variables	
  to	
  truly	
  capture	
  the	
  
dynamics	
  of	
  processes.	
  This	
  is	
  probably	
  the	
  most	
  essential	
  contribution	
  of	
  this	
  dissertation	
  
in	
  that	
  it	
  empirically	
  explicates	
  that	
  investigating	
  trajectory	
  changes	
  is	
  insufficient	
  for	
  
examining	
  a	
  process.	
  Without	
  examining	
  the	
  relationships,	
  it	
  would	
  have	
  been	
  assumed	
  
from	
  the	
  trajectories	
  of	
  the	
  motivational	
  variables,	
  that	
  there	
  would	
  be	
  a	
  negative	
  
relationship	
  with	
  performance	
  in	
  both	
  of	
  the	
  processes.	
  However,	
  the	
  cross-­‐lag	
  analysis	
  
shows	
  that	
  this	
  was	
  not	
  the	
  case,	
  likely	
  because	
  this	
  analysis	
  examines	
  the	
  data	
  in	
  a	
  very	
  
different	
  way.	
  First,	
  it	
  accounts	
  for	
  the	
  variance	
  due	
  to	
  the	
  variable	
  at	
  the	
  last	
  time	
  point	
  
tending	
  to	
  remain	
  on	
  the	
  same	
  path	
  at	
  the	
  next	
  point	
  (the	
  autoregressive	
  component).	
  
Second,	
  it	
  allows	
  the	
  remaining	
  variance	
  to	
  then	
  be	
  distributed	
  among	
  the	
  variables	
  that	
  
are	
  predicting	
  it.	
  Therefore,	
  even	
  though	
  goals	
  were	
  decreasing	
  in	
  the	
  adaptation	
  process	
  
while	
  performance	
  was	
  increasing,	
  there	
  can	
  still	
  be	
  a	
  positive	
  relationship	
  with	
  the	
  
remaining	
  variance.	
  This	
  is	
  directly	
  interpreted	
  as:	
  controlling	
  for	
  the	
  fact	
  that	
  previously	
  
low	
  levels	
  of	
  goals	
  tended	
  to	
  elicit	
  similarly	
  low	
  goals	
  in	
  the	
  subsequent	
  trial,	
  the	
  changes	
  in	
  
goal	
  level	
  were	
  positively	
  related	
  to	
  performance	
  such	
  that	
  as	
  performance	
  increased	
  
between	
  those	
  two	
  trials,	
  and	
  goals	
  followed	
  suit	
  and	
  deviated	
  from	
  the	
  path	
  that	
  they	
  

	
  

124	
  

	
  
would	
  take	
  otherwise.	
  This	
  research	
  not	
  only	
  is	
  among	
  the	
  first	
  to	
  provide	
  specific	
  
dynamics	
  in	
  the	
  theorizing	
  of	
  the	
  adaptation	
  process,	
  but	
  it	
  is	
  the	
  first	
  endeavor	
  to	
  utilize	
  
an	
  analytical	
  model	
  to	
  test	
  whether	
  the	
  anticipated	
  dynamics	
  provided	
  an	
  accurate	
  picture	
  
of	
  the	
  process.	
  
	
  
	
  Practical	
  Implications	
  
	
  

As	
  it	
  was	
  mentioned	
  briefly	
  above,	
  this	
  research	
  extends	
  the	
  adaptation	
  literature	
  by	
  

not	
  only	
  adding	
  to	
  the	
  few	
  efforts	
  that	
  have	
  conceptualized	
  adaptation	
  as	
  a	
  process	
  
incorporating	
  self-­‐regulatory	
  mechanisms,	
  but	
  it	
  is	
  the	
  first	
  to	
  specifically	
  theorize	
  the	
  
trajectory	
  and	
  relationship	
  dynamics	
  involved.	
  Furthermore,	
  this	
  study	
  utilized	
  analyses	
  
that	
  appropriately	
  examined	
  a	
  process	
  framework.	
  This	
  will	
  hopefully	
  encourage	
  and	
  assist	
  
future	
  researchers	
  to	
  make	
  similarly	
  specific	
  dynamic	
  hypotheses	
  and	
  test	
  them	
  
appropriately.	
  The	
  utility	
  in	
  examining	
  adaptation	
  as	
  a	
  process	
  does	
  not	
  merely	
  lie	
  in	
  its	
  
novelty,	
  but	
  it	
  is	
  because	
  processes	
  are	
  typically	
  generalizable,	
  at	
  least	
  to	
  a	
  certain	
  extent,	
  
and	
  processes	
  can	
  be	
  influenced	
  (e.g.,	
  through	
  training),	
  unlike	
  an	
  individual	
  difference	
  or	
  a	
  
performance	
  dimension.	
  The	
  similarity	
  between	
  the	
  trajectories,	
  and	
  relationships	
  
between	
  the	
  two	
  complexity	
  conditions,	
  lend	
  further	
  support	
  to	
  the	
  robustness	
  of	
  the	
  
adaptation	
  process	
  in	
  explaining	
  behavior	
  across	
  different	
  types	
  of	
  changes.	
  Therefore,	
  
contributing	
  to	
  this	
  perspective	
  of	
  adaptation	
  has	
  the	
  potential	
  to	
  cause	
  change	
  in	
  behavior,	
  
not	
  just	
  to	
  understand	
  behavior.	
  
The	
  findings	
  not	
  only	
  have	
  relevance	
  to	
  theory,	
  but	
  also	
  to	
  practice.	
  As	
  mentioned	
  
above,	
  with	
  a	
  process	
  perspective	
  to	
  adaptation	
  there	
  is	
  opportunity	
  for	
  manipulations	
  to	
  
be	
  implemented	
  to	
  enhance	
  the	
  outcome	
  of	
  the	
  process	
  or	
  possible	
  speed	
  with	
  which	
  it	
  

	
  

125	
  

	
  
occurs.	
  Previous	
  research	
  has	
  demonstrated	
  that	
  training	
  impacts	
  adaptive	
  performance	
  
(e.g.,	
  Baard,	
  2013;	
  Bell	
  &	
  Kozlowski,	
  2008).	
  Therefore,	
  it	
  may	
  be	
  that	
  training	
  can	
  enhance	
  
the	
  process	
  itself	
  as	
  well.	
  Cognitively	
  focused	
  training	
  may	
  assist	
  individuals	
  in	
  where	
  to	
  
look	
  for	
  information	
  should	
  an	
  adaptive	
  event	
  occur,	
  prepare	
  them	
  in	
  how	
  to	
  test	
  different	
  
strategies	
  without	
  wasting	
  valuable	
  time	
  and	
  resources,	
  and	
  help	
  them	
  recognize	
  when	
  to	
  
stop	
  strategizing	
  and	
  gathering	
  information	
  and	
  maximize	
  on	
  performance	
  as	
  over-­‐
strategizing	
  can	
  hurt	
  adaptive	
  performance.	
  Motivationally	
  focused	
  training	
  can	
  help	
  
encourage	
  employees	
  to	
  keep	
  putting	
  forth	
  effort	
  despite	
  performance	
  drops	
  and	
  push	
  
through	
  fatigue	
  or	
  discouragement	
  that	
  may	
  accompany	
  change	
  as	
  maintaining	
  high	
  goals	
  
and	
  effort	
  was	
  seen	
  to	
  be	
  a	
  critical	
  component	
  for	
  effective	
  performance	
  during	
  adaptation.	
  
	
  
Limitations	
  and	
  Future	
  Research	
  	
  
One	
  of	
  the	
  primary	
  limitations	
  of	
  this	
  study	
  was	
  the	
  number	
  of	
  adaption	
  trials.	
  
Although	
  this	
  study	
  goes	
  beyond	
  most	
  of	
  the	
  literature	
  in	
  measuring	
  15	
  adaptation	
  
performance	
  trials,	
  and	
  is	
  the	
  only	
  study	
  in	
  the	
  adaptation	
  literature	
  to	
  examine	
  self-­‐
regulatory	
  processes	
  for	
  an	
  extended	
  period	
  of	
  time	
  after	
  a	
  change,	
  it	
  appears	
  that	
  seven	
  
trials	
  was	
  not	
  sufficient	
  to	
  capture	
  the	
  performance	
  process.	
  Given	
  the	
  lack	
  of	
  performance	
  
stabilization	
  in	
  the	
  coordinative	
  complexity	
  condition,	
  it	
  appears	
  that	
  equilibrium	
  in	
  
performance	
  was	
  not	
  yet	
  met,	
  resulting	
  in	
  an	
  incomplete	
  picture	
  of	
  the	
  performance	
  
process.	
  Future	
  researchers	
  should	
  extend	
  their	
  research	
  design	
  to	
  incorporate	
  more	
  
measurements	
  after	
  introducing	
  the	
  change.	
  	
  
In	
  that	
  same	
  line	
  of	
  thinking,	
  given	
  the	
  lack	
  of	
  consistency	
  in	
  performance	
  during	
  all	
  
of	
  the	
  adaptation	
  trials,	
  it	
  would	
  be	
  beneficial	
  to	
  attempt	
  to	
  create	
  a	
  more	
  stable	
  task	
  

	
  

126	
  

	
  
environment.	
  The	
  sawtooth	
  pattern	
  may	
  have	
  been	
  a	
  result	
  of	
  the	
  scenarios	
  being	
  slightly	
  
altered	
  in	
  order	
  to	
  make	
  the	
  task	
  more	
  engaging	
  for	
  individuals	
  so	
  they	
  did	
  not	
  cease	
  to	
  
investigate	
  information	
  about	
  the	
  targets	
  when	
  they	
  recognized	
  the	
  pattern.	
  However,	
  it	
  is	
  
possible	
  that	
  individuals	
  attempted	
  to	
  engage	
  with	
  the	
  task	
  as	
  if	
  it	
  were	
  a	
  recurring	
  pattern	
  
every	
  trial,	
  resulting	
  in	
  some	
  key	
  targets	
  being	
  misidentified	
  every	
  other	
  trial.	
  	
  
Furthermore,	
  the	
  laboratory	
  setting	
  itself	
  presents	
  a	
  limitation	
  to	
  the	
  study	
  of	
  
adaptation.	
  One	
  reason	
  is	
  that	
  additional	
  trials	
  were	
  not	
  possible	
  in	
  a	
  laboratory	
  design	
  
given	
  the	
  fatigue	
  individuals	
  reported	
  and	
  splitting	
  adaptive	
  performance	
  trials	
  was	
  not	
  
deemed	
  feasible.	
  	
  In	
  addition,	
  a	
  lab	
  presents	
  a	
  very	
  sterile	
  environment	
  for	
  the	
  participants	
  
and	
  it	
  may	
  have	
  been	
  that	
  the	
  reward	
  was	
  not	
  great	
  enough	
  to	
  merit	
  continued	
  effort	
  in	
  the	
  
face	
  of	
  a	
  difficult	
  change.	
  This	
  may	
  have	
  contributed	
  to	
  why	
  a	
  sawtooth	
  pattern	
  of	
  
performance	
  was	
  seen.	
  If	
  individuals	
  were	
  not	
  engaged	
  or	
  thought	
  they	
  were	
  being	
  tricked,	
  
as	
  they	
  often	
  are	
  in	
  studies,	
  they	
  may	
  not	
  have	
  tried	
  to	
  adjust	
  their	
  behaviors,	
  but	
  rather	
  
came	
  to	
  expect	
  a	
  pattern	
  of	
  performance.	
  Therefore,	
  it	
  is	
  recommended	
  that	
  future	
  
research	
  attempt	
  to	
  attempt	
  to	
  make	
  experimental	
  platform	
  as	
  similar	
  to	
  what	
  individuals	
  
would	
  experience	
  on	
  a	
  job	
  as	
  much	
  as	
  possible.	
  For	
  instance,	
  measuring	
  individuals	
  over	
  
the	
  course	
  of	
  a	
  semester	
  would	
  allow	
  for	
  an	
  expanded	
  view	
  of	
  the	
  adaptation	
  and	
  
performance	
  processes	
  where	
  the	
  period	
  of	
  time	
  would	
  now	
  extend	
  to	
  months	
  rather	
  than	
  
days.	
  Furthermore,	
  such	
  a	
  setting	
  may	
  have	
  more	
  relevant	
  tasks	
  (e.g.,	
  papers	
  and	
  projects)	
  
and	
  rewards	
  (e.g.,	
  grades)	
  for	
  individuals,	
  presenting	
  them	
  with	
  a	
  stronger	
  reason	
  to	
  
effectively	
  adapt.	
  
Finally,	
  additional	
  research	
  can	
  be	
  done	
  on	
  the	
  utility	
  of	
  the	
  partially	
  crossed,	
  
partially	
  time	
  varying	
  cross-­‐lag	
  panel	
  regression	
  model.	
  Although	
  DeShon	
  (2012)	
  posed	
  the	
  

	
  

127	
  

	
  
utility	
  of	
  dynamic	
  analyses	
  to	
  investigate	
  phenomena	
  such	
  as	
  the	
  complex	
  relationship	
  
between	
  self-­‐efficacy	
  and	
  performance,	
  it	
  appears	
  that	
  research	
  has	
  not	
  yet	
  started	
  
applying	
  these	
  dynamic	
  analyses	
  in	
  empirical	
  works.	
  The	
  cross-­‐lag	
  analysis	
  used	
  in	
  this	
  
research	
  is	
  one	
  of	
  the	
  first	
  to	
  attempt	
  to	
  bring	
  dynamic	
  analyses	
  to	
  the	
  organizational	
  
literature	
  to	
  investigate	
  processes	
  dynamically.	
  However,	
  there	
  is	
  still	
  much	
  to	
  learn	
  about	
  
the	
  model,	
  assumptions,	
  limitations,	
  and	
  applications.	
  The	
  series	
  of	
  simulations	
  conducted	
  
for	
  this	
  study	
  in	
  Appendix	
  A	
  presented	
  a	
  small	
  set	
  of	
  parameters	
  that	
  can	
  be	
  investigated.	
  	
  
One	
  key	
  limitation	
  of	
  the	
  cross	
  lag	
  panel	
  model	
  is	
  that	
  it	
  is	
  a	
  pooled	
  model	
  of	
  the	
  
individual	
  dynamics,	
  making	
  an	
  assumption	
  that	
  each	
  individual	
  has	
  the	
  same	
  dynamic	
  
system.	
  This	
  assumption	
  was	
  retained	
  in	
  the	
  current	
  study,	
  as	
  the	
  goal	
  of	
  this	
  research	
  was	
  
to	
  capture	
  the	
  overall	
  performance	
  adaptation	
  process.	
  Although	
  the	
  cross	
  lag	
  model	
  is	
  
very	
  sophisticated	
  and	
  extremely	
  informative	
  about	
  the	
  overall	
  dynamic	
  system,	
  it	
  does	
  not	
  
provide	
  an	
  easy	
  way	
  to	
  understand	
  the	
  dynamics	
  of	
  the	
  individuals.	
  There	
  are,	
  however,	
  
two	
  extensions	
  of	
  this	
  analysis	
  that	
  provide	
  avenues	
  for	
  future	
  research.	
  First,	
  instead	
  of	
  
averaging	
  the	
  paths	
  across	
  the	
  trials	
  in	
  the	
  adaptation	
  process,	
  it	
  is	
  possible	
  to	
  unconstrain	
  
the	
  paths	
  of	
  the	
  relationships	
  to	
  allow	
  them	
  to	
  change	
  over	
  each	
  time	
  period.	
  In	
  other	
  
words,	
  in	
  model	
  used	
  in	
  this	
  research,	
  two	
  sets	
  of	
  relationships	
  were	
  obtained,	
  
representing	
  the	
  adaptation	
  and	
  performance	
  processes	
  in	
  Trials	
  19-­‐25	
  and	
  Trials	
  26-­‐33,	
  
respectively.	
  However,	
  future	
  research	
  could	
  estimate	
  different	
  relationships	
  between	
  each	
  
trial,	
  resulting	
  in	
  15	
  sets	
  of	
  relationships,	
  thereby	
  providing	
  a	
  more	
  fine-­‐grained	
  view	
  of	
  
both	
  the	
  adaptation	
  and	
  performance	
  processes.	
  
Second,	
  there	
  is	
  a	
  need	
  to	
  explore	
  the	
  person	
  in	
  the	
  process.	
  That	
  is,	
  the	
  current	
  
study	
  did	
  not	
  capture	
  either	
  the	
  individuals’	
  phenomenological	
  experiences	
  when	
  they	
  

	
  

128	
  

	
  
were	
  exposed	
  to	
  the	
  change	
  (e.g.,	
  How	
  did	
  they	
  feel?	
  Stressed?	
  Confident?),	
  nor	
  did	
  it	
  focus	
  
on	
  the	
  individual	
  trajectories,	
  though	
  there	
  was	
  evidence	
  of	
  differences	
  among	
  individuals	
  
in	
  how	
  they	
  responded	
  with	
  regard	
  to	
  the	
  self-­‐regulatory	
  variables	
  under	
  investigation.	
  
With	
  regard	
  to	
  the	
  first,	
  when	
  individuals	
  are	
  faced	
  with	
  a	
  change,	
  several	
  responses	
  can,	
  
theoretically,	
  ensue.	
  Individuals	
  may	
  feel	
  intensely	
  motivated	
  to	
  adapt	
  and	
  put	
  extra	
  effort	
  
toward	
  pursuing	
  a	
  high	
  goal	
  to	
  re-­‐attain	
  their	
  pre-­‐change	
  performance.	
  Others	
  may	
  feel	
  
defeated	
  by	
  a	
  large	
  drop	
  in	
  performance	
  and	
  either	
  give	
  up	
  or	
  settle	
  for	
  a	
  very	
  low	
  
performance	
  level.	
  Although	
  these	
  ideas	
  cannot	
  be	
  answered	
  with	
  the	
  data	
  in	
  this	
  research,	
  
Figure	
  13	
  provides	
  a	
  brief	
  overview	
  of	
  what	
  the	
  individuals’	
  performance	
  dynamics	
  were	
  in	
  
current	
  study	
  in	
  the	
  two	
  adaptive	
  conditions:	
  component	
  and	
  coordinative	
  complexity	
  
increases.	
  The	
  figure	
  shows	
  not	
  only	
  the	
  mean	
  trajectories,	
  as	
  presented	
  earlier,	
  but	
  also	
  
the	
  error	
  bars	
  display	
  the	
  differences	
  in	
  individual	
  responses	
  across	
  the	
  trials	
  in	
  the	
  
adaptation	
  and	
  performance	
  environments.	
  To	
  make	
  this	
  even	
  more	
  explicit,	
  the	
  “spaghetti	
  
plots”	
  below	
  those	
  graphs	
  show	
  the	
  change	
  in	
  performance	
  levels	
  of	
  30	
  of	
  the	
  individuals	
  in	
  
each	
  condition.	
  These	
  plots	
  show	
  large	
  variability	
  in	
  how	
  individuals	
  change	
  in	
  their	
  
performance	
  levels	
  over	
  time,	
  suggesting	
  that	
  individuals	
  may	
  follow	
  different	
  patterns	
  in	
  
their	
  processing	
  of	
  a	
  change.	
  	
  
In	
  addition	
  to	
  the	
  differential	
  effect	
  on	
  performance,	
  individuals	
  may	
  also	
  have	
  
different	
  affective,	
  cognitive	
  or	
  motivational	
  reactions.	
  As	
  referenced	
  to	
  above,	
  a	
  high	
  stress	
  
situation	
  may	
  elicit	
  a	
  fight	
  reaction	
  in	
  some,	
  where	
  these	
  individuals	
  become	
  excited	
  and	
  
determined.	
  However,	
  others,	
  having	
  a	
  flight	
  reaction,	
  may	
  shut	
  down	
  and	
  no	
  longer	
  care	
  
about	
  the	
  task.	
  It	
  may	
  also	
  be	
  that	
  these	
  motivational	
  or	
  affective	
  components	
  have	
  
differing	
  relationships	
  with	
  performance	
  or	
  with	
  other	
  self-­‐regulatory	
  variables	
  in	
  the	
  face	
  

	
  

129	
  

	
  
of	
  a	
  change.	
  For	
  instance,	
  some	
  individuals	
  may	
  consistently	
  have	
  large	
  goal-­‐performance	
  
discrepancies;	
  others	
  may	
  have	
  consistently	
  small	
  discrepancies;	
  and	
  still	
  others	
  may	
  
change	
  in	
  their	
  responses	
  over	
  time	
  where	
  they	
  start	
  with	
  large	
  discrepancies	
  and	
  then	
  
calibrate	
  their	
  goals	
  more	
  effectively	
  as	
  the	
  learn	
  about	
  the	
  change.	
  It	
  would	
  be	
  interesting	
  
to	
  examine	
  whether	
  these	
  differences	
  impact	
  the	
  relationships	
  in	
  the	
  overall	
  adaptation	
  
process.	
  In	
  order	
  to	
  demonstrate	
  the	
  differences	
  the	
  individual	
  variability	
  in	
  the	
  self-­‐
regulatory	
  variables,	
  Figure	
  14	
  adds	
  error	
  bars	
  to	
  the	
  graphs	
  of	
  the	
  trajectories	
  of	
  these	
  
variables	
  in	
  the	
  adaptive	
  trials.	
  It	
  is	
  clear	
  that	
  individuals	
  differ	
  in	
  their	
  responses,	
  but	
  it	
  is	
  
unclear	
  why	
  these	
  differences	
  appear.	
  Were	
  some	
  individuals	
  more	
  engaged	
  than	
  others?	
  
Were	
  some	
  more	
  stressed	
  than	
  others?	
  Did	
  the	
  different	
  types	
  of	
  changes	
  trigger	
  different	
  
responses	
  among	
  individuals,	
  causing	
  different	
  patterns	
  to	
  emerge?	
  	
  
These	
  questions,	
  among	
  others,	
  would	
  be	
  fruitful	
  avenues	
  for	
  additional	
  research	
  
and	
  it	
  is	
  strongly	
  encouraged	
  that	
  future	
  endeavors	
  explore	
  the	
  differences	
  among	
  
individuals’	
  dynamics	
  in	
  the	
  investigation	
  of	
  the	
  process	
  of	
  adaptation.	
  Specifically,	
  it	
  is	
  
recommended	
  that	
  future	
  researchers	
  investigate	
  several	
  questions	
  with	
  regard	
  to	
  
individual	
  variability.	
  First,	
  researchers	
  must	
  determine	
  what	
  types	
  of	
  patterns	
  individuals	
  
may	
  follow	
  in	
  response	
  to	
  a	
  change	
  (e.g.,	
  a	
  failure	
  to	
  adapt,	
  the	
  decision	
  to	
  adopt	
  a	
  lower	
  
standard,	
  or	
  a	
  desire	
  to	
  increase	
  in	
  performance	
  and	
  thrive	
  in	
  adaptation).	
  Then	
  efforts	
  can	
  
be	
  devoted	
  to	
  understanding	
  whether	
  individual	
  differences	
  impact	
  or	
  predict	
  those	
  
patterns	
  (e.g.,	
  individuals	
  with	
  higher	
  learning	
  goal	
  orientations	
  tend	
  to	
  follow	
  the	
  “desire	
  
to	
  increase	
  in	
  performance	
  and	
  thrive	
  in	
  adaptation”	
  pattern).	
  Finally,	
  researchers	
  can	
  
examine	
  whether	
  these	
  patterns	
  are	
  also	
  seen	
  in	
  differences	
  in	
  the	
  relationships	
  between	
  
the	
  self-­‐regulatory	
  variables	
  involved	
  in	
  the	
  adaptation	
  process.	
  	
  

	
  

130	
  

	
  
Figure	
  13	
  
	
  
Individual	
  Variability	
  in	
  Performance	
  Trajectories	
  

	
  

	
  

	
  

	
  
	
  

131	
  

	
  
Figure	
  14	
  
Individual	
  Variability	
  in	
  Self-­‐Regulatory	
  Variables	
  	
  

	
  

132	
  

	
  
Conclusion	
  
	
  

The	
  primary	
  purpose	
  of	
  this	
  research	
  effort	
  was	
  to	
  extend	
  the	
  theory	
  on	
  the	
  process	
  

of	
  adaptation	
  to	
  incorporate	
  specific	
  dynamics,	
  and	
  to	
  test	
  these	
  dynamics	
  with	
  the	
  most	
  
appropriate	
  analyses	
  available.	
  The	
  results	
  show	
  support	
  for	
  the	
  majority	
  of	
  the	
  
hypotheses,	
  suggesting	
  that	
  this	
  theory	
  provides	
  a	
  reasonable	
  stepping-­‐stone	
  for	
  additional	
  
work	
  to	
  be	
  conducted	
  on	
  the	
  adaptation	
  process.	
  Given	
  the	
  consistency	
  of	
  the	
  trajectories	
  
and	
  relationship	
  changes	
  in	
  the	
  component	
  and	
  complexity	
  change	
  conditions,	
  the	
  
adaptation	
  and	
  post-­‐change	
  performance	
  processes	
  appear	
  to	
  be	
  consistent	
  across	
  
different	
  types	
  of	
  change.	
  This	
  suggests	
  that	
  these	
  processes	
  are	
  generalizable	
  to	
  multiple	
  
types	
  of	
  change,	
  and	
  that,	
  if	
  trainings	
  or	
  interventions	
  are	
  implemented,	
  organizations	
  
could	
  assist	
  individuals	
  in	
  dealing	
  with	
  different	
  types	
  of	
  adaptive	
  situations.	
  Furthermore,	
  
it	
  is	
  clear	
  from	
  the	
  results	
  that	
  it	
  is	
  critical	
  to	
  push	
  research	
  beyond	
  investigating	
  
trajectories	
  (i.e.,	
  through	
  growth	
  curve	
  analyses)	
  and	
  start	
  pursuing	
  more	
  complex	
  ways	
  of	
  
representing	
  dynamic	
  data.	
  The	
  cross-­‐lag	
  analysis	
  allowed	
  for	
  the	
  investigation	
  of	
  all	
  the	
  
variables	
  in	
  one	
  model,	
  resulting	
  in	
  the	
  conclusion	
  that	
  relationship	
  changes	
  are	
  not	
  fully	
  
driven	
  by	
  trajectory	
  changes,	
  but	
  incorporate	
  autoregressive	
  relationships	
  and	
  the	
  other	
  
relationships	
  accounting	
  for	
  variance	
  in	
  the	
  model.	
  This	
  research	
  provides	
  a	
  first	
  step	
  into	
  
what	
  could	
  be	
  a	
  very	
  fruitful	
  path	
  investigating	
  the	
  dynamics	
  of	
  the	
  adaptation	
  process.	
  

	
  

133	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
APPENDICES

	
  

134	
  

	
  
APPENDIX	
  A:	
  
Longitudinal	
  Cross-­‐Lag	
  Simulations	
  
	
  
In	
  order	
  to	
  examine	
  the	
  limitations	
  of	
  the	
  longitudinal	
  cross-­‐lag	
  simulation,	
  I	
  
conducted	
  a	
  two-­‐part	
  simulation	
  study	
  where	
  I	
  examined	
  the	
  requirements	
  of	
  the	
  model	
  
with	
  respect	
  to	
  the	
  number	
  of	
  time	
  points	
  (T),	
  people	
  (N),	
  error	
  variance,	
  and	
  effect	
  sizes	
  of	
  
the	
  autoregressive	
  and	
  cross-­‐lag	
  estimations.	
  In	
  the	
  first	
  set	
  of	
  simulations	
  with	
  only	
  two	
  
variables	
  with	
  time	
  invariant	
  relationships,	
  I	
  asked	
  the	
  question:	
  are	
  there	
  differences	
  in	
  
the	
  effectiveness	
  of	
  this	
  model	
  in	
  re-­‐creating	
  the	
  data	
  based	
  on	
  the	
  T	
  and	
  the	
  N,	
  and	
  is	
  that	
  
dependent	
  on	
  the	
  effect	
  sizes	
  of	
  the	
  relationships	
  or	
  the	
  error	
  in	
  the	
  data?	
  
In	
  the	
  second	
  set	
  of	
  simulations,	
  I	
  use	
  my	
  findings	
  from	
  the	
  first	
  set	
  of	
  simulations	
  to	
  
test	
  whether	
  the	
  model	
  could	
  handle	
  seven	
  variables	
  and	
  partially	
  time-­‐varying	
  
relationships	
  (i.e.,	
  two	
  sets	
  of	
  relationships	
  –	
  adaptation	
  process	
  and	
  performance	
  process).	
  
In	
  these	
  simulations	
  I	
  asked	
  the	
  question:	
  given	
  the	
  expected	
  effect	
  sizes	
  and	
  the	
  T	
  I	
  will	
  
have	
  in	
  my	
  study,	
  what	
  is	
  the	
  impact	
  of	
  various	
  effect	
  sizes,	
  the	
  N,	
  and	
  when	
  the	
  shift	
  
between	
  the	
  processes	
  occurs?	
  This	
  final	
  part	
  of	
  the	
  question	
  is	
  critical	
  because	
  it	
  is	
  likely	
  
that	
  the	
  individuals	
  in	
  the	
  component	
  complexity	
  condition	
  go	
  through	
  the	
  adaptation	
  
process	
  more	
  rapidly	
  and	
  I	
  wanted	
  to	
  ensure	
  there	
  is	
  sufficient	
  statistical	
  power	
  to	
  detect	
  
the	
  relationships	
  in	
  the	
  adaptation	
  process	
  in	
  only	
  a	
  few	
  trials,	
  even	
  with	
  small	
  effect	
  sizes	
  
and	
  larger	
  error	
  rates.	
  
The	
  results	
  of	
  these	
  simulations	
  were	
  used	
  in	
  the	
  design	
  of	
  the	
  experiment	
  as	
  well	
  as	
  
in	
  the	
  indication	
  of	
  the	
  needed	
  number	
  of	
  participants	
  in	
  each	
  condition,	
  given	
  the	
  expected	
  
effect	
  sizes,	
  ease	
  of	
  complexity	
  change,	
  and	
  duration	
  of	
  the	
  experiment.	
  
Simulation	
  1:	
  Time	
  Invariant	
  Longitudinal	
  Cross-­‐Lag	
  Panel	
  Regression	
  Models	
  With	
  Two	
  
	
  

135	
  

	
  
Variables	
  
Method.	
  In	
  order	
  to	
  examine	
  the	
  impact	
  of	
  the	
  number	
  of	
  time	
  points	
  (T),	
  number	
  of	
  
individuals	
  (N),	
  error	
  variance,	
  and	
  effect	
  sizes	
  of	
  the	
  autoregressive	
  and	
  cross-­‐lag	
  
parameters	
  on	
  the	
  validity	
  of	
  the	
  model	
  estimates,	
  Monte	
  Carlo	
  simulations	
  were	
  used.	
  In	
  
this	
  series	
  of	
  models,	
  two	
  variables	
  with	
  time	
  invariant	
  relationships	
  were	
  simulated	
  with	
  
varying	
  Ts,	
  Ns,	
  and	
  parameter	
  estimates.	
  Variations	
  of	
  the	
  below	
  equations	
  were	
  used	
  to	
  
estimate	
  the	
  parameters.	
  	
  
	
  

	
  

	
  

	
  

Y(t)	
  =	
  ARy(t-­‐1)	
  +	
  CLx(t-­‐1)	
  +	
  e	
  

	
  

	
  

X(t)	
  =	
  ARx(t-­‐1)	
  +	
  CLy(t-­‐1)	
  +	
  e	
  

	
  
Ten	
  datasets	
  were	
  estimated	
  for	
  each	
  of	
  the	
  243	
  models	
  investigated.	
  The	
  models	
  
were	
  based	
  on	
  the	
  manipulation	
  of	
  the	
  above	
  5	
  parameters	
  (note	
  that	
  the	
  autoregressive	
  
and	
  cross-­‐lag	
  parameters	
  were	
  estimated	
  as	
  the	
  same	
  in	
  the	
  simulated	
  data	
  for	
  X	
  and	
  Y).	
  
The	
  number	
  of	
  time	
  points	
  in	
  the	
  model	
  was	
  5,	
  10	
  or	
  15;	
  the	
  number	
  of	
  people	
  was	
  100,	
  
150	
  or	
  200;	
  and	
  the	
  effect	
  sizes,	
  autoregressive	
  and	
  cross-­‐lag	
  parameters	
  were	
  .1,	
  .3,	
  or	
  .5.	
  
The	
  values	
  chosen	
  for	
  these	
  final	
  parameters	
  were	
  based	
  on	
  Cohen	
  1988	
  who	
  provided	
  a	
  
series	
  of	
  benchmarks	
  from	
  which	
  to	
  make	
  comparisons.	
  He	
  presents	
  certain	
  thresholds	
  
that	
  should	
  be	
  used	
  when	
  determining	
  the	
  effect	
  sizes	
  of	
  correlations.	
  In	
  this	
  paper	
  he	
  
proposed	
  that	
  small	
  effects	
  are	
  .1,	
  medium	
  are	
  .3	
  and	
  large	
  are	
  .5.	
  Given	
  that	
  the	
  basis	
  of	
  
cross-­‐lag	
  regression	
  is	
  correlational,	
  I	
  used	
  this	
  framework	
  to	
  examine	
  the	
  impact	
  of	
  
different	
  relationship	
  strengths	
  on	
  the	
  model.	
  
	
  
Monte	
  Carlo	
  Evidence.	
  Table	
  13	
  shows	
  the	
  results	
  of	
  the	
  simulations.	
  The	
  first	
  set	
  of	
  

	
  

136	
  

	
  
columns	
  show	
  the	
  parameters	
  set	
  in	
  the	
  data;	
  the	
  second	
  set	
  of	
  columns	
  show	
  the	
  values	
  of	
  
Y	
  that	
  were	
  estimated	
  from	
  the	
  model;	
  and	
  the	
  third	
  set	
  of	
  columns	
  show	
  the	
  same	
  for	
  X.	
  
When	
  the	
  estimate	
  from	
  the	
  analysis	
  was	
  .01	
  (or	
  greater)	
  less	
  than	
  the	
  initial	
  parameters	
  
set	
  in	
  the	
  simulated	
  data,	
  the	
  value	
  in	
  Table	
  13	
  was	
  bolded;	
  when	
  the	
  estimate	
  was	
  .01	
  (or	
  
greater)	
  more	
  than	
  the	
  simulated	
  data	
  parameter,	
  the	
  value	
  in	
  Table	
  13	
  is	
  bolded	
  and	
  
italicized.	
  This	
  highlights	
  instances	
  of	
  under-­‐	
  and	
  overestimation.	
  
Specifically	
  examining	
  situations	
  where	
  the	
  estimates	
  were	
  beyond	
  the	
  accepted	
  .01	
  
absolute	
  difference,	
  the	
  Monte	
  Carlo	
  simulations	
  suggest	
  that	
  when	
  there	
  was	
  a	
  small	
  T	
  (5)	
  
and	
  a	
  small	
  N	
  (100)	
  and	
  the	
  standard	
  error	
  was	
  moderate	
  (.3)	
  or	
  large	
  (.5),	
  there	
  were	
  
multiple	
  errors	
  in	
  the	
  cross-­‐lag	
  and	
  autoregression	
  estimates.	
  Furthermore,	
  unlike	
  the	
  
other	
  models,	
  where	
  the	
  parameters	
  estimates	
  were	
  statistically	
  significant,	
  when	
  the	
  
standard	
  error	
  was	
  large	
  and	
  the	
  autoregressive	
  parameter	
  was	
  small	
  (.1),	
  the	
  pvalue	
  of	
  the	
  
predicted	
  autoregressive	
  component	
  was	
  not	
  always	
  statistically	
  significant.	
  The	
  
simulations	
  also	
  revealed	
  that	
  when	
  there	
  was	
  a	
  small	
  T	
  (5)	
  and	
  a	
  moderate	
  N	
  (150),	
  when	
  
the	
  standard	
  error	
  was	
  moderate	
  and	
  the	
  autoregressive	
  parameter	
  was	
  large,	
  the	
  cross-­‐
lag	
  estimate	
  can	
  be	
  incorrect;	
  when	
  the	
  standard	
  error	
  is	
  large,	
  both	
  the	
  autoregressive	
  and	
  
the	
  cross-­‐lag	
  estimates	
  can	
  be	
  beyond	
  the	
  .01	
  threshold.	
  Fewer	
  errors	
  were	
  found	
  when	
  
there	
  was	
  a	
  moderate	
  T	
  (10)	
  and	
  a	
  small	
  N	
  (100),	
  although	
  some	
  inaccuracies	
  were	
  
detected	
  in	
  the	
  cross-­‐lag	
  and	
  autoregressive	
  estimates	
  when	
  the	
  standard	
  errors	
  were	
  
moderate	
  or	
  large	
  and	
  the	
  cross-­‐lag	
  or	
  autoregressive	
  parameters	
  were	
  set	
  to	
  be	
  small.	
  
Also,	
  very	
  few	
  errors	
  in	
  the	
  cross-­‐lag	
  and	
  autoregressive	
  estimates	
  were	
  found	
  when	
  there	
  
was	
  a	
  large	
  T	
  (15)	
  and	
  a	
  small	
  N	
  (100)	
  if	
  the	
  standard	
  error	
  was	
  high.	
  	
  
	
  

All	
  other	
  values	
  were	
  within	
  the	
  accepted	
  range,	
  suggesting	
  that	
  the	
  following	
  

	
  

137	
  

	
  
situations	
  provide	
  a	
  level	
  of	
  confidence	
  in	
  the	
  output	
  of	
  the	
  models.	
  When	
  a	
  small	
  T	
  (5)	
  
must	
  be	
  accepted,	
  a	
  large	
  N	
  (200)	
  is	
  needed.	
  If	
  a	
  large	
  N	
  is	
  not	
  possible,	
  a	
  moderate	
  N	
  (150)	
  
is	
  acceptable	
  if	
  the	
  amount	
  of	
  error	
  expected	
  is	
  low	
  to	
  moderate.	
  A	
  moderate	
  or	
  large	
  T	
  is	
  
preferred	
  in	
  these	
  models	
  as	
  they	
  are	
  fairly	
  robust	
  to	
  large	
  error	
  rates	
  in	
  the	
  ability	
  of	
  the	
  
model	
  to	
  accurately	
  estimate	
  even	
  small	
  effect	
  sizes,	
  particularly	
  with	
  moderate	
  to	
  high	
  Ns.	
  
If	
  there	
  were	
  errors	
  made	
  in	
  the	
  estimation,	
  they	
  were	
  typically	
  overestimated	
  by	
  a	
  small	
  
amount	
  when	
  there	
  was	
  a	
  large	
  T,	
  regardless	
  of	
  the	
  N.	
  Finally,	
  in	
  general,	
  the	
  size	
  of	
  the	
  
cross-­‐lags	
  and	
  autoregressive	
  parameters	
  did	
  not	
  have	
  much	
  impact	
  on	
  the	
  estimates.	
  
In	
  conclusion,	
  having	
  five	
  time	
  points	
  is	
  risky,	
  but	
  if	
  that	
  is	
  needed,	
  a	
  large	
  N	
  is	
  
required.	
  Having	
  10	
  time	
  points	
  is	
  preferred,	
  but	
  if	
  there	
  is	
  a	
  possibility	
  of	
  high	
  amounts	
  of	
  
error	
  in	
  the	
  data,	
  a	
  large	
  N	
  or	
  large	
  expected	
  effects	
  sizes	
  are	
  necessary.	
  Finally,	
  having	
  15	
  
time	
  points	
  is	
  the	
  best	
  option	
  as	
  it	
  allows	
  for	
  flexibility	
  in	
  expected	
  error	
  rates	
  and	
  effect	
  
sizes	
  regardless	
  of	
  N.	
  
	
  
Simulation	
  2:	
  Partially	
  Time	
  Varying	
  Longitudinal	
  Cross-­‐Lag	
  Panel	
  Regression	
  Models	
  With	
  
Seven	
  Variables	
  
Method.	
  Based	
  on	
  the	
  first	
  set	
  of	
  simulations,	
  I	
  constrained	
  the	
  next	
  set	
  of	
  
simulations	
  to	
  have	
  15	
  total	
  time	
  points.	
  I	
  also	
  limited	
  the	
  variance	
  of	
  the	
  N	
  to	
  either	
  
moderate	
  (150)	
  or	
  large	
  (200).	
  The	
  amount	
  of	
  error	
  was	
  manipulated	
  (.1,	
  .3,	
  or	
  .5),	
  as	
  was	
  
the	
  point	
  at	
  which	
  the	
  relationships	
  were	
  expected	
  to	
  change.	
  This	
  is	
  based	
  on	
  the	
  
expectation	
  that	
  when	
  individuals	
  are	
  exposed	
  to	
  a	
  change,	
  they	
  engage	
  in	
  an	
  adaptation	
  
process;	
  however,	
  once	
  the	
  source	
  of	
  the	
  change	
  is	
  determined	
  and	
  a	
  strategy	
  is	
  chosen,	
  
individuals	
  will	
  enter	
  a	
  performance	
  process.	
  As	
  these	
  two	
  processes	
  have	
  different	
  

	
  

138	
  

	
  
expected	
  relationships,	
  there	
  are	
  two	
  sets	
  of	
  regression	
  equations	
  estimated	
  (see	
  the	
  below	
  
equations).	
  The	
  autoregressive	
  effects	
  were	
  constrained	
  to	
  be	
  large	
  (.5)	
  as	
  this	
  is	
  a	
  typical	
  
finding	
  in	
  the	
  literature,	
  small	
  hypothesized	
  effects	
  were	
  set	
  as	
  .2	
  while	
  large	
  effects	
  were	
  
simulated	
  as	
  .4,	
  and	
  these	
  effect	
  sizes	
  were	
  constrained	
  to	
  be	
  consistent	
  with	
  the	
  
hypothesized	
  relationships	
  across	
  all	
  models.	
  Most	
  importantly,	
  since	
  it	
  is	
  unknown	
  at	
  
which	
  point	
  individuals	
  will	
  switch	
  from	
  an	
  adaptation	
  to	
  a	
  performance	
  process,	
  this	
  
Monte	
  Carlo	
  simulation	
  was	
  conducted	
  to	
  investigate	
  whether	
  the	
  point	
  at	
  which	
  this	
  
transition	
  will	
  occur	
  will	
  influence	
  the	
  reliability	
  of	
  the	
  model	
  estimates.	
  

	
  

	
  

Performance(t)	
  =	
  ARperformance(t-­‐1)	
  +	
  CLmetacognition(t-­‐1)	
  +	
  CLoutcome-­‐effort(t-­‐1)	
  +	
  e	
  
Learning-­‐effort(t)	
  =	
  ARlearning-­‐effort(t-­‐1)	
  +	
  CLperformance(t-­‐1)	
  +	
  CLevaluation(t-­‐1)	
  +	
  e	
  
Metacognition(t)	
  =	
  ARmetacognition(t-­‐1)	
  +	
  CLlearning-­‐effort	
  (t-­‐1)	
  +	
  CLevaluation	
  (t-­‐1)	
  +	
  e	
  
Evaluation(t)	
  =	
  ARevaluation(t-­‐1)	
  +	
  CLperformance(t-­‐1)	
  +	
  e	
  
Goals(t)	
  =	
  ARgoals(t-­‐1)	
  +	
  CLperformance(t-­‐1)	
  +	
  CLself-­‐efficacy(t-­‐1)	
  +	
  e	
  
Outcome-­‐effort(t)	
  =	
  ARoutcome-­‐effort(t-­‐1)	
  +	
  CLgoals(t-­‐1)	
  +	
  CLself-­‐efficacy(t-­‐1)	
  +	
  e	
  
Self-­‐efficacy(t)	
  =	
  ARself-­‐efficacy(t-­‐1)	
  +	
  CLperformance(t-­‐1)	
  +	
  e	
  
	
  
Ten	
  datasets	
  were	
  estimated	
  for	
  each	
  of	
  the	
  36	
  models	
  investigated.	
  The	
  models	
  
were	
  based	
  on	
  the	
  manipulation	
  of	
  the	
  duration	
  of	
  time	
  individuals	
  are	
  engaging	
  in	
  an	
  
adaptation	
  process	
  (for	
  3,	
  4,	
  5,	
  6,	
  7,	
  or	
  8	
  trials	
  of	
  the	
  15	
  total	
  trials),	
  the	
  number	
  of	
  
individuals	
  in	
  the	
  dataset	
  (150	
  or	
  200),	
  and	
  the	
  error	
  estimate	
  (.1,	
  .3,	
  or	
  .5).	
  	
  
Monte	
  Carlo	
  Evidence.	
  Table	
  14	
  presents	
  the	
  results	
  from	
  the	
  36	
  models.	
  Estimates	
  
were	
  specified	
  for	
  both	
  the	
  adaptation	
  and	
  performance	
  processes.	
  The	
  first	
  column	
  
indicates	
  which	
  variable	
  was	
  being	
  predicted.	
  The	
  column	
  labeled	
  trial	
  change	
  indicates	
  

	
  

139	
  

	
  
which	
  was	
  the	
  first	
  performance	
  process	
  trial	
  (e.g.,	
  if	
  the	
  trial	
  change	
  variable	
  reads	
  7,	
  then	
  
the	
  first	
  6	
  trials	
  where	
  the	
  adaptation	
  process	
  was	
  occurring	
  and	
  Trials	
  7	
  through	
  15	
  were	
  
when	
  the	
  performance	
  process	
  was	
  taking	
  place).	
  Since	
  the	
  autoregressive	
  relationships	
  
were	
  constrained	
  to	
  be	
  the	
  same	
  across	
  both	
  the	
  adaptation	
  and	
  performance	
  process,	
  
there	
  is	
  only	
  one	
  estimate	
  reported	
  for	
  this	
  parameter.	
  The	
  following	
  cross-­‐lags	
  (CL),	
  
standard	
  error	
  (Se),	
  and	
  pvalues	
  (P)	
  are	
  separated	
  by	
  the	
  process	
  investigated	
  and	
  the	
  
relationship	
  estimated.	
  For	
  instance,	
  performance	
  was	
  predicted	
  by	
  the	
  autoregressive	
  
effect	
  of	
  performance,	
  the	
  cross-­‐lag	
  effect	
  of	
  metacognition,	
  and	
  the	
  cross-­‐lag	
  of	
  outcome-­‐
oriented	
  effort.	
  
Similar	
  to	
  the	
  previous	
  set	
  of	
  simulations,	
  instances	
  of	
  under-­‐	
  and	
  overestimation	
  
were	
  determined	
  by	
  the	
  difference	
  between	
  the	
  model	
  estimate	
  and	
  the	
  simulated	
  data	
  
parameter	
  exceeding	
  .01.	
  Underestimates	
  were	
  bolded	
  and	
  overestimates	
  were	
  bolded	
  and	
  
highlighted	
  in	
  Table	
  14.	
  The	
  results	
  of	
  the	
  simulation	
  suggest	
  that	
  not	
  only	
  can	
  the	
  cross-­‐lag	
  
model	
  run	
  with	
  the	
  number	
  of	
  parameters	
  and	
  individuals,	
  but	
  the	
  estimates	
  are	
  within	
  .01	
  
of	
  the	
  specified	
  simulated	
  values.	
  This	
  indicates	
  that	
  the	
  model	
  will	
  be	
  able	
  to	
  analyze	
  the	
  
hypothesized	
  relationships	
  with	
  accuracy.	
  When	
  there	
  is	
  a	
  discrepancy	
  below	
  .01	
  in	
  the	
  
simulated	
  and	
  estimated	
  values,	
  it	
  tends	
  to	
  be	
  when	
  there	
  is	
  a	
  small	
  adaptation	
  process	
  
phase,	
  resulting	
  in	
  fewer	
  trials	
  to	
  estimate	
  the	
  process.	
  However,	
  this	
  occurs	
  rarely	
  and	
  
does	
  not	
  seem	
  to	
  be	
  dependent	
  on	
  the	
  effect	
  size	
  of	
  the	
  variables	
  but	
  rather	
  on	
  the	
  amount	
  
of	
  error	
  apparent	
  in	
  the	
  data.	
  Therefore,	
  minimizing	
  error	
  would	
  be	
  beneficial	
  to	
  the	
  
model’s	
  accuracy,	
  but	
  this	
  set	
  of	
  simulations	
  clearly	
  show	
  that	
  this	
  analysis	
  technique	
  will	
  
be	
  able	
  to	
  reliably	
  test	
  the	
  hypotheses	
  presented	
  in	
  this	
  research	
  endeavor.

	
  

140	
  

	
  

T
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5

	
  

Table	
  13	
  
Fully	
  Crossed	
  Time	
  Invariant	
  Longitudinal	
  Cross-­‐Lag	
  Simulation	
  With	
  Two	
  Variables	
  
Data Simulation Parameters
Model Parameter Estimates (Y)
Model Parameter Estimates (X)
Std
Auto Cross
AR
CL
Se
Se (CL
P
P (CL
AR
CL
Se
Se (CL
P
N
error
reg
-Lag
(Y)
(x->y) (AR Y)
x->y)
(AR Y) x->y)
(X)
(y->x) (AR X)
y->x)
(AR X)
100
0.1
0.1
0.1
0.097
0.105
0.010
0.009
0.000
0.000
0.097
0.106
0.009
0.010
0.000
100
0.1
0.1
0.3
0.099
0.297
0.010
0.009
0.000
0.000
0.102
0.298
0.008
0.009
0.000
100
0.1
0.1
0.5
0.098
0.500
0.009
0.008
0.000
0.000
0.103
0.497
0.008
0.009
0.000
100
0.1
0.3
0.1
0.300
0.098
0.010
0.008
0.000
0.000
0.298
0.101
0.009
0.010
0.000
100
0.1
0.3
0.3
0.301
0.296
0.009
0.008
0.000
0.000
0.300
0.302
0.008
0.009
0.000
100
0.1
0.3
0.5
0.297
0.502
0.008
0.008
0.000
0.000
0.303
0.500
0.008
0.008
0.000
100
0.1
0.5
0.1
0.495
0.101
0.009
0.008
0.000
0.000
0.498
0.096
0.008
0.009
0.000
100
0.1
0.5
0.3
0.499
0.302
0.008
0.008
0.000
0.000
0.499
0.305
0.007
0.008
0.000
100
0.1
0.5
0.5
0.496
0.500
0.008
0.007
0.000
0.000
0.502
0.495
0.007
0.008
0.000
100
0.3
0.1
0.1
0.093
0.117
0.027
0.024
0.010
0.001
0.092
0.114
0.024
0.026
0.002
100
0.3
0.1
0.3
0.096
0.293
0.027
0.024
0.032
0.000
0.105
0.291
0.023
0.025
0.001
100
0.3
0.1
0.5
0.097
0.497
0.024
0.022
0.028
0.000
0.108
0.492
0.022
0.023
0.000
100
0.3
0.3
0.1
0.301
0.098
0.026
0.023
0.000
0.001
0.297
0.099
0.024
0.027
0.000
100
0.3
0.3
0.3
0.299
0.290
0.025
0.022
0.000
0.000
0.301
0.306
0.023
0.025
0.000
100
0.3
0.3
0.5
0.291
0.508
0.022
0.021
0.000
0.000
0.304
0.503
0.021
0.022
0.000
100
0.3
0.5
0.1
0.483
0.101
0.024
0.022
0.000
0.000
0.494
0.092
0.022
0.024
0.000
100
0.3
0.5
0.3
0.492
0.311
0.023
0.021
0.000
0.000
0.498
0.310
0.021
0.023
0.000
100
0.3
0.5
0.5
0.489
0.502
0.021
0.020
0.000
0.000
0.504
0.489
0.020
0.021
0.000
100
0.5
0.1
0.1
0.093
0.128
0.038
0.036
0.072
0.018
0.090
0.117
0.034
0.037
0.048
100
0.5
0.1
0.3
0.095
0.290
0.037
0.035
0.086
0.000
0.106
0.284
0.033
0.036
0.031
100
0.5
0.1
0.5
0.098
0.492
0.034
0.032
0.125
0.000 0.110
0.488
0.032
0.034
0.007
100
0.5
0.3
0.1
0.301
0.100
0.037
0.034
0.000
0.025
0.300
0.096
0.034
0.037
0.000
100
0.5
0.3
0.3
0.295
0.287
0.035
0.033
0.000
0.000
0.302
0.310
0.033
0.036
0.000
100
0.5
0.3
0.5
0.286
0.511
0.032
0.031
0.000
0.000
0.302
0.509
0.031
0.032
0.000
100
0.5
0.5
0.1
0.475
0.100
0.035
0.032
0.000
0.012
0.490
0.090
0.032
0.035
0.000
100
0.5
0.5
0.3
0.485
0.320
0.033
0.031
0.000
0.000
0.498
0.309
0.030
0.033
0.000
100
0.5
0.5
0.5
0.484
0.502
0.031
0.030
0.000
0.000
0.503
0.488
0.029
0.030
0.000
150
0.1
0.1
0.1
0.101
0.097
0.008
0.008
0.000
0.000
0.100
0.101
0.008
0.008
0.000
150
0.1
0.1
0.3
0.102
0.299
0.008
0.007
0.000
0.000
0.099
0.305
0.007
0.008
0.000
150
0.1
0.1
0.5
0.101
0.501
0.007
0.007
0.000
0.000
0.094
0.501
0.007
0.007
0.000
150
0.1
0.3
0.1
0.304
0.101
0.008
0.007
0.000
0.000
0.303
0.094
0.008
0.008
0.000
150
0.1
0.3
0.3
0.299
0.297
0.008
0.007
0.000
0.000
0.302
0.300
0.007
0.007
0.000
150
0.1
0.3
0.5
0.298
0.500
0.007
0.007
0.000
0.000
0.301
0.497
0.007
0.007
0.000

141	
  

P (CL
y->x)
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.008
0.000
0.000
0.007
0.000
0.000
0.005
0.000
0.000
0.070
0.000
0.000
0.087
0.000
0.000
0.061
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000

	
  
Table	
  13	
  (cont’d)	
  

	
  
T
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5

	
  

Data Simulation Parameters
Std
Auto Cross
N
error
reg
-Lag
150
0.1
0.5
0.1
150
0.1
0.5
0.3
150
0.1
0.5
0.5
150
0.3
0.1
0.1
150
0.3
0.1
0.3
150
0.3
0.1
0.5
150
0.3
0.3
0.1
150
0.3
0.3
0.3
150
0.3
0.3
0.5
150
0.3
0.5
0.1
150
0.3
0.5
0.3
150
0.3
0.5
0.5
150
0.5
0.1
0.1
150
0.5
0.1
0.3
150
0.5
0.1
0.5
150
0.5
0.3
0.1
150
0.5
0.3
0.3
150
0.5
0.3
0.5
150
0.5
0.5
0.1
150
0.5
0.5
0.3
150
0.5
0.5
0.5
200
0.1
0.1
0.1
200
0.1
0.1
0.3
200
0.1
0.1
0.5
200
0.1
0.3
0.1
200
0.1
0.3
0.3
200
0.1
0.3
0.5
200
0.1
0.5
0.1
200
0.1
0.5
0.3
200
0.1
0.5
0.5
200
0.3
0.1
0.1
200
0.3
0.1
0.3
200
0.3
0.1
0.5

AR
(Y)
0.500
0.499
0.499
0.102
0.107
0.104
0.310
0.295
0.297
0.501
0.497
0.502
0.103
0.113
0.107
0.314
0.291
0.298
0.504
0.497
0.506
0.100
0.099
0.101
0.296
0.300
0.304
0.499
0.502
0.499
0.101
0.095
0.102

Model Parameter Estimates (Y)
CL
Se
Se (CL
P
(x->y) (AR Y)
x->y)
(AR Y)
0.096
0.007
0.007
0.000
0.304
0.007
0.007
0.000
0.501
0.006
0.006
0.000
0.091
0.022
0.021
0.000
0.298
0.021
0.020
0.002
0.500
0.019
0.019
0.000
0.100
0.021
0.020
0.000
0.291
0.021
0.020
0.000
0.501
0.019
0.018
0.000
0.090
0.019
0.018
0.000
0.310
0.019
0.018
0.000
0.501
0.018
0.017
0.000
0.088
0.031
0.030
0.010
0.299
0.030
0.029
0.009
0.499
0.027
0.027
0.002
0.099
0.030
0.029
0.000
0.287
0.029
0.029
0.000
0.502
0.027
0.026
0.000
0.087
0.027
0.026
0.000
0.314
0.027
0.026
0.000
0.500
0.025
0.025
0.000
0.102
0.007
0.008
0.000
0.303
0.007
0.007
0.000
0.504
0.006
0.006
0.000
0.101
0.006
0.007
0.000
0.299
0.006
0.007
0.000
0.499
0.006
0.006
0.000
0.103
0.006
0.007
0.000
0.301
0.006
0.006
0.000
0.501
0.006
0.006
0.000
0.105
0.018
0.020
0.000
0.307
0.018
0.020
0.000
0.510
0.016
0.018
0.000

142	
  

P (CL
x->y)
0.000
0.000
0.000
0.001
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.032
0.000
0.000
0.014
0.000
0.000
0.011
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000

AR
(X)
0.502
0.499
0.501
0.097
0.096
0.084
0.309
0.302
0.302
0.502
0.496
0.498
0.092
0.093
0.081
0.313
0.300
0.302
0.499
0.494
0.495
0.100
0.101
0.102
0.299
0.301
0.299
0.502
0.499
0.499
0.098
0.103
0.105

Model Parameter Estimates (X)
CL
Se
Se (CL
P
(y->x) (AR X)
y->x)
(AR X)
0.101
0.007
0.007
0.000
0.301
0.007
0.007
0.000
0.501
0.006
0.006
0.000
0.103
0.021
0.022
0.000
0.312
0.020
0.021
0.000
0.502
0.019
0.019
0.000
0.088
0.020
0.021
0.000
0.302
0.019
0.020
0.000
0.495
0.018
0.019
0.000
0.101
0.019
0.020
0.000
0.301
0.018
0.019
0.000
0.508
0.017
0.017
0.000
0.104
0.030
0.031
0.027
0.315
0.029
0.029
0.024
0.501
0.027
0.028
0.013
0.087
0.029
0.030
0.000
0.305
0.028
0.028
0.000
0.496
0.026
0.027
0.000
0.099
0.027
0.028
0.000
0.301
0.026
0.027
0.000
0.516
0.025
0.025
0.000
0.098
0.008
0.007
0.000
0.297
0.007
0.006
0.000
0.496
0.006
0.006
0.000
0.096
0.007
0.007
0.000
0.296
0.007
0.006
0.000
0.501
0.006
0.006
0.000
0.102
0.007
0.006
0.000
0.301
0.006
0.006
0.000
0.500
0.006
0.006
0.000
0.097
0.020
0.018
0.000
0.297
0.019
0.018
0.000
0.490
0.017
0.016
0.000

P (CL
y->x)
0.000
0.000
0.000
0.000
0.000
0.000
0.003
0.000
0.000
0.000
0.000
0.000
0.017
0.000
0.000
0.044
0.000
0.000
0.002
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000

	
  

Data Simulation Parameters
Std
Auto Cross
T
N
error
reg
-Lag
5
200
0.3
0.3
0.1
5
200
0.3
0.3
0.3
5
200
0.3
0.3
0.5
5
200
0.3
0.5
0.1
5
200
0.3
0.5
0.3
5
200
0.3
0.5
0.5
5
200
0.5
0.1
0.1
5
200
0.5
0.1
0.3
5
200
0.5
0.1
0.5
5
200
0.5
0.3
0.1
5
200
0.5
0.3
0.3
5
200
0.5
0.3
0.5
5
200
0.5
0.5
0.1
5
200
0.5
0.5
0.3
5
200
0.5
0.5
0.5
10
100
0.1
0.1
0.1
10
100
0.1
0.1
0.3
10
100
0.1
0.1
0.5
10
100
0.1
0.3
0.1
10
100
0.1
0.3
0.3
10
100
0.1
0.3
0.5
10
100
0.1
0.5
0.1
10
100
0.1
0.5
0.3
10
100
0.1
0.5
0.5
10
100
0.3
0.1
0.1
10
100
0.3
0.1
0.3
10
100
0.3
0.1
0.5
10
100
0.3
0.3
0.1
10
100
0.3
0.3
0.3
10
100
0.3
0.3
0.5
10
100
0.3
0.5
0.1
10
100
0.3
0.5
0.3
10
100
0.3
0.5
0.5

	
  

AR
(Y)
0.291
0.299
0.309
0.495
0.504
0.496
0.103
0.092
0.103
0.288
0.297
0.310
0.491
0.504
0.494
0.095
0.095
0.097
0.299
0.302
0.300
0.499
0.504
0.501
0.095
0.097
0.089
0.295
0.305
0.302
0.496
0.510
0.503

Model
CL
(x->y)
0.101
0.296
0.499
0.108
0.305
0.503
0.107
0.308
0.511
0.100
0.293
0.500
0.110
0.308
0.505
0.101
0.293
0.497
0.102
0.303
0.500
0.105
0.298
0.498
0.103
0.285
0.493
0.108
0.307
0.499
0.114
0.294
0.493

Table	
  13	
  (cont’d)	
  
	
  	
  
Parameter Estimates (Y)
Se
Se (CL
P
P (CL
(AR Y)
x->y)
(AR Y) x->y)
0.018
0.019
0.000
0.000
0.017
0.019
0.000
0.000
0.016
0.017
0.000
0.000
0.016
0.018
0.000
0.000
0.016
0.017
0.000
0.000
0.015
0.016
0.000
0.000
0.026
0.028
0.005
0.004
0.026
0.027
0.008
0.000
0.023
0.025
0.003
0.000
0.025
0.027
0.000
0.012
0.025
0.026
0.000
0.000
0.023
0.024
0.000
0.000
0.023
0.025
0.000
0.000
0.023
0.024
0.000
0.000
0.022
0.022
0.000
0.000
0.010
0.009
0.000
0.000
0.009
0.008
0.000
0.000
0.008
0.008
0.000
0.000
0.009
0.008
0.000
0.000
0.009
0.008
0.000
0.000
0.008
0.007
0.000
0.000
0.009
0.008
0.000
0.000
0.008
0.007
0.000
0.000
0.007
0.007
0.000
0.000
0.024
0.022
0.012
0.001
0.022
0.020
0.001
0.000
0.020
0.019
0.000
0.000
0.022
0.021
0.000
0.000
0.020
0.019
0.000
0.000
0.019
0.018
0.000
0.000
0.021
0.019
0.000
0.000
0.019
0.018
0.000
0.000
0.017
0.017
0.000
0.000

143	
  

AR
(X)
0.296
0.305
0.298
0.502
0.500
0.497
0.095
0.104
0.106
0.295
0.308
0.299
0.500
0.501
0.495
0.099
0.100
0.098
0.300
0.295
0.300
0.501
0.498
0.502
0.096
0.104
0.092
0.297
0.288
0.301
0.506
0.498
0.501

Model Parameter Estimates (X)
CL
Se
Se (CL
P
(y->x) (AR X)
y->x)
(AR X)
0.089
0.020
0.018
0.000
0.288
0.018
0.017
0.000
0.501
0.017
0.016
0.000
0.105
0.018
0.016
0.000
0.301
0.017
0.015
0.000
0.501
0.016
0.015
0.000
0.097
0.028
0.026
0.040
0.301
0.027
0.025
0.002
0.488
0.025
0.023
0.000
0.086
0.027
0.025
0.000
0.281
0.026
0.024
0.000
0.500
0.024
0.023
0.000
0.107
0.025
0.023
0.000
0.301
0.024
0.022
0.000
0.502
0.023
0.022
0.000
0.103
0.009
0.010
0.000
0.296
0.009
0.010
0.000
0.502
0.008
0.008
0.000
0.097
0.008
0.009
0.000
0.300
0.008
0.009
0.000
0.498
0.007
0.008
0.000
0.105
0.008
0.009
0.000
0.299
0.007
0.008
0.000
0.496
0.007
0.007
0.000
0.102
0.021
0.023
0.001
0.290
0.021
0.023
0.000
0.506
0.019
0.020
0.001
0.093
0.020
0.022
0.000
0.300
0.019
0.021
0.000
0.493
0.018
0.019
0.000
0.108
0.019
0.020
0.000
0.298
0.018
0.019
0.000
0.493
0.017
0.017
0.000

P (CL
y->x)
0.000
0.000
0.000
0.000
0.000
0.000
0.007
0.000
0.000
0.036
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.006
0.000
0.000
0.001
0.000
0.000
0.000
0.000
0.000

	
  
Table	
  13	
  (cont’d)	
  

	
  
Data Simulation Parameters
Std
Auto Cross
T
N
error
reg
-Lag
10
100
0.5
0.1
0.1
10
100
0.5
0.1
0.3
10
100
0.5
0.1
0.5
10
100
0.5
0.3
0.1
10
100
0.5
0.3
0.3
10
100
0.5
0.3
0.5
10
100
0.5
0.5
0.1
10
100
0.5
0.5
0.3
10
100
0.5
0.5
0.5
10
150
0.1
0.1
0.1
10
150
0.1
0.1
0.3
10
150
0.1
0.1
0.5
10
150
0.1
0.3
0.1
10
150
0.1
0.3
0.3
10
150
0.1
0.3
0.5
10
150
0.1
0.5
0.1
10
150
0.1
0.5
0.3
10
150
0.1
0.5
0.5
10
150
0.3
0.1
0.1
10
150
0.3
0.1
0.3
10
150
0.3
0.1
0.5
10
150
0.3
0.3
0.1
10
150
0.3
0.3
0.3
10
150
0.3
0.3
0.5
10
150
0.3
0.5
0.1
10
150
0.3
0.5
0.3
10
150
0.3
0.5
0.5
10
150
0.5
0.1
0.1
10
150
0.5
0.1
0.3
10
150
0.5
0.1
0.5
10
150
0.5
0.3
0.1
10
150
0.5
0.3
0.3
10
150
0.5
0.3
0.5

	
  

AR
(Y)
0.098
0.102
0.085
0.293
0.305
0.305
0.494
0.512
0.505
0.101
0.100
0.101
0.301
0.301
0.298
0.501
0.499
0.501
0.105
0.102
0.103
0.304
0.301
0.298
0.504
0.498
0.504
0.108
0.102
0.105
0.307
0.300
0.300

Model Parameter Estimates (Y)
CL
Se
Se (CL
P
(x->y) (AR Y)
x->y)
(AR Y)
0.104
0.029
0.028
0.019
0.283
0.028
0.027
0.006
0.492
0.025
0.024
0.013
0.112
0.028
0.027
0.000
0.308
0.026
0.025
0.000
0.497
0.024
0.023
0.000
0.118
0.026
0.024
0.000
0.292
0.024
0.023
0.000
0.491
0.022
0.022
0.000
0.098
0.008
0.007
0.000
0.301
0.008
0.007
0.000
0.500
0.007
0.007
0.000
0.101
0.007
0.007
0.000
0.300
0.007
0.007
0.000
0.499
0.006
0.006
0.000
0.096
0.007
0.007
0.000
0.304
0.007
0.006
0.000
0.499
0.006
0.006
0.000
0.095
0.019
0.018
0.000
0.301
0.018
0.017
0.000
0.498
0.016
0.016
0.000
0.103
0.018
0.017
0.000
0.298
0.017
0.017
0.000
0.499
0.015
0.015
0.000
0.095
0.016
0.016
0.000
0.308
0.016
0.015
0.000
0.496
0.014
0.014
0.000
0.094
0.023
0.023
0.001
0.299
0.022
0.022
0.004
0.497
0.021
0.020
0.000
0.103
0.022
0.022
0.000
0.296
0.022
0.021
0.000
0.499
0.020
0.019
0.000

144	
  

P (CL
x->y)
0.012
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.005
0.000
0.000
0.000
0.000
0.000

AR
(X)
0.094
0.107
0.088
0.295
0.286
0.302
0.510
0.500
0.499
0.102
0.099
0.102
0.304
0.304
0.301
0.504
0.503
0.498
0.099
0.101
0.103
0.309
0.307
0.302
0.508
0.507
0.495
0.096
0.103
0.102
0.310
0.306
0.302

Model Parameter Estimates (X)
CL
Se
Se (CL
P
(y->x) (AR X)
y->x)
(AR X)
0.100
0.027
0.028
0.023
0.288
0.027
0.028
0.001
0.508
0.024
0.025
0.019
0.092
0.026
0.028
0.000
0.301
0.025
0.026
0.000
0.491
0.023
0.024
0.000
0.107
0.024
0.025
0.000
0.298
0.023
0.024
0.000
0.493
0.022
0.022
0.000
0.103
0.007
0.008
0.000
0.296
0.007
0.008
0.000
0.499
0.007
0.007
0.000
0.100
0.007
0.007
0.000
0.294
0.007
0.007
0.000
0.500
0.006
0.006
0.000
0.100
0.007
0.007
0.000
0.299
0.006
0.006
0.000
0.502
0.006
0.006
0.000
0.106
0.018
0.019
0.000
0.296
0.017
0.018
0.000
0.498
0.016
0.016
0.000
0.103
0.017
0.018
0.000
0.287
0.017
0.017
0.000
0.502
0.015
0.015
0.000
0.101
0.016
0.016
0.000
0.298
0.015
0.015
0.000
0.504
0.014
0.014
0.000
0.106
0.023
0.023
0.004
0.298
0.022
0.023
0.000
0.497
0.020
0.021
0.000
0.106
0.022
0.022
0.000
0.287
0.021
0.021
0.000
0.502
0.019
0.020
0.000

P (CL
y->x)
0.031
0.000
0.000
0.008
0.000
0.000
0.005
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.001
0.000
0.000
0.000
0.000
0.000

	
  
Table	
  13	
  (cont’d)	
  

	
  
Data Simulation Parameters
Std
Auto Cross
T
N
error
reg
-Lag
10
150
0.5
0.5
0.1
10
150
0.5
0.5
0.3
10
150
0.5
0.5
0.5
10
200
0.1
0.1
0.1
10
200
0.1
0.1
0.3
10
200
0.1
0.1
0.5
10
200
0.1
0.3
0.1
10
200
0.1
0.3
0.3
10
200
0.1
0.3
0.5
10
200
0.1
0.5
0.1
10
200
0.1
0.5
0.3
10
200
0.1
0.5
0.5
10
200
0.3
0.1
0.1
10
200
0.3
0.1
0.3
10
200
0.3
0.1
0.5
10
200
0.3
0.3
0.1
10
200
0.3
0.3
0.3
10
200
0.3
0.3
0.5
10
200
0.3
0.5
0.1
10
200
0.3
0.5
0.3
10
200
0.3
0.5
0.5
10
200
0.5
0.1
0.1
10
200
0.5
0.1
0.3
10
200
0.5
0.1
0.5
10
200
0.5
0.3
0.1
10
200
0.5
0.3
0.3
10
200
0.5
0.3
0.5
10
200
0.5
0.5
0.1
10
200
0.5
0.5
0.3
10
200
0.5
0.5
0.5
15
100
0.1
0.1
0.1
15
100
0.1
0.1
0.3
15
100
0.1
0.1
0.5

	
  

AR
(Y)
0.505
0.497
0.504
0.098
0.100
0.100
0.303
0.295
0.299
0.503
0.499
0.500
0.093
0.099
0.099
0.305
0.291
0.299
0.504
0.497
0.501
0.090
0.098
0.098
0.305
0.292
0.300
0.503
0.496
0.501
0.098
0.103
0.097

Model Parameter Estimates (Y)
CL
Se
Se (CL
P
(x->y)
(AR Y)
x->y)
(AR Y)
0.097
0.021
0.020
0.000
0.308
0.020
0.020
0.000
0.494
0.018
0.018
0.000
0.104
0.007
0.007
0.000
0.300
0.007
0.007
0.000
0.502
0.006
0.006
0.000
0.101
0.006
0.007
0.000
0.301
0.006
0.007
0.000
0.502
0.006
0.006
0.000
0.099
0.006
0.006
0.000
0.305
0.006
0.006
0.000
0.500
0.005
0.005
0.000
0.108
0.016
0.017
0.000
0.297
0.015
0.016
0.000
0.502
0.014
0.015
0.000
0.103
0.015
0.016
0.000
0.306
0.015
0.016
0.000
0.505
0.013
0.014
0.000
0.100
0.014
0.015
0.000
0.311
0.013
0.014
0.000
0.500
0.012
0.013
0.000
0.110
0.020
0.021
0.011
0.293
0.019
0.020
0.000
0.502
0.018
0.018
0.000
0.105
0.019
0.020
0.000
0.308
0.019
0.019
0.000
0.505
0.017
0.017
0.000
0.101
0.017
0.018
0.000
0.313
0.017
0.017
0.000
0.500
0.016
0.016
0.000
0.100
0.010
0.009
0.000
0.298
0.009
0.008
0.000
0.499
0.008
0.008
0.000

145	
  

P (CL
x->y)
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000

AR
(X)
0.508
0.508
0.492
0.101
0.098
0.100
0.298
0.300
0.301
0.500
0.503
0.499
0.102
0.095
0.100
0.296
0.298
0.304
0.498
0.505
0.496
0.102
0.094
0.099
0.296
0.297
0.306
0.496
0.505
0.493
0.103
0.101
0.101

Model Parameter Estimates (X)
CL
Se
Se (CL
P
(y->x) (AR X)
y->x)
(AR X)
0.102
0.020
0.021
0.000
0.299
0.019
0.019
0.000
0.504
0.018
0.018
0.000
0.102
0.007
0.007
0.000
0.298
0.007
0.006
0.000
0.502
0.006
0.006
0.000
0.101
0.007
0.006
0.000
0.300
0.007
0.006
0.000
0.502
0.006
0.006
0.000
0.098
0.006
0.006
0.000
0.300
0.006
0.006
0.000
0.501
0.005
0.005
0.000
0.105
0.017
0.016
0.000
0.292
0.016
0.015
0.000
0.505
0.015
0.014
0.000
0.104
0.016
0.015
0.000
0.299
0.016
0.015
0.000
0.506
0.014
0.013
0.000
0.095
0.015
0.014
0.000
0.298
0.014
0.014
0.000
0.503
0.012
0.012
0.000
0.107
0.021
0.020
0.000
0.289
0.020
0.019
0.000
0.504
0.018
0.018
0.000
0.105
0.020
0.019
0.000
0.299
0.019
0.019
0.000
0.509
0.017
0.017
0.000
0.094
0.018
0.018
0.000
0.296
0.018
0.017
0.000
0.504
0.016
0.016
0.000
0.097
0.008
0.009
0.000
0.308
0.008
0.009
0.000
0.503
0.008
0.008
0.000

P (CL
y->x)
0.001
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000

	
  
Table	
  13	
  (cont’d)	
  

	
  
Data Simulation Parameters
Std
Auto Cross
T
N
error
reg
-Lag
15
100
0.1
0.3
0.1
15
100
0.1
0.3
0.3
15
100
0.1
0.3
0.5
15
100
0.1
0.5
0.1
15
100
0.1
0.5
0.3
15
100
0.1
0.5
0.5
15
100
0.3
0.1
0.1
15
100
0.3
0.1
0.3
15
100
0.3
0.1
0.5
15
100
0.3
0.3
0.1
15
100
0.3
0.3
0.3
15
100
0.3
0.3
0.5
15
100
0.3
0.5
0.1
15
100
0.3
0.5
0.3
15
100
0.3
0.5
0.5
15
100
0.5
0.1
0.1
15
100
0.5
0.1
0.3
15
100
0.5
0.1
0.5
15
100
0.5
0.3
0.1
15
100
0.5
0.3
0.3
15
100
0.5
0.3
0.5
15
100
0.5
0.5
0.1
15
100
0.5
0.5
0.3
15
100
0.5
0.5
0.5
15
150
0.1
0.1
0.1
15
150
0.1
0.1
0.3
15
150
0.1
0.1
0.5
15
150
0.1
0.3
0.1
15
150
0.1
0.3
0.3
15
150
0.1
0.3
0.5
15
150
0.1
0.5
0.1
15
150
0.1
0.5
0.3
15
150
0.1
0.5
0.5

	
  

AR
(Y)
0.302
0.297
0.305
0.502
0.504
0.499
0.100
0.102
0.094
0.308
0.297
0.311
0.510
0.507
0.499
0.103
0.099
0.093
0.310
0.299
0.313
0.514
0.507
0.500
0.102
0.103
0.101
0.299
0.300
0.303
0.502
0.502
0.500

Model Parameter Estimates (Y)
CL
Se
Se (CL
P
(x->y) (AR Y)
x->y)
(AR Y)
0.098
0.009
0.008
0.000
0.302
0.009
0.008
0.000
0.496
0.008
0.007
0.000
0.102
0.008
0.007
0.000
0.299
0.008
0.007
0.000
0.502
0.007
0.007
0.000
0.101
0.020
0.019
0.001
0.297
0.019
0.018
0.000
0.502
0.018
0.017
0.000
0.098
0.020
0.018
0.000
0.307
0.019
0.018
0.000
0.489
0.017
0.016
0.000
0.102
0.018
0.017
0.000
0.299
0.017
0.016
0.000
0.501
0.015
0.015
0.000
0.102
0.024
0.024
0.003
0.297
0.023
0.022
0.001
0.505
0.021
0.021
0.000
0.101
0.023
0.022
0.000
0.310
0.022
0.021
0.000
0.487
0.020
0.020
0.000
0.101
0.021
0.020
0.000
0.300
0.020
0.020
0.000
0.500
0.018
0.018
0.000
0.098
0.008
0.007
0.000
0.300
0.008
0.007
0.000
0.502
0.007
0.007
0.000
0.100
0.007
0.007
0.000
0.304
0.007
0.007
0.000
0.497
0.006
0.006
0.000
0.096
0.007
0.006
0.000
0.297
0.006
0.006
0.000
0.499
0.006
0.006
0.000

146	
  

P (CL
x->y)
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000

AR
(X)
0.304
0.302
0.298
0.502
0.501
0.500
0.106
0.100
0.100
0.309
0.305
0.298
0.503
0.502
0.500
0.105
0.099
0.098
0.311
0.306
0.299
0.503
0.502
0.499
0.099
0.101
0.101
0.300
0.303
0.302
0.501
0.504
0.499

Model Parameter Estimates (X)
CL
Se
Se (CL
P
(y->x) (AR X)
y->x)
(AR X)
0.093
0.008
0.009
0.000
0.296
0.008
0.009
0.000
0.496
0.007
0.008
0.000
0.106
0.007
0.008
0.000
0.303
0.007
0.008
0.000
0.501
0.007
0.007
0.000
0.091
0.019
0.020
0.000
0.314
0.018
0.020
0.000
0.505
0.017
0.018
0.000
0.089
0.019
0.020
0.000
0.294
0.018
0.019
0.000
0.494
0.016
0.017
0.000
0.112
0.016
0.018
0.000
0.304
0.016
0.016
0.000
0.503
0.015
0.015
0.000
0.089
0.023
0.024
0.000
0.314
0.022
0.023
0.000
0.505
0.020
0.021
0.000
0.091
0.022
0.023
0.000
0.294
0.022
0.022
0.000
0.494
0.020
0.020
0.000
0.112
0.020
0.021
0.000
0.303
0.019
0.020
0.000
0.504
0.018
0.018
0.000
0.104
0.007
0.008
0.000
0.300
0.007
0.008
0.000
0.499
0.006
0.007
0.000
0.101
0.007
0.007
0.000
0.296
0.007
0.007
0.000
0.499
0.006
0.006
0.000
0.099
0.006
0.007
0.000
0.296
0.006
0.006
0.000
0.501
0.006
0.006
0.000

P (CL
y->x)
0.000
0.000
0.000
0.000
0.000
0.000
0.001
0.000
0.000
0.001
0.000
0.000
0.000
0.000
0.000
0.009
0.000
0.000
0.003
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000

	
  
Table	
  13	
  (cont’d)	
  

	
  
Data	
  Simulation	
  Parameters	
  

T
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15

	
  

N
150
150
150
150
150
150
150
150
150
150
150
150
150
150
150
150
150
150
200
200
200
200
200
200
200
200
200
200
200
200
200
200
200

Std
error
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.3
0.3
0.3
0.3
0.3
0.3

Auto
reg
0.1
0.1
0.1
0.3
0.3
0.3
0.5
0.5
0.5
0.1
0.1
0.1
0.3
0.3
0.3
0.5
0.5
0.5
0.1
0.1
0.1
0.3
0.3
0.3
0.5
0.5
0.5
0.1
0.1
0.1
0.3
0.3
0.3

Cross
-Lag
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5

Model	
  Parameter	
  Estimates	
  (Y)	
  

AR
(Y)
0.110
0.106
0.104
0.302
0.301
0.305
0.505
0.502
0.502
0.114
0.106
0.106
0.304
0.301
0.304
0.505
0.502
0.504
0.099
0.102
0.098
0.300
0.302
0.301
0.501
0.499
0.502
0.097
0.100
0.096
0.303
0.303
0.301

CL
(x->y)
0.097
0.297
0.502
0.100
0.306
0.494
0.092
0.292
0.498
0.096
0.295
0.502
0.101
0.305
0.493
0.092
0.291
0.497
0.103
0.306
0.502
0.102
0.300
0.503
0.099
0.303
0.500
0.102
0.310
0.505
0.106
0.300
0.505

Se
(AR Y)
0.016
0.016
0.015
0.016
0.015
0.014
0.015
0.014
0.012
0.020
0.019
0.017
0.019
0.018
0.016
0.017
0.016
0.015
0.006
0.006
0.006
0.006
0.006
0.005
0.006
0.005
0.005
0.014
0.013
0.012
0.014
0.013
0.012

Se (CL
x->y)
0.016
0.016
0.014
0.015
0.015
0.014
0.014
0.013
0.012
0.019
0.019
0.017
0.018
0.018
0.016
0.017
0.016
0.015
0.007
0.007
0.006
0.007
0.007
0.006
0.006
0.006
0.005
0.015
0.014
0.013
0.014
0.014
0.012

P
(AR Y)
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000

147	
  

Model	
  Parameter	
  Estimates	
  (X)	
  

P (CL
x->y)
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000

AR
(X)
0.098
0.098
0.100
0.297
0.309
0.304
0.504
0.509
0.496
0.098
0.096
0.099
0.294
0.310
0.304
0.505
0.510
0.494
0.098
0.101
0.099
0.297
0.297
0.300
0.502
0.501
0.500
0.094
0.102
0.097
0.298
0.296
0.300

CL
(y->x)
0.105
0.303
0.499
0.104
0.295
0.497
0.097
0.291
0.504
0.104
0.305
0.500
0.105
0.297
0.497
0.096
0.290
0.506
0.100
0.299
0.498
0.104
0.299
0.501
0.099
0.298
0.502
0.099
0.297
0.495
0.113
0.295
0.502

Se
(AR X)
0.016
0.016
0.014
0.016
0.015
0.013
0.014
0.013
0.012
0.019
0.019
0.017
0.019
0.018
0.016
0.017
0.016
0.015
0.007
0.007
0.006
0.007
0.007
0.006
0.006
0.006
0.005
0.015
0.014
0.013
0.014
0.014
0.012

Se (CL
y->x)
0.016
0.016
0.014
0.016
0.015
0.014
0.014
0.014
0.012
0.019
0.019
0.017
0.019
0.018
0.016
0.017
0.016
0.015
0.006
0.006
0.006
0.006
0.006
0.005
0.006
0.005
0.005
0.014
0.013
0.012
0.013
0.013
0.012

P
(AR X)
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000

P (CL
y->x)
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000

	
  
Table	
  13	
  (cont’d)	
  

	
  
Data Simulation Parameters
T
15
15
15
15
15
15
15
15
15
15
15
15

N
200
200
200
200
200
200
200
200
200
200
200
200

Std
error
0.3
0.3
0.3
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5

Auto
reg
0.5
0.5
0.5
0.1
0.1
0.1
0.3
0.3
0.3
0.5
0.5
0.5

Cross
-Lag
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5

AR
(Y)
0.503
0.499
0.504
0.096
0.097
0.096
0.305
0.302
0.300
0.504
0.499
0.506

Model Parameter Estimates (Y)
Se
Se
CL
(AR
(CL
P
(x->y)
Y)
x->y) (AR Y)
0.100
0.012
0.013
0.000
0.303
0.012
0.012
0.000
0.499
0.011
0.011
0.000
0.099
0.017
0.017
0.000
0.309
0.016
0.017
0.000
0.505
0.015
0.015
0.000
0.107
0.016
0.017
0.000
0.300
0.016
0.016
0.000
0.506
0.014
0.014
0.000
0.101
0.015
0.015
0.000
0.301
0.014
0.014
0.000
0.499
0.013
0.013
0.000

	
  

	
  

148	
  

Model Parameter Estimates (X)
P (CL
x->y)
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000

AR
(X)
0.502
0.499
0.501
0.093
0.102
0.097
0.300
0.297
0.299
0.500
0.498
0.502

CL
(y->x)
0.099
0.299
0.503
0.098
0.296
0.494
0.116
0.293
0.503
0.100
0.300
0.502

Se
(AR X)
0.013
0.012
0.011
0.017
0.017
0.015
0.017
0.016
0.014
0.015
0.014
0.013

Se (CL
y->x)
0.012
0.012
0.011
0.017
0.016
0.015
0.016
0.016
0.014
0.015
0.014
0.013

P
(AR X)
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000

P (CL
y->x)
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000

	
  

Y
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf
Perf

	
  

Model
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36

Trial Δ
4
4
4
4
4
4
5
5
5
5
5
5
6
6
6
6
6
6
7
7
7
7
7
7
8
8
8
8
8
8
9
9
9
9
9
9

Table	
  14	
  
Partially	
  Cross	
  Partially	
  Time	
  Varying	
  Longitudinal	
  Cross-­‐Lag	
  Simulation	
  With	
  Seven	
  Variables	
  
Both Processes
Adaptation Process
Performance Process
N
Error AR perf
Se
p
CL meta
Se
P
CL outef
Se
P
CL meta Se
P
CL outef
150
0.1
0.498 0.01 0.00
0.399 0.01 0.00
0.199 0.01 0.00
-0.196 0.01 0.00
0.405
150
0.3
0.495 0.01 0.00
0.398 0.02 0.00
0.197 0.02 0.00
-0.195 0.01 0.00
0.411
150
0.5
0.495 0.01 0.00
0.398 0.03 0.00
0.197 0.03 0.00
-0.196 0.02 0.00
0.413
200
0.1
0.500 0.00 0.00
0.402 0.01 0.00
0.196 0.01 0.00
-0.199 0.01 0.00
0.401
200
0.3
0.500 0.01 0.00
0.406 0.02 0.00
0.190 0.02 0.00
-0.200 0.01 0.00
0.400
200
0.5
0.501 0.01 0.00
0.410 0.03 0.00
0.187 0.03 0.00
-0.201 0.01 0.00
0.400
150
0.1
0.498 0.01 0.00
0.401 0.01 0.00
0.199 0.01 0.00
-0.196 0.01 0.00
0.405
150
0.3
0.495 0.01 0.00
0.403 0.02 0.00
0.198 0.02 0.00
-0.195 0.01 0.00
0.411
150
0.5
0.495 0.01 0.00
0.405 0.03 0.00
0.198 0.03 0.00
-0.196 0.02 0.00
0.412
200
0.1
0.499 0.00 0.00
0.403 0.01 0.00
0.197 0.01 0.00
-0.201 0.01 0.00
0.401
200
0.3
0.500 0.01 0.00
0.407 0.01 0.00
0.191 0.02 0.00
-0.203 0.01 0.00
0.401
200
0.5
0.501 0.01 0.00
0.409 0.02 0.00
0.188 0.02 0.00
-0.204 0.01 0.00
0.400
150
0.1
0.499 0.01 0.00
0.401 0.01 0.00
0.199 0.01 0.00
-0.194 0.01 0.00
0.404
150
0.3
0.497 0.01 0.00
0.404 0.02 0.00
0.199 0.02 0.00
-0.195 0.02 0.00
0.409
150
0.5
0.497 0.01 0.00
0.407 0.02 0.00
0.200 0.02 0.00
-0.196 0.02 0.00
0.411
200
0.1
0.499 0.00 0.00
0.402 0.00 0.00
0.197 0.01 0.00
-0.203 0.01 0.00
0.400
200
0.3
0.499 0.01 0.00
0.406 0.01 0.00
0.193 0.02 0.00
-0.206 0.01 0.00
0.400
200
0.5
0.500 0.01 0.00
0.407 0.02 0.00
0.192 0.02 0.00
-0.206 0.01 0.00
0.400
150
0.1
0.499 0.01 0.00
0.402 0.01 0.00
0.199 0.01 0.00
-0.195 0.01 0.00
0.404
150
0.3
0.499 0.01 0.00
0.405 0.01 0.00
0.200 0.02 0.00
-0.197 0.02 0.00
0.407
150
0.5
0.499 0.01 0.00
0.408 0.02 0.00
0.203 0.02 0.00
-0.198 0.02 0.00
0.408
200
0.1
0.500 0.00 0.00
0.402 0.00 0.00
0.198 0.01 0.00
-0.204 0.01 0.00
0.400
200
0.3
0.500 0.01 0.00
0.404 0.01 0.00
0.195 0.01 0.00
-0.206 0.01 0.00
0.399
200
0.5
0.501 0.01 0.00
0.405 0.02 0.00
0.194 0.02 0.00
-0.206 0.02 0.00
0.398
150
0.1
0.499 0.01 0.00
0.402 0.01 0.00
0.199 0.01 0.00
-0.198 0.02 0.00
0.401
150
0.3
0.499 0.01 0.00
0.407 0.01 0.00
0.201 0.01 0.00
-0.200 0.02 0.00
0.406
150
0.5
0.499 0.01 0.00
0.410 0.02 0.00
0.205 0.02 0.00
-0.202 0.02 0.00
0.408
200
0.1
0.499 0.00 0.00
0.402 0.00 0.00
0.197 0.01 0.00
-0.202 0.01 0.00
0.404
200
0.3
0.499 0.01 0.00
0.403 0.01 0.00
0.193 0.01 0.00
-0.204 0.02 0.00
0.403
200
0.5
0.499 0.01 0.00
0.402 0.02 0.00
0.192 0.02 0.00
-0.203 0.02 0.00
0.402
150
0.1
0.499 0.01 0.00
0.403 0.01 0.00
0.199 0.01 0.00
-0.198 0.02 0.00
0.403
150
0.3
0.498 0.01 0.00
0.408 0.01 0.00
0.200 0.01 0.00
-0.202 0.02 0.00
0.411
150
0.5
0.498 0.01 0.00
0.411 0.02 0.00
0.202 0.02 0.00
-0.204 0.02 0.00
0.413
200
0.1
0.499 0.00 0.00
0.402 0.00 0.00
0.197 0.01 0.00
-0.206 0.02 0.00
0.403
200
0.3
0.498 0.01 0.00
0.404 0.01 0.00
0.193 0.01 0.00
-0.206 0.02 0.00
0.404
200
0.5
0.497 0.01 0.00
0.403 0.01 0.00
0.191 0.02 0.00
-0.206 0.02 0.00
0.404

149	
  

Se
0.01
0.01
0.02
0.01
0.01
0.01
0.01
0.02
0.02
0.01
0.01
0.01
0.01
0.02
0.02
0.01
0.01
0.01
0.01
0.02
0.02
0.01
0.01
0.02
0.01
0.02
0.02
0.01
0.02
0.02
0.01
0.02
0.02
0.01
0.02
0.02

P
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00

	
  

Y
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef
Lrnef

	
  

Model
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36

Trial Δ
4
4
4
4
4
4
5
5
5
5
5
5
6
6
6
6
6
6
7
7
7
7
7
7
8
8
8
8
8
8
9
9
9
9
9
9

N
150
150
150
200
200
200
150
150
150
200
200
200
150
150
150
200
200
200
150
150
150
200
200
200
150
150
150
200
200
200
150
150
150
200
200
200

Error
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5

Both Processes
AR lrnef Se
p
0.501 0.01 0.00
0.500 0.01 0.00
0.499 0.01 0.00
0.499 0.01 0.00
0.496 0.01 0.00
0.494 0.01 0.00
0.501 0.01 0.00
0.501 0.01 0.00
0.500 0.01 0.00
0.499 0.01 0.00
0.496 0.01 0.00
0.494 0.01 0.00
0.501 0.01 0.00
0.501 0.01 0.00
0.500 0.01 0.00
0.499 0.01 0.00
0.496 0.01 0.00
0.494 0.01 0.00
0.501 0.01 0.00
0.501 0.01 0.00
0.500 0.01 0.00
0.499 0.01 0.00
0.496 0.01 0.00
0.494 0.01 0.00
0.501 0.01 0.00
0.500 0.01 0.00
0.499 0.01 0.00
0.499 0.01 0.00
0.496 0.01 0.00
0.494 0.01 0.00
0.501 0.01 0.00
0.500 0.01 0.00
0.499 0.01 0.00
0.499 0.00 0.00
0.496 0.01 0.00
0.494 0.01 0.00

Table	
  14	
  (cont’d)	
  
	
  
Adaptation Process
CL perf Se
P
CL eval
-0.199 0.01 0.00
0.396
-0.196 0.02 0.00
0.392
-0.194 0.03 0.00
0.390
-0.201 0.01 0.00
0.400
-0.204 0.02 0.00
0.399
-0.207 0.03 0.00
0.398
-0.199 0.01 0.00
0.396
-0.199 0.02 0.00
0.390
-0.200 0.03 0.00
0.387
-0.202 0.01 0.00
0.399
-0.207 0.01 0.00
0.397
-0.211 0.02 0.00
0.394
-0.199 0.01 0.00
0.396
-0.199 0.02 0.00
0.390
-0.200 0.02 0.00
0.386
-0.202 0.00 0.00
0.399
-0.207 0.01 0.00
0.396
-0.212 0.02 0.00
0.392
-0.200 0.01 0.00
0.396
-0.200 0.02 0.00
0.390
-0.202 0.02 0.00
0.388
-0.202 0.00 0.00
0.399
-0.208 0.01 0.00
0.396
-0.213 0.02 0.00
0.392
-0.200 0.01 0.00
0.396
-0.202 0.01 0.00
0.392
-0.203 0.02 0.00
0.391
-0.202 0.00 0.00
0.399
-0.208 0.01 0.00
0.394
-0.212 0.02 0.00
0.390
-0.200 0.01 0.00
0.396
-0.202 0.01 0.00
0.392
-0.203 0.02 0.00
0.391
-0.202 0.00 0.00
0.399
-0.207 0.01 0.00
0.393
-0.210 0.02 0.00
0.389

150	
  

Se
0.01
0.02
0.04
0.01
0.02
0.03
0.01
0.02
0.03
0.01
0.02
0.02
0.01
0.02
0.03
0.01
0.02
0.02
0.01
0.02
0.03
0.01
0.02
0.02
0.01
0.02
0.02
0.01
0.01
0.02
0.01
0.02
0.02
0.01
0.01
0.02

P
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00

CL perf
-0.399
-0.402
-0.404
-0.403
-0.405
-0.406
-0.398
-0.400
-0.401
-0.402
-0.404
-0.405
-0.400
-0.401
-0.402
-0.403
-0.404
-0.405
-0.401
-0.402
-0.402
-0.404
-0.405
-0.405
-0.401
-0.402
-0.403
-0.400
-0.401
-0.401
-0.405
-0.404
-0.404
-0.399
-0.401
-0.401

Performance Process
Se
P
CL eval
0.01 0.00
0.197
0.01 0.00
0.198
0.02 0.00
0.199
0.01 0.00
0.195
0.01 0.00
0.193
0.01 0.00
0.194
0.01 0.00
0.201
0.02 0.00
0.201
0.02 0.00
0.201
0.01 0.00
0.197
0.01 0.00
0.196
0.01 0.00
0.196
0.01 0.00
0.201
0.02 0.00
0.202
0.02 0.00
0.202
0.01 0.00
0.197
0.01 0.00
0.197
0.01 0.00
0.197
0.01 0.00
0.201
0.02 0.00
0.202
0.02 0.00
0.202
0.01 0.00
0.195
0.01 0.00
0.197
0.02 0.00
0.197
0.01 0.00
0.200
0.02 0.00
0.199
0.02 0.00
0.199
0.01 0.00
0.202
0.02 0.00
0.201
0.02 0.00
0.201
0.02 0.00
0.195
0.02 0.00
0.197
0.02 0.00
0.198
0.01 0.00
0.205
0.02 0.00
0.204
0.02 0.00
0.204

Se
0.01
0.02
0.02
0.01
0.02
0.02
0.01
0.02
0.02
0.01
0.02
0.02
0.02
0.02
0.02
0.01
0.02
0.02
0.02
0.02
0.02
0.01
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02

P
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00

	
  

Y
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta
Meta

	
  

Model
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36

Trial Δ
4
4
4
4
4
4
5
5
5
5
5
5
6
6
6
6
6
6
7
7
7
7
7
7
8
8
8
8
8
8
9
9
9
9
9
9

N
150
150
150
200
200
200
150
150
150
200
200
200
150
150
150
200
200
200
150
150
150
200
200
200
150
150
150
200
200
200
150
150
150
200
200
200

Error
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5

Both Processes
AR meta Se
p
0.500 0.01 0.00
0.497 0.01 0.00
0.496 0.02 0.00
0.500 0.01 0.00
0.498 0.01 0.00
0.497 0.01 0.00
0.500 0.01 0.00
0.498 0.01 0.00
0.497 0.01 0.00
0.500 0.00 0.00
0.499 0.01 0.00
0.497 0.01 0.00
0.500 0.01 0.00
0.498 0.01 0.00
0.497 0.01 0.00
0.500 0.00 0.00
0.499 0.01 0.00
0.497 0.01 0.00
0.499 0.01 0.00
0.498 0.01 0.00
0.497 0.01 0.00
0.500 0.00 0.00
0.499 0.01 0.00
0.497 0.01 0.00
0.499 0.01 0.00
0.498 0.01 0.00
0.498 0.01 0.00
0.500 0.00 0.00
0.499 0.01 0.00
0.497 0.01 0.00
0.499 0.01 0.00
0.499 0.01 0.00
0.499 0.01 0.00
0.500 0.00 0.00
0.498 0.01 0.00
0.497 0.01 0.00

Table	
  14	
  (cont’d)	
  
	
  
Adaptation Process
CL lrnef Se
P
CL eval
0.400 0.01 0.00
0.400
0.399 0.02 0.00
0.402
0.397 0.03 0.00
0.404
0.398 0.01 0.00
0.397
0.396 0.02 0.00
0.391
0.394 0.03 0.00
0.387
0.400 0.01 0.00
0.400
0.400 0.02 0.00
0.401
0.399 0.03 0.00
0.402
0.399 0.01 0.00
0.398
0.398 0.02 0.00
0.394
0.397 0.03 0.00
0.393
0.400 0.01 0.00
0.400
0.401 0.02 0.00
0.402
0.400 0.02 0.00
0.404
0.399 0.01 0.00
0.397
0.398 0.02 0.00
0.393
0.397 0.02 0.00
0.391
0.400 0.01 0.00
0.400
0.400 0.02 0.00
0.402
0.400 0.02 0.00
0.404
0.399 0.01 0.00
0.397
0.398 0.02 0.00
0.391
0.399 0.02 0.00
0.387
0.400 0.01 0.00
0.401
0.399 0.02 0.00
0.403
0.398 0.02 0.00
0.404
0.399 0.01 0.00
0.397
0.399 0.01 0.00
0.392
0.399 0.02 0.00
0.388
0.400 0.01 0.00
0.400
0.399 0.02 0.00
0.399
0.397 0.02 0.00
0.399
0.399 0.01 0.00
0.397
0.398 0.01 0.00
0.393
0.397 0.02 0.00
0.391

151	
  

Se
0.01
0.02
0.04
0.01
0.02
0.03
0.01
0.02
0.03
0.01
0.02
0.03
0.01
0.02
0.03
0.01
0.02
0.02
0.01
0.02
0.03
0.01
0.02
0.02
0.01
0.02
0.02
0.01
0.02
0.02
0.01
0.02
0.02
0.01
0.01
0.02

P
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00

CL lrnef
0.197
0.197
0.198
0.203
0.203
0.203
0.195
0.196
0.197
0.205
0.204
0.204
0.193
0.195
0.196
0.204
0.204
0.204
0.193
0.195
0.197
0.203
0.203
0.203
0.195
0.198
0.199
0.203
0.203
0.204
0.194
0.199
0.200
0.206
0.208
0.209

Performance Process
Se
P
CL eval
0.01 0.00
0.205
0.02 0.00
0.208
0.02 0.00
0.210
0.01 0.00
0.194
0.01 0.00
0.192
0.01 0.00
0.192
0.01 0.00
0.207
0.02 0.00
0.209
0.02 0.00
0.210
0.01 0.00
0.191
0.01 0.00
0.189
0.01 0.00
0.189
0.01 0.00
0.208
0.02 0.00
0.209
0.02 0.00
0.210
0.01 0.00
0.192
0.01 0.00
0.191
0.02 0.00
0.191
0.02 0.00
0.207
0.02 0.00
0.208
0.02 0.00
0.209
0.01 0.00
0.195
0.02 0.00
0.194
0.02 0.00
0.194
0.02 0.00
0.206
0.02 0.00
0.209
0.02 0.00
0.210
0.01 0.00
0.196
0.02 0.00
0.194
0.02 0.00
0.194
0.02 0.00
0.211
0.02 0.00
0.214
0.02 0.00
0.215
0.02 0.00
0.193
0.02 0.00
0.192
0.02 0.00
0.192

Se
0.02
0.02
0.02
0.01
0.02
0.02
0.02
0.02
0.02
0.01
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02

P
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00

	
  

Y
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval
Eval

	
  

Model
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36

Trial Δ
4
4
4
4
4
4
5
5
5
5
5
5
6
6
6
6
6
6
7
7
7
7
7
7
8
8
8
8
8
8
9
9
9
9
9
9

N
150
150
150
200
200
200
150
150
150
200
200
200
150
150
150
200
200
200
150
150
150
200
200
200
150
150
150
200
200
200
150
150
150
200
200
200

Error
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5

Both Processes
AR eval
Se
p
0.500 0.01 0.00
0.499 0.01 0.00
0.499 0.02 0.00
0.500 0.01 0.00
0.498 0.01 0.00
0.496 0.01 0.00
0.500 0.01 0.00
0.500 0.01 0.00
0.500 0.02 0.00
0.500 0.01 0.00
0.498 0.01 0.00
0.496 0.01 0.00
0.500 0.01 0.00
0.500 0.01 0.00
0.500 0.02 0.00
0.501 0.01 0.00
0.498 0.01 0.00
0.496 0.01 0.00
0.500 0.01 0.00
0.500 0.01 0.00
0.500 0.02 0.00
0.500 0.01 0.00
0.498 0.01 0.00
0.497 0.01 0.00
0.500 0.01 0.00
0.500 0.01 0.00
0.500 0.02 0.00
0.501 0.01 0.00
0.499 0.01 0.00
0.497 0.01 0.00
0.501 0.01 0.00
0.501 0.01 0.00
0.500 0.02 0.00
0.501 0.01 0.00
0.499 0.01 0.00
0.497 0.01 0.00

Table	
  14	
  (cont’d)	
  
	
  
Adaptation Process
Performance Process
CL perf
Se
P
CL perf Se
P
-0.199 0.01 0.00
-0.399 0.01 0.00
-0.194 0.02 0.00
-0.402 0.01 0.00
-0.188 0.03 0.00
-0.404 0.01 0.00
-0.198 0.01 0.00
-0.403 0.01 0.00
-0.195 0.02 0.00
-0.406 0.01 0.00
-0.193 0.03 0.00
-0.406 0.01 0.00
-0.198 0.01 0.00
-0.400 0.01 0.00
-0.194 0.02 0.00
-0.404 0.01 0.00
-0.189 0.03 0.00
-0.405 0.02 0.00
-0.199 0.01 0.00
-0.403 0.01 0.00
-0.197 0.02 0.00
-0.406 0.01 0.00
-0.195 0.02 0.00
-0.407 0.01 0.00
-0.197 0.01 0.00
-0.404 0.01 0.00
-0.191 0.02 0.00
-0.408 0.01 0.00
-0.186 0.02 0.00
-0.410 0.02 0.00
-0.200 0.00 0.00
-0.402 0.01 0.00
-0.199 0.01 0.00
-0.406 0.01 0.00
-0.197 0.02 0.00
-0.406 0.01 0.00
-0.198 0.01 0.00
-0.404 0.01 0.00
-0.194 0.02 0.00
-0.409 0.02 0.00
-0.191 0.02 0.00
-0.410 0.02 0.00
-0.200 0.00 0.00
-0.403 0.01 0.00
-0.198 0.01 0.00
-0.406 0.01 0.00
-0.196 0.02 0.00
-0.406 0.01 0.00
-0.198 0.01 0.00
-0.405 0.01 0.00
-0.196 0.01 0.00
-0.408 0.02 0.00
-0.194 0.02 0.00
-0.408 0.02 0.00
-0.200 0.00 0.00
-0.403 0.01 0.00
-0.200 0.01 0.00
-0.405 0.01 0.00
-0.200 0.02 0.00
-0.404 0.01 0.00
-0.199 0.01 0.00
-0.403 0.01 0.00
-0.197 0.01 0.00
-0.405 0.02 0.00
-0.195 0.02 0.00
-0.405 0.02 0.00
-0.200 0.00 0.00
-0.403 0.01 0.00
-0.201 0.01 0.00
-0.404 0.01 0.00
-0.200 0.02 0.00
-0.405 0.02 0.00

152	
  

	
  

Y
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals
Goals

	
  

Model
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36

Trial Δ
4
4
4
4
4
4
5
5
5
5
5
5
6
6
6
6
6
6
7
7
7
7
7
7
8
8
8
8
8
8
9
9
9
9
9
9

N
150
150
150
200
200
200
150
150
150
200
200
200
150
150
150
200
200
200
150
150
150
200
200
200
150
150
150
200
200
200
150
150
150
200
200
200

Error
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5

Both Processes
AR goals Se
p
0.498 0.01 0.00
0.495 0.01 0.00
0.495 0.01 0.00
0.503 0.00 0.00
0.506 0.01 0.00
0.505 0.01 0.00
0.498 0.01 0.00
0.496 0.01 0.00
0.495 0.01 0.00
0.503 0.00 0.00
0.506 0.01 0.00
0.506 0.01 0.00
0.499 0.01 0.00
0.498 0.01 0.00
0.498 0.01 0.00
0.503 0.00 0.00
0.505 0.01 0.00
0.505 0.01 0.00
0.499 0.01 0.00
0.498 0.01 0.00
0.497 0.01 0.00
0.503 0.00 0.00
0.506 0.01 0.00
0.507 0.01 0.00
0.500 0.01 0.00
0.498 0.01 0.00
0.497 0.01 0.00
0.503 0.00 0.00
0.506 0.01 0.00
0.506 0.01 0.00
0.500 0.01 0.00
0.498 0.01 0.00
0.496 0.01 0.00
0.503 0.00 0.00
0.507 0.01 0.00
0.508 0.01 0.00

Table	
  14	
  (cont’d)	
  
	
  
Adaptation
CL perf Se
P
0.201 0.01 0.00
0.201 0.02 0.00
0.200 0.03 0.00
0.200 0.01 0.00
0.200 0.02 0.00
0.199 0.03 0.00
0.201 0.01 0.00
0.202 0.02 0.00
0.201 0.03 0.00
0.200 0.01 0.00
0.200 0.02 0.00
0.199 0.02 0.00
0.201 0.01 0.00
0.203 0.02 0.00
0.202 0.02 0.00
0.200 0.01 0.00
0.197 0.01 0.00
0.195 0.02 0.00
0.200 0.01 0.00
0.200 0.02 0.00
0.198 0.02 0.00
0.200 0.00 0.00
0.199 0.01 0.00
0.197 0.02 0.00
0.200 0.01 0.00
0.202 0.01 0.00
0.202 0.02 0.00
0.200 0.00 0.00
0.200 0.01 0.00
0.199 0.02 0.00
0.200 0.01 0.00
0.198 0.01 0.00
0.196 0.02 0.00
0.200 0.00 0.00
0.200 0.01 0.00
0.199 0.02 0.00

153	
  

Process
CL se
0.402
0.405
0.409
0.400
0.402
0.406
0.401
0.405
0.409
0.400
0.401
0.402
0.401
0.404
0.408
0.400
0.400
0.401
0.401
0.406
0.410
0.399
0.397
0.396
0.401
0.406
0.411
0.399
0.398
0.397
0.402
0.407
0.410
0.399
0.396
0.395

Se
0.01
0.02
0.03
0.01
0.02
0.03
0.01
0.02
0.03
0.01
0.02
0.02
0.01
0.02
0.02
0.01
0.02
0.02
0.01
0.02
0.02
0.01
0.01
0.02
0.01
0.02
0.02
0.01
0.01
0.02
0.01
0.02
0.02
0.01
0.01
0.02

P
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00

CL perf
0.401
0.402
0.402
0.399
0.398
0.399
0.400
0.401
0.401
0.400
0.399
0.400
0.396
0.397
0.397
0.402
0.402
0.403
0.399
0.400
0.400
0.400
0.401
0.402
0.396
0.397
0.397
0.403
0.404
0.404
0.398
0.398
0.398
0.400
0.402
0.402

Performance Process
Se
P
CL se
0.01 0.00
-0.197
0.01 0.00
-0.192
0.02 0.00
-0.191
0.01 0.00
-0.202
0.01 0.00
-0.203
0.01 0.00
-0.203
0.01 0.00
-0.195
0.01 0.00
-0.191
0.02 0.00
-0.191
0.01 0.00
-0.203
0.01 0.00
-0.203
0.01 0.00
-0.202
0.01 0.00
-0.190
0.02 0.00 -0.189
0.02 0.00 -0.189
0.01 0.00
-0.204
0.01 0.00
-0.202
0.01 0.00
-0.201
0.01 0.00
-0.191
0.02 0.00
-0.191
0.02 0.00
-0.191
0.01 0.00
-0.198
0.01 0.00
-0.198
0.01 0.00
-0.198
0.01 0.00
-0.190
0.02 0.00
-0.192
0.02 0.00
-0.193
0.01 0.00
-0.202
0.01 0.00
-0.199
0.02 0.00
-0.199
0.01 0.00
-0.194
0.02 0.00
-0.193
0.02 0.00
-0.193
0.01 0.00
-0.196
0.02 0.00
-0.195
0.02 0.00
-0.195

Se
0.01
0.02
0.02
0.01
0.02
0.02
0.01
0.02
0.02
0.01
0.02
0.02
0.02
0.02
0.02
0.01
0.02
0.02
0.02
0.02
0.02
0.01
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02

P
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00

	
  
Table	
  14	
  (cont’d)	
  

Y
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef
Outef

	
  

Model
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36

Trial Δ
4
4
4
4
4
4
5
5
5
5
5
5
6
6
6
6
6
6
7
7
7
7
7
7
8
8
8
8
8
8
9
9
9
9
9
9

N
150
150
150
200
200
200
150
150
150
200
200
200
150
150
150
200
200
200
150
150
150
200
200
200
150
150
150
200
200
200
150
150
150
200
200
200

Error
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5

Both Processes
AR outef Se
p
0.499 0.01 0.00
0.499 0.01 0.00
0.499 0.01 0.00
0.502 0.01 0.00
0.504 0.01 0.00
0.504 0.01 0.00
0.499 0.01 0.00
0.499 0.01 0.00
0.499 0.01 0.00
0.502 0.00 0.00
0.504 0.01 0.00
0.504 0.01 0.00
0.499 0.01 0.00
0.499 0.01 0.00
0.500 0.01 0.00
0.501 0.00 0.00
0.503 0.01 0.00
0.504 0.01 0.00
0.499 0.01 0.00
0.499 0.01 0.00
0.499 0.01 0.00
0.502 0.00 0.00
0.505 0.01 0.00
0.505 0.01 0.00
0.499 0.01 0.00
0.499 0.01 0.00
0.500 0.01 0.00
0.501 0.00 0.00
0.503 0.01 0.00
0.504 0.01 0.00
0.499 0.01 0.00
0.500 0.01 0.00
0.500 0.01 0.00
0.501 0.00 0.00
0.503 0.01 0.00
0.503 0.01 0.00

CL goals
0.199
0.196
0.196
0.200
0.198
0.194
0.199
0.201
0.206
0.200
0.198
0.196
0.200
0.202
0.205
0.200
0.198
0.195
0.200
0.201
0.203
0.200
0.197
0.193
0.200
0.203
0.206
0.200
0.197
0.194
0.200
0.203
0.207
0.200
0.198
0.195

Adaptation Process
Se
P
CL se
0.01 0.00
0.200
0.02 0.00
0.200
0.03 0.00
0.199
0.01 0.00
0.200
0.02 0.00
0.201
0.03 0.00
0.204
0.01 0.00
0.201
0.02 0.00
0.199
0.03 0.00
0.196
0.01 0.00
0.201
0.02 0.00
0.202
0.02 0.00
0.204
0.01 0.00
0.200
0.02 0.00
0.197
0.02 0.00
0.193
0.01 0.00
0.200
0.01 0.00
0.202
0.02 0.00
0.204
0.01 0.00
0.200
0.02 0.00
0.196
0.02 0.00
0.192
0.00 0.00
0.201
0.01 0.00
0.206
0.02 0.00
0.211
0.01 0.00
0.200
0.02 0.00
0.197
0.02 0.00
0.194
0.00 0.00
0.201
0.01 0.00
0.204
0.02 0.00
0.207
0.01 0.00
0.200
0.01 0.00
0.196
0.02 0.00
0.193
0.00 0.00
0.201
0.01 0.00
0.203
0.02 0.00
0.205

154	
  

Se
0.01
0.02
0.03
0.01
0.02
0.03
0.01
0.02
0.03
0.01
0.02
0.02
0.01
0.02
0.03
0.01
0.02
0.02
0.01
0.02
0.02
0.01
0.01
0.02
0.01
0.02
0.02
0.01
0.01
0.02
0.01
0.02
0.02
0.01
0.01
0.02

P
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00

CL goals
0.406
0.407
0.407
0.398
0.397
0.397
0.407
0.405
0.404
0.400
0.398
0.397
0.405
0.404
0.403
0.399
0.398
0.398
0.404
0.405
0.405
0.400
0.399
0.398
0.405
0.404
0.403
0.402
0.401
0.400
0.403
0.402
0.402
0.400
0.401
0.401

Performance
Se
P
0.01 0.00
0.02 0.00
0.02 0.00
0.01 0.00
0.01 0.00
0.01 0.00
0.01 0.00
0.02 0.00
0.02 0.00
0.01 0.00
0.01 0.00
0.01 0.00
0.01 0.00
0.02 0.00
0.02 0.00
0.01 0.00
0.01 0.00
0.02 0.00
0.01 0.00
0.02 0.00
0.02 0.00
0.01 0.00
0.02 0.00
0.02 0.00
0.02 0.00
0.02 0.00
0.02 0.00
0.01 0.00
0.02 0.00
0.02 0.00
0.02 0.00
0.02 0.00
0.02 0.00
0.01 0.00
0.02 0.00
0.02 0.00

Process
CL se
0.196
0.196
0.196
0.203
0.203
0.202
0.193
0.195
0.195
0.201
0.202
0.202
0.195
0.196
0.196
0.205
0.204
0.203
0.198
0.198
0.197
0.201
0.199
0.198
0.192
0.192
0.193
0.200
0.200
0.199
0.194
0.194
0.194
0.201
0.202
0.202

Se
0.01
0.02
0.02
0.01
0.02
0.02
0.02
0.02
0.02
0.01
0.02
0.02
0.02
0.02
0.02
0.01
0.02
0.02
0.02
0.02
0.02
0.01
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02

P
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00

	
  
Table	
  14	
  (cont’d)	
  

Y
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se
Se

	
  

Model
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36

Trial Δ
4
4
4
4
4
4
5
5
5
5
5
5
6
6
6
6
6
6
7
7
7
7
7
7
8
8
8
8
8
8
9
9
9
9
9
9

N
150
150
150
200
200
200
150
150
150
200
200
200
150
150
150
200
200
200
150
150
150
200
200
200
150
150
150
200
200
200
150
150
150
200
200
200

Error
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5
0.1
0.3
0.5

Both Processes
AR se
Se
p
0.498 0.01 0.00
0.496 0.01 0.00
0.495 0.02 0.00
0.501 0.01 0.00
0.504 0.01 0.00
0.505 0.01 0.00
0.499 0.01 0.00
0.496 0.01 0.00
0.495 0.02 0.00
0.501 0.01 0.00
0.504 0.01 0.00
0.505 0.01 0.00
0.498 0.01 0.00
0.496 0.01 0.00
0.496 0.02 0.00
0.501 0.01 0.00
0.504 0.01 0.00
0.505 0.01 0.00
0.498 0.01 0.00
0.497 0.01 0.00
0.496 0.02 0.00
0.501 0.00 0.00
0.504 0.01 0.00
0.505 0.01 0.00
0.499 0.01 0.00
0.497 0.01 0.00
0.496 0.02 0.00
0.501 0.00 0.00
0.504 0.01 0.00
0.505 0.01 0.00
0.498 0.01 0.00
0.496 0.01 0.00
0.496 0.02 0.00
0.501 0.00 0.00
0.503 0.01 0.00
0.504 0.01 0.00

Adaptation Process
CL perf
Se
P
0.401 0.01 0.00
0.402 0.02 0.00
0.403 0.03 0.00
0.398 0.01 0.00
0.395 0.02 0.00
0.393 0.03 0.00
0.401 0.01 0.00
0.404 0.02 0.00
0.404 0.03 0.00
0.398 0.01 0.00
0.395 0.02 0.00
0.392 0.02 0.00
0.400 0.01 0.00
0.398 0.02 0.00
0.395 0.02 0.00
0.398 0.01 0.00
0.394 0.01 0.00
0.392 0.02 0.00
0.401 0.01 0.00
0.402 0.02 0.00
0.402 0.02 0.00
0.398 0.00 0.00
0.395 0.01 0.00
0.394 0.02 0.00
0.401 0.01 0.00
0.401 0.01 0.00
0.400 0.02 0.00
0.398 0.00 0.00
0.395 0.01 0.00
0.394 0.02 0.00
0.401 0.01 0.00
0.402 0.01 0.00
0.400 0.02 0.00
0.398 0.00 0.00
0.397 0.01 0.00
0.397 0.02 0.00

155	
  

Performance Process
CL perf Se
P
0.199 0.01 0.00
0.200 0.01 0.00
0.201 0.01 0.00
0.198 0.01 0.00
0.198 0.01 0.00
0.198 0.01 0.00
0.199 0.01 0.00
0.200 0.01 0.00
0.200 0.01 0.00
0.199 0.01 0.00
0.198 0.01 0.00
0.199 0.01 0.00
0.202 0.01 0.00
0.203 0.01 0.00
0.202 0.01 0.00
0.199 0.01 0.00
0.199 0.01 0.00
0.199 0.01 0.00
0.202 0.01 0.00
0.201 0.01 0.00
0.200 0.02 0.00
0.200 0.01 0.00
0.199 0.01 0.00
0.198 0.01 0.00
0.202 0.01 0.00
0.201 0.02 0.00
0.200 0.02 0.00
0.202 0.01 0.00
0.201 0.01 0.00
0.200 0.01 0.00
0.202 0.01 0.00
0.199 0.02 0.00
0.197 0.02 0.00
0.202 0.01 0.00
0.199 0.01 0.00
0.198 0.02 0.00

	
  
APPENDIX	
  B:	
  Overall	
  Flow	
  of	
  the	
  Two-­‐Day	
  Experiment	
  
Table	
  15	
  
Day	
  1:	
  Training	
  
Action	
  
Consent,	
  Demographics	
  and	
  Trait	
  Goal	
  Orientation	
  	
  
TANDEM	
  Demo	
  by	
  Experimenter	
  

Time	
   Minutes	
   Description	
  
0:10	
  
0:10	
   Familiarization	
  
0:15	
  
0:05	
   Familiarization	
  
0:17	
  
0:02	
   Familiarization	
  
Manual	
  Familiarization	
  
0:18	
  
0:01	
   Familiarization	
  
Familiarization	
  Trial	
  
0:19	
  
0:01	
   Familiarization	
  
Feedback	
  Familiarization	
  
0:23	
  
0:04	
   Training	
  -­‐	
  Block	
  1	
   Instructions	
  –	
  Block	
  1	
  
0:25	
  
0:02	
   Training	
  -­‐	
  Block	
  1	
   Manual	
  
0:29	
  
0:04	
   Training	
  -­‐	
  Block	
  1	
   Training	
  Scenario	
  1	
  
0:30	
  
0:01	
   Training	
  -­‐	
  Block	
  1	
   Feedback	
  
0:32	
  
0:02	
   Training	
  -­‐	
  Block	
  1	
   Manual	
  
0:36	
  
0:04	
   Training	
  -­‐	
  Block	
  1	
   Training	
  Scenario	
  2	
  
0:37	
  
0:01	
   Training	
  -­‐	
  Block	
  1	
   Feedback	
  
0:39	
  
0:02	
   Training	
  -­‐	
  Block	
  1	
   Manual	
  
0:43	
  
0:04	
   Training	
  -­‐	
  Block	
  1	
   Training	
  Scenario	
  3	
  
0:44	
  
0:01	
   Training	
  -­‐	
  Block	
  1	
   Feedback	
  
0:48	
  
0:04	
   Training	
  -­‐	
  Block	
  2	
   Instructions	
  –	
  Block	
  2	
  
0:50	
  
0:02	
   Training	
  -­‐	
  Block	
  2	
   Manual	
  
0:54	
  
0:04	
   Training	
  -­‐	
  Block	
  2	
   Training	
  Scenario	
  4	
  
0:55	
  
0:01	
   Training	
  -­‐	
  Block	
  2	
   Feedback	
  
0:57	
  
0:02	
   Training	
  -­‐	
  Block	
  2	
   Manual	
  
1:01	
  
0:04	
   Training	
  -­‐	
  Block	
  2	
   Training	
  Scenario	
  5	
  
1:02	
  
0:01	
   Training	
  -­‐	
  Block	
  2	
   Feedback	
  
1:04	
  
0:02	
   Training	
  -­‐	
  Block	
  2	
   Manual	
  
1:08	
  
0:04	
   Training	
  -­‐	
  Block	
  2	
   Training	
  Scenario	
  6	
  
1:09	
  
0:01	
   Training	
  -­‐	
  Block	
  2	
   Feedback	
  
1:13	
  
0:04	
   Training	
  -­‐	
  Block	
  3	
   Instructions	
  –	
  Block	
  3	
  
1:15	
  
0:02	
   Training	
  -­‐	
  Block	
  3	
   Manual	
  
1:19	
  
0:04	
   Training	
  -­‐	
  Block	
  3	
   Training	
  Scenario	
  7	
  
1:20	
  
0:01	
   Training	
  -­‐	
  Block	
  3	
   Feedback	
  
1:22	
  
0:02	
   Training	
  -­‐	
  Block	
  3	
   Manual	
  
1:26	
  
0:04	
   Training	
  -­‐	
  Block	
  3	
   Training	
  Scenario	
  8	
  
1:27	
  
0:01	
   Training	
  -­‐	
  Block	
  3	
   Feedback	
  
1:29	
  
0:02	
   Training	
  -­‐	
  Block	
  3	
   Manual	
  
1:33	
  
0:04	
   Training	
  -­‐	
  Block	
  3	
   Training	
  Scenario	
  9	
  
1:34	
  
0:01	
   Training	
  -­‐	
  Block	
  3	
   Feedback	
  
1:49	
  
0:15	
   Measures	
  
TANDEM	
  Knowledge	
  Test,	
  State	
  Goal	
  Orientation	
  	
  
Day	
  1	
  	
  	
  TOTAL	
  TIME	
  =	
  2:00	
  =	
  4	
  credits	
  
	
  

156	
  

	
  
Table	
  16	
  
Day	
  2:	
  Performance	
  
Action	
  
Re-­‐Read	
  Training	
  Topics	
  

Time	
   Minutes	
   Description	
  
0:08	
  
0:08	
   Re-­‐training	
  
0:10	
  
0:02	
   Re-­‐training	
  
Manual	
  familiarization	
  
0:14	
  
0:04	
   Re-­‐training	
  
Training	
  Scenario	
  
0:15	
  
0:01	
   Re-­‐training	
  
Feedback	
  
0:25	
  
0:10	
   Measures	
  
TANDEM	
  Knowledge	
  Test,	
  State	
  Goal	
  Orientation	
  
0:27	
  
0:02	
   Routine	
  	
  
Instructions	
  
0:28	
  
0:01	
   Routine	
  
SR	
  Measures	
  	
  
0:29	
  
0:01	
   Routine	
  	
  
Manual	
  
0:33	
  
0:04	
   Routine	
  	
  
Scenario	
  1	
  (Routine	
  Performance	
  Trial)	
  
0:34	
  
0:01	
   Routine	
  	
  
Feedback	
  
0:41	
  
0:07	
   Routine	
  	
  
SR	
  Measures,	
  Manual,	
  Scenario	
  2,	
  Feedback	
  
0:48	
  
0:07	
   Routine	
  	
  
SR	
  Measures,	
  Manual,	
  Scenario	
  3,	
  Feedback	
  
0:55	
  
0:07	
   Routine	
  	
  
SR	
  Measures,	
  Manual,	
  Scenario	
  4,	
  Feedback	
  
1:02	
  
0:07	
   Routine	
  	
  
SR	
  Measures,	
  Manual,	
  Scenario	
  5,	
  Feedback	
  
1:05	
  
0:03	
   Adaptive	
  
Instructions	
  (and	
  Reminder	
  of	
  Reward)	
  
1:06	
  
0:01	
   Adaptive	
  
SR	
  Measures	
  
1:07	
  
0:01	
   Adaptive	
  
Manual	
  
1:11	
  
0:04	
   Adaptive	
  
Scenario	
  6	
  (Adaptive	
  Performance	
  Trial)	
  
1:12	
  
0:01	
   Adaptive	
  
Feedback	
  
1:19	
  
0:07	
   Adaptive	
  
SR	
  Measures,	
  Manual,	
  Scenario	
  7,	
  Feedback	
  
1:26	
  
0:07	
   Adaptive	
  
SR	
  Measures,	
  Manual,	
  Scenario	
  8,	
  Feedback	
  
1:33	
  
0:07	
   Adaptive	
  
SR	
  Measures,	
  Manual,	
  Scenario	
  9,	
  Feedback	
  
1:40	
  
0:07	
   Adaptive	
  
SR	
  Measures,	
  Manual,	
  Scenario10,	
  Feedback	
  
1:47	
  
0:07	
   Adaptive	
  
SR	
  Measures,	
  Manual,	
  Scenario11,	
  Feedback	
  
1:54	
  
0:07	
   Adaptive	
  
SR	
  Measures,	
  Manual,	
  Scenario12,	
  Feedback	
  
2:01	
  
0:07	
   Adaptive	
  
SR	
  Measures,	
  Manual,	
  Scenario13,	
  Feedback	
  
2:08	
  
0:07	
   Adaptive	
  
SR	
  Measures,	
  Manual,	
  Scenario14,	
  Feedback	
  
2:15	
  
0:07	
   Adaptive	
  
SR	
  Measures,	
  Manual,	
  Scenario15,	
  Feedback	
  
2:22	
  
0:07	
   Adaptive	
  
SR	
  Measures,	
  Manual,	
  Scenario16,	
  Feedback	
  
2:29	
  
0:07	
   Adaptive	
  
SR	
  Measures,	
  Manual,	
  Scenario17,	
  Feedback	
  
2:36	
  
0:07	
   Adaptive	
  
SR	
  Measures,	
  Manual,	
  Scenario18,	
  Feedback	
  
2:43	
  
0:07	
   Adaptive	
  
SR	
  Measures,	
  Manual,	
  Scenario19,	
  Feedback	
  
2:50	
  
0:07	
   Adaptive	
  
SR	
  Measures,	
  Manual,	
  Scenario20,	
  Feedback	
  
2:51	
  
0:01	
   Adaptive	
  
SR	
  Measures	
  
2:52	
  
0:01	
   Adaptive	
  
Debrief	
  
Day	
  2	
  	
  	
  TOTAL	
  TIME	
  =	
  4:00	
  =	
  6	
  credits	
  

	
  

157	
  

	
  
APPENDIX	
  C:	
  
Training	
  Materials	
  
Practice	
  Topics	
  
	
  
Training	
  Topics	
  for	
  Training	
  Block	
  1	
  
In	
  this	
  first	
  block	
  of	
  three	
  trials,	
  the	
  major	
  focus	
  of	
  training	
  is	
  getting	
  familiar	
  with	
  the	
  
simulation	
  and	
  making	
  contact	
  decisions.	
  You	
  should	
  focus	
  on	
  the	
  following	
  training	
  topics:	
  
	
  
1. Using	
  the	
  mouse	
  and	
  other	
  equipment	
  to	
  operate	
  the	
  simulation.	
  
	
  
2. Hooking	
  contacts	
  and	
  accessing	
  the	
  pull-­‐down	
  menus.	
  
	
  
3. Making	
  TYPE	
  contact	
  decisions.	
  
	
  
4. Making	
  CLASS	
  contact	
  decisions.	
  
	
  
5. Making	
  INTENT	
  contact	
  decisions.	
  
	
  
6. Making	
  FINAL	
  ENGAGEMENT	
  contact	
  decisions.	
  
	
  
7. Viewing	
  right	
  button	
  feedback	
  after	
  making	
  contact	
  decisions.	
  
	
  
	
  
Training	
  Topics	
  for	
  Training	
  Block	
  2	
  
In	
  this	
  second	
  block	
  of	
  three	
  trials,	
  the	
  major	
  focus	
  of	
  training	
  is	
  preventing	
  contacts	
  from	
  
crossing	
  the	
  defensive	
  perimeters.	
  You	
  should	
  focus	
  on	
  the	
  following	
  training	
  topics:	
  
	
  
1. Using	
  the	
  zoom	
  function	
  to	
  view	
  the	
  “big	
  picture”	
  and	
  monitoring	
  the	
  inner	
  and	
  
outer	
  perimeters.	
  
	
  
2. Using	
  marker	
  contacts	
  to	
  locate	
  the	
  outer	
  defensive	
  perimeter.	
  
	
  
3. Watching	
  for	
  pop-­‐up	
  contacts	
  that	
  appear	
  suddenly	
  on	
  your	
  screen.	
  
	
  
	
  
Training	
  Topics	
  for	
  Training	
  Block	
  3	
  
In	
  this	
  last	
  block	
  of	
  three	
  trials,	
  the	
  major	
  focus	
  of	
  training	
  is	
  being	
  able	
  to	
  apply	
  strategies	
  
that	
  are	
  used	
  to	
  better	
  prevent	
  contacts	
  from	
  crossing	
  the	
  defensive	
  perimeters.	
  You	
  should	
  
focus	
  on	
  the	
  following	
  training	
  topics:	
  
	
  
1. Prioritizing	
  contacts	
  located	
  on	
  the	
  radar	
  screen	
  to	
  determine	
  high	
  and	
  low	
  priority	
  
contacts	
  and	
  the	
  order	
  in	
  which	
  contacts	
  should	
  be	
  prosecuted.	
  
	
  
2. Making	
  trade-­‐offs	
  between	
  contacts	
  that	
  are	
  approaching	
  your	
  inner	
  and	
  outer	
  
defensive	
  perimeters.	
  
	
  

158	
  

	
  
Possible	
  Errors	
  
	
  
Errors	
  –	
  Block	
  1	
  
For	
  each	
  of	
  the	
  training	
  topics	
  listed	
  above,	
  there	
  is	
  the	
  potential	
  for	
  a	
  number	
  of	
  errors.	
  	
  
Some	
  of	
  the	
  mistakes	
  that	
  can	
  be	
  made	
  in	
  these	
  areas	
  are	
  listed	
  below:	
  
1. Clicking	
  on	
  the	
  wrong	
  mouse	
  button	
  (left/right)	
  to	
  hook	
  a	
  contact	
  or	
  access	
  a	
  menu.	
  
2. Not	
  properly	
  evaluating	
  contact	
  information	
  and	
  making	
  incorrect	
  contact	
  sub-­‐
decisions	
  (TYPE,	
  CLASS,	
  INTENT)	
  and	
  decisions	
  (FINAL	
  ENGAGEMENT).	
  
3. Making	
  contact	
  sub-­‐decisions	
  based	
  on	
  a	
  single	
  cue	
  value.	
  	
  For	
  example,	
  deciding	
  a	
  
contact’s	
  TYPE	
  based	
  only	
  on	
  speed	
  information.	
  
4. Making	
  contact	
  decisions	
  too	
  quickly.	
  
	
  
Errors	
  –	
  Block	
  2	
  
For	
  each	
  of	
  the	
  training	
  topics	
  listed	
  above,	
  there	
  is	
  the	
  potential	
  for	
  a	
  number	
  of	
  errors.	
  
Some	
  of	
  the	
  common	
  mistakes	
  in	
  these	
  areas	
  are	
  listed	
  below	
  
1. Focusing	
  on	
  only	
  the	
  inner	
  perimeter	
  rather	
  than	
  zooming	
  out	
  to	
  see	
  the	
  “big	
  
picture”	
  and	
  to	
  monitor	
  the	
  outer	
  perimeter.	
  
2. Hooking	
  the	
  wrong	
  marker	
  contacts	
  or	
  not	
  using	
  marker	
  contacts	
  to	
  locate	
  the	
  outer	
  
perimeter.	
  
3. Focusing	
  only	
  on	
  stable	
  contacts	
  and	
  ignoring	
  contacts	
  that	
  pop-­‐up	
  suddenly	
  on	
  the	
  
screen.	
  Often	
  people	
  do	
  not	
  monitor	
  their	
  screen	
  for	
  pop-­‐up	
  contacts.	
  
4. Letting	
  contacts	
  cross	
  the	
  inner	
  and	
  outer	
  defensive	
  perimeters.	
  
	
  
Errors	
  –	
  Block	
  3	
  
For	
  each	
  of	
  the	
  training	
  topics	
  listed	
  above,	
  there	
  is	
  the	
  potential	
  for	
  a	
  number	
  of	
  errors.	
  	
  
Some	
  of	
  the	
  common	
  mistakes	
  in	
  these	
  areas	
  are	
  listed	
  below:	
  
1. Focusing	
  on	
  low	
  priority	
  rather	
  than	
  high	
  priority	
  contacts.	
  
2. Not	
  checking	
  the	
  speeds	
  of	
  contacts	
  close	
  to	
  the	
  perimeters.	
  
3. Preventing	
  all	
  contacts	
  from	
  crossing	
  one	
  perimeter	
  while	
  ignoring	
  the	
  other	
  
perimeter.	
  
	
  
	
  
	
  
	
  
Error	
  Encouragement	
  Framing	
  
	
  
During	
  training,	
  you	
  are	
  encouraged	
  to	
  make	
  these	
  errors.	
  For	
  training	
  to	
  be	
  effective,	
  you	
  
should	
  make	
  these	
  errors.	
  Errors	
  are	
  a	
  positive	
  part	
  of	
  the	
  learning	
  experience.	
  As	
  a	
  result	
  
of	
  making	
  errors,	
  you	
  can	
  learn	
  from	
  your	
  mistakes	
  and	
  develop	
  a	
  better	
  understanding	
  of	
  
the	
  simulation.	
  The	
  more	
  errors	
  you	
  make	
  the	
  more	
  you	
  learn.	
  
	
  

	
  

159	
  

	
  
Exploratory	
  Learning	
  Instructions	
  
	
  
	
  
Task	
  Instructions	
  –	
  Block	
  1	
  
	
  
An	
  effective	
  method	
  for	
  learning	
  the	
  skills	
  just	
  discussed	
  is	
  to	
  explore	
  the	
  task	
  and	
  develop	
  
your	
  own	
  understanding	
  of	
  it.	
  As	
  you	
  practice	
  the	
  scenarios,	
  explore	
  the	
  task	
  to	
  understand	
  
what	
  is	
  occurring	
  in	
  the	
  scenario,	
  and	
  discover	
  the	
  best	
  strategy	
  to	
  deal	
  with	
  the	
  situation.	
  
Also,	
  experiment	
  with	
  different	
  strategies	
  and	
  methods	
  as	
  you	
  explore	
  the	
  task	
  and	
  learn	
  
important	
  task	
  skills.	
  Remember,	
  your	
  task	
  is	
  to	
  learn	
  the	
  basic	
  features	
  of	
  the	
  simulation,	
  
hook	
  the	
  contacts	
  and	
  use	
  the	
  pull-­‐down	
  menus,	
  make	
  contact	
  decisions,	
  and	
  view	
  right-­‐
button	
  feedback	
  following	
  contact	
  decisions.	
  
	
  
	
  
Task	
  Instructions	
  –	
  Block	
  2	
  
	
  
An	
  effective	
  method	
  for	
  learning	
  the	
  skills	
  just	
  discussed	
  is	
  to	
  explore	
  the	
  task	
  and	
  develop	
  
your	
  own	
  understanding	
  of	
  it.	
  As	
  you	
  practice	
  the	
  scenarios,	
  explore	
  the	
  task	
  to	
  understand	
  
what	
  is	
  occurring	
  in	
  the	
  scenario,	
  and	
  discover	
  the	
  best	
  strategy	
  to	
  deal	
  with	
  the	
  situation.	
  
Also,	
  experiment	
  with	
  different	
  strategies	
  and	
  methods	
  as	
  you	
  explore	
  the	
  task	
  and	
  learn	
  
important	
  task	
  skills.	
  Remember,	
  your	
  task	
  is	
  to	
  learn	
  how	
  to	
  prevent	
  contacts	
  from	
  
crossing	
  your	
  perimeters.	
  To	
  do	
  this	
  effectively,	
  you	
  will	
  need	
  to	
  learn	
  how	
  to	
  use	
  the	
  zoom	
  
function,	
  how	
  to	
  use	
  marker	
  contacts	
  to	
  locate	
  the	
  outer	
  defensive	
  perimeter,	
  and	
  how	
  to	
  
watch	
  for	
  pop-­‐up	
  contacts	
  that	
  appear	
  suddenly	
  on	
  your	
  screen.	
  
	
  
	
  
Task	
  Instructions	
  –	
  Block	
  3	
  
	
  
An	
  effective	
  method	
  for	
  learning	
  the	
  skills	
  just	
  discussed	
  is	
  to	
  explore	
  the	
  task	
  and	
  develop	
  
your	
  own	
  understanding	
  of	
  it.	
  As	
  you	
  practice	
  the	
  scenarios,	
  explore	
  the	
  task	
  to	
  understand	
  
what	
  is	
  occurring	
  in	
  the	
  scenario,	
  and	
  discover	
  the	
  best	
  strategy	
  to	
  deal	
  with	
  the	
  situation.	
  
Also,	
  experiment	
  with	
  different	
  strategies	
  and	
  methods	
  as	
  you	
  explore	
  the	
  task	
  and	
  learn	
  
important	
  task	
  skills.	
  Remember,	
  your	
  task	
  is	
  to	
  learn	
  how	
  to	
  prioritize	
  contacts	
  and	
  make	
  
tradeoffs	
  between	
  contacts	
  that	
  are	
  approaching	
  your	
  inner	
  and	
  outer	
  perimeters.	
  
	
  
	
  

	
  

160	
  

	
  
	
  
Contact	
  Cue	
  Values	
  
	
  
Listed	
  below	
  are	
  the	
  cue	
  values	
  for	
  different	
  type	
  of	
  contacts.	
  Remember,	
  as	
  you	
  make	
  
TYPE,	
  CLASS,	
  and	
  INTENT	
  decisions	
  you	
  want	
  to	
  select	
  the	
  option	
  indicated	
  by	
  the	
  
MAJORITY	
  of	
  the	
  cue	
  values.	
  Note:	
  you	
  will	
  not	
  be	
  able	
  to	
  use	
  this	
  sheet	
  in	
  the	
  final	
  two	
  
trials.	
  
	
  	
  
Contact	
  type	
  
AIR	
  
Speed	
  >	
  35	
  knots	
  
Altitude/Depth	
  >	
  0	
  feet	
  
Communication	
  	
  Time	
  =	
  0	
  –	
  40	
  s	
  
SURFACE	
  
Speed	
  =	
  25	
  –	
  35	
  knots	
  
Altitude/Depth	
  =	
  0	
  feet	
  
Communication	
  	
  Time	
  =	
  41	
  –	
  80	
  s	
  
SUB	
  
Speed	
  =	
  0	
  –	
  24	
  knots	
  
Altitude/Depth	
  <	
  0	
  feet	
  
Communication	
  	
  Time	
  =	
  81	
  –	
  120	
  s	
  
	
  
Contact	
  class	
  
CIVILIAN	
  
Intelligence	
  =	
  Private	
  
Direction	
  of	
  Origin	
  =	
  Blue	
  Lagoon	
  
Maneuvering	
  Pattern	
  =	
  Code	
  Foxtrot	
  
UNKNOWN	
  
Intelligence	
  =	
  Unavailable	
  
Direction	
  of	
  Origin	
  =	
  Unknown	
  
Maneuvering	
  Pattern	
  =	
  Code	
  Echo	
  
MILITARY	
  
Intelligence	
  =	
  Platform	
  
Direction	
  of	
  Origin	
  =	
  Red	
  Sea	
  
Maneuvering	
  Pattern	
  =	
  Code	
  Delta	
  
	
  
Contact	
  intent	
  
PEACEFUL	
  
Countermeasures	
  =	
  None	
  
Threat	
  Level	
  =	
  1	
  
Response	
  =	
  Given	
  
UNKNOWN	
  
Countermeasures	
  =	
  Unknown	
  
Threat	
  Level	
  =	
  2	
  
Response	
  =	
  Inaudible	
  
HOSTILE	
  
Countermeasures	
  =	
  Jamming	
  
Threat	
  Level	
  =	
  3	
  
Response	
  =	
  No	
  Response	
  
	
  
Final	
  engagement	
  decision	
  
CLEAR	
  
If	
  INTENT	
  =	
  Peaceful	
  
SHOOT	
  
If	
  INTENT	
  =	
  Hostile	
  
	
  

	
  

161	
  

	
  
APPENDIX	
  D:	
  
Measures	
  
	
  

Metacognition	
  	
  
For	
  each	
  of	
  the	
  items	
  below,	
  rate	
  the	
  extent	
  to	
  which	
  you	
  were	
  thinking	
  about	
  these	
  issues	
  
during	
  THE	
  LAST	
  TRIAL.	
  Please	
  use	
  the	
  scale	
  shown	
  below	
  to	
  make	
  your	
  ratings.	
  
	
  
	
  	
  	
  	
  1	
   	
  
	
  
2	
  
	
  
	
  
3	
  
	
  
	
  
4	
  
	
  
	
  
5	
  
Never	
   	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  Rarely	
   	
  
	
  	
  	
  	
  Sometimes	
   	
  
	
  	
  	
  	
  	
  Frequently	
  	
  
	
  	
  	
  	
  Constantly	
  
	
  
1) During	
  the	
  trial,	
  I	
  monitored	
  how	
  well	
  I	
  was	
  learning	
  its	
  requirements.	
  
2) When	
  my	
  methods	
  were	
  not	
  successful,	
  I	
  experimented	
  with	
  different	
  procedures	
  
for	
  performing	
  the	
  task.	
  
3) I	
  thought	
  about	
  new	
  strategies	
  for	
  improving	
  my	
  performance.	
  
4) I	
  monitored	
  closely	
  the	
  areas	
  where	
  I	
  needed	
  the	
  most	
  study	
  and	
  practice.	
  
	
  
	
  
	
  
Self-­‐efficacy	
  
This	
  set	
  of	
  questions	
  asks	
  you	
  to	
  describe	
  how	
  you	
  feel	
  about	
  your	
  capabilities	
  for	
  
performing	
  ON	
  THE	
  NEXT	
  TRIAL	
  of	
  this	
  simulation	
  using	
  the	
  following	
  scale.	
  
	
  
	
  	
  	
  	
  	
  	
  1	
   	
  
	
  
2	
  
	
  
	
  
3	
  
	
  
	
  
4	
  
	
  
	
  
5	
  
Strongly	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  
Neither	
  Agree	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  
	
  
	
  	
  	
  	
  	
  	
  Strongly	
  
Disagree	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Disagree	
   	
  
Nor	
  Disagree	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
   	
  	
  	
  	
  	
  	
  	
  	
  	
  Agree	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
   	
  	
  	
  	
  	
  	
  	
  	
  Agree	
  
	
  	
  
1) I	
  can	
  meet	
  the	
  challenges	
  of	
  this	
  simulation.	
  
2) I	
  am	
  certain	
  that	
  I	
  can	
  manage	
  the	
  requirements	
  of	
  this	
  task.	
  
3) I	
  believe	
  I	
  can	
  develop	
  methods	
  to	
  handle	
  changing	
  aspects	
  of	
  this	
  task.	
  
4) I	
  am	
  certain	
  I	
  can	
  cope	
  with	
  the	
  task	
  components	
  competing	
  for	
  my	
  time.	
  
	
  
	
  
	
  
Goal	
  level	
  
Please	
  indicate	
  your	
  desired	
  level	
  of	
  performance	
  ON	
  THE	
  NEXT	
  TRIAL.	
  
	
  
1) How	
  many	
  points	
  do	
  you	
  plan	
  to	
  score	
  during	
  the	
  next	
  trial?	
  ________	
  
2) How	
  many	
  targets	
  do	
  you	
  plan	
  to	
  prosecute	
  correctly?	
  ______	
  
3) Please	
  indicate	
  which	
  goal	
  is	
  more	
  important	
  to	
  you	
  for	
  THE	
  NEXT	
  TRIAL	
  using	
  the	
  
following	
  scale.	
  
	
  
Figure	
  out	
   	
  
	
  
	
  
	
  
	
  
	
  
	
  	
  	
  	
  	
  	
  
	
  
	
  	
  Do	
  	
  
as	
  much	
  as	
  I	
  can	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  	
  	
  	
  as	
  much	
  as	
  I	
  can	
  	
  
|__________|__________|__________|__________|__________|__________|__________|__________|	
  
1	
  
	
  	
  	
  	
  2	
   	
  	
  	
  	
  	
  	
  	
  	
  	
  3	
   	
  
4	
  
	
  	
  	
  	
  5	
   	
  	
  	
  	
  	
  	
  	
  	
  	
  6	
   	
  
7	
  
	
  	
  	
  	
  	
  8	
   	
  	
  	
  	
  	
  	
  	
  	
  	
  9	
  
	
  

162	
  

	
  
	
  
Evaluation	
  	
  
(Behavioral)	
  Total	
  amount	
  of	
  time	
  spent	
  reviewing	
  performance	
  feedback	
  information.	
  
	
  
	
  
Learning-­‐oriented	
  Effort	
  
(Behavioral)	
  Amount	
  of	
  information	
  sought	
  (number	
  of	
  pages	
  viewed)	
  from	
  the	
  manual	
  
(Behavioral)	
  Amount	
  of	
  time	
  spent	
  investigating	
  information	
  from	
  the	
  manual	
  
	
  
1) Please	
  use	
  the	
  following	
  scale	
  to	
  indicate	
  the	
  extent	
  to	
  which	
  you	
  focused	
  on	
  
understanding	
  what	
  changed	
  in	
  the	
  task	
  IN	
  THE	
  LAST	
  TRIAL.	
  
	
  
	
  	
  	
  	
  	
  1	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  2	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  3	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  4	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  5	
  
Not	
  at	
   	
  	
  	
  	
  	
  	
  	
  	
  	
  A	
  very	
  small	
  
	
  	
  A	
  small	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  A	
  moderate	
   	
  	
  	
  	
  	
  A	
  great	
  
	
  	
  	
  All	
   	
  
extent	
  	
  
	
  	
  extent	
  
	
  	
  	
  	
  extent	
  
	
  	
  	
  	
  	
  	
  extent	
  
	
  
	
  
Outcome-­‐oriented	
  Effort	
  	
  
(Behavioral)	
  Total	
  amount	
  of	
  behaviors	
  (total	
  number	
  of	
  clicks)	
  
	
  
	
  
Performance	
  
(Behavioral)	
  Total	
  score	
  =	
  (Correct	
  execution	
  of	
  targets	
  -­‐	
  penalties	
  for	
  incorrect	
  decisions	
  -­‐	
  
penalties	
  for	
  inner	
  perimeter	
  crosses	
  -­‐	
  penalties	
  for	
  outer	
  perimeter	
  crosses)	
  
	
  
	
  
Additional	
  Measures	
  	
  
1) When	
  you	
  think	
  you’ve	
  figured	
  out	
  what	
  changed	
  in	
  the	
  task,	
  answer	
  this	
  question:	
  
What	
  changed	
  in	
  the	
  task?	
  	
  ___________________________________________________________________	
  
2) Please	
  use	
  the	
  scale	
  below	
  to	
  indicate	
  the	
  focus	
  of	
  your	
  attention	
  and	
  action	
  IN	
  THE	
  
LAST	
  TRIAL.	
  
	
  
Executing	
  as	
  many	
  	
   	
  
	
  
	
  
	
  
	
  
	
  
	
  	
  Figuring	
  out	
  
targets	
  as	
  I	
  could	
  
	
  
	
  
	
  
	
  
	
  
	
  
what	
  changed	
  
|__________|__________|__________|__________|__________|__________|__________|__________|	
  
1	
  
	
  	
  	
  	
  2	
   	
  	
  	
  	
  	
  	
  	
  	
  	
  3	
   	
  
4	
  
	
  	
  	
  	
  5	
   	
  	
  	
  	
  	
  	
  	
  	
  	
  6	
   	
  
7	
  
	
  	
  	
  	
  	
  8	
   	
  	
  	
  	
  	
  	
  	
  	
  	
  9	
  
	
  
3) Choose	
  the	
  option	
  below	
  that	
  best	
  represents	
  the	
  level	
  of	
  the	
  goal	
  you	
  are	
  making:	
  	
  	
  
	
  
	
  	
  	
  	
  	
  1	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  2	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  3	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  4	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  5	
  
	
  	
  	
  	
  Very	
  low	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Low	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
   Medium	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
   	
  	
  	
  	
  	
  High	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
   	
  	
  Very	
  High	
  
	
  

	
  

163	
  

	
  
TANDEM	
  Knowledge	
  Test	
  
The	
  following	
  is	
  a	
  knowledge	
  test	
  about	
  the	
  simulation.	
  Please	
  select	
  the	
  response	
  that	
  best	
  
answers	
  the	
  question.	
  
	
  
1. If	
  a	
  Response	
  is	
  Given,	
  what	
  is	
  the	
  likely	
  Intent	
  of	
  the	
  contact?	
  
a. Military	
  
b. Hostile	
  
c. Civilian	
  
d. Peaceful	
  
	
  
2. A	
  submarine	
  may	
  have	
  which	
  of	
  the	
  following	
  characteristics?	
  
a. Speed	
  30	
  knots,	
  Altitude/Depth	
  -­‐20,	
  Communication	
  time	
  85	
  seconds	
  
b. Speed	
  30	
  knots,	
  Altitude/Depth	
  0,	
  Communication	
  time	
  30	
  seconds	
  
c. Speed	
  20	
  knots,	
  Altitude/Depth	
  0,	
  Communication	
  time	
  80	
  seconds	
  
d. Speed	
  20	
  knots,	
  Altitude/Depth	
  -­‐20,	
  Communication	
  time	
  90	
  seconds	
  
	
  
3. A	
  Maneuvering	
  Pattern	
  of	
  Code	
  Delta	
  indicates	
  the	
  contact	
  is	
  which	
  of	
  the	
  following?	
  
a. Air	
  
b. Military	
  
c. Surface	
  
d. Civilian	
  
	
  
4. A	
  Blue	
  Lagoon	
  Direction	
  of	
  Origin	
  indicates	
  the	
  contact	
  is	
  which	
  of	
  the	
  following?	
  
a. Unknown	
  
b. Sub	
  
c. Civilian	
  
d. Military	
  
	
  
5. If	
  a	
  contact’s	
  Altitude/Depth	
  is	
  10	
  feet,	
  what	
  is	
  the	
  Type	
  of	
  the	
  contact?	
  
a. Air	
  
b. Surface	
  
c. Submarine	
  
d. Unknown	
  
	
  
6. If	
  a	
  contact’s	
  Intelligence	
  is	
  Unavailable,	
  what	
  Class	
  does	
  this	
  suggest	
  for	
  the	
  
contact?	
  
a. Air	
  
b. Civilian	
  
c. Military	
  
d. Unknown	
  
	
  
7. If	
  a	
  contact’s	
  characteristics	
  are	
  Communication	
  Time	
  =	
  20	
  seconds	
  and	
  Speed	
  =	
  50	
  
knots,	
  which	
  of	
  the	
  following	
  actions	
  should	
  you	
  take?	
  
a. Choose	
  Intent	
  is	
  Peaceful	
  
b. Choose	
  Type	
  is	
  Surface	
  
	
  

164	
  

	
  
c. Get	
  another	
  piece	
  of	
  information	
  
d. Choose	
  Type	
  is	
  Air	
  
	
  
8. A	
  communication	
  Time	
  of	
  52	
  seconds	
  indicates	
  that	
  the	
  contact	
  is	
  likely:	
  
a. Air	
  
b. Surface	
  
c. Submarine	
  
d. Unknown	
  
	
  
9. If	
  a	
  contact’s	
  characteristics	
  are	
  Intelligence	
  is	
  Private	
  and	
  Maneuvering	
  Pattern	
  is	
  
Code	
  Foxtrot,	
  which	
  of	
  the	
  following	
  actions	
  should	
  you	
  take?	
  
a. Choose	
  Class	
  is	
  Military	
  
b. Choose	
  Intent	
  is	
  Peaceful	
  
c. Choose	
  Class	
  is	
  Civilian	
  
d. Choose	
  Intent	
  is	
  Unknown	
  
	
  
10. If	
  a	
  contact’s	
  Maneuvering	
  Pattern	
  is	
  Code	
  Echo,	
  this	
  suggests	
  that	
  the	
  contact	
  falls	
  
into	
  which	
  category?	
  
a. Class	
  is	
  Unknown	
  
b. Class	
  is	
  Military	
  
c. Class	
  is	
  Hostile	
  
d. Class	
  is	
  Peaceful	
  
	
  
11. If	
  a	
  contact’s	
  Speed	
  is	
  40	
  knots,	
  what	
  does	
  this	
  suggest	
  about	
  the	
  contact?	
  
a. The	
  contact	
  is	
  Air	
  
b. The	
  contact	
  is	
  Surface	
  
c. The	
  contact	
  is	
  Civilian	
  
d. The	
  contact	
  is	
  Military	
  
	
  
12. Your	
  Outer	
  Defensive	
  Perimeter	
  is	
  located	
  at:	
  
a. 64	
  nm	
  
b. 128	
  nm	
  
c. 256	
  nm	
  
d. 512	
  nm	
  
	
  
13. If	
  you’ve	
  just	
  noticed	
  three	
  contacts	
  near	
  your	
  inner	
  perimeter,	
  which	
  of	
  the	
  
following	
  should	
  you	
  do	
  next?	
  
a. Engage	
  the	
  contact	
  nearest	
  the	
  inner	
  perimeter	
  
b. Engage	
  the	
  faster	
  contact	
  near	
  the	
  inner	
  perimeter	
  
c. Zoom-­‐Out	
  to	
  check	
  the	
  outer	
  perimeter	
  
d. Zoom-­‐In	
  to	
  check	
  how	
  close	
  the	
  contacts	
  are	
  to	
  the	
  inner	
  perimeter	
  
	
  
14. If	
  you	
  Zoom-­‐Out	
  to	
  find	
  three	
  contacts	
  around	
  your	
  Outer	
  Perimeter,	
  how	
  would	
  you	
  
determine	
  which	
  contact	
  is	
  the	
  marker	
  contact?	
  
a. Check	
  to	
  see	
  which	
  contact	
  is	
  closest	
  to	
  the	
  outer	
  perimeter	
  
b. Check	
  the	
  speeds	
  of	
  the	
  contacts	
  
	
  

165	
  

	
  
c. Check	
  to	
  see	
  which	
  contact	
  is	
  Civilian	
  
d. Check	
  to	
  see	
  which	
  contact	
  is	
  Hostile	
  
	
  
15. What	
  is	
  the	
  purpose	
  of	
  marker	
  contacts?	
  
a. To	
  determine	
  which	
  Contacts	
  are	
  Hostile	
  and	
  which	
  are	
  Peaceful	
  
b. To	
  locate	
  your	
  Inner	
  Defensive	
  Perimeter	
  
c. To	
  quickly	
  determine	
  the	
  speeds	
  of	
  contacts	
  near	
  your	
  perimeters	
  
d. To	
  locate	
  your	
  Outer	
  Defensive	
  Perimeter	
  
	
  
16. Which	
  of	
  the	
  following	
  pieces	
  of	
  information	
  is	
  NOT	
  useful	
  for	
  prioritizing	
  contacts?	
  
a. The	
  distance	
  of	
  contacts	
  from	
  the	
  Outer	
  Defensive	
  Perimeter	
  
b. Whether	
  the	
  contact	
  is	
  Peaceful	
  or	
  Hostile	
  
c. The	
  distance	
  of	
  contacts	
  from	
  the	
  Inner	
  Defensive	
  Perimeter	
  
d. The	
  Speed	
  of	
  contacts	
  near	
  your	
  Inner	
  and	
  Outer	
  Defensive	
  Perimeter	
  
	
  
17. Which	
  of	
  the	
  following	
  functions	
  is	
  most	
  useful	
  for	
  identifying	
  marker	
  contacts?	
  
a. Zoom-­‐In	
  
b. Right-­‐button	
  feedback	
  
c. Engage	
  Shoot	
  or	
  Clear	
  
d. Zoom-­‐Out	
  
	
  
18. If	
  three	
  contacts	
  are	
  about	
  10	
  miles	
  outside	
  your	
  Outer	
  Defensive	
  Perimeter,	
  which	
  
of	
  the	
  following	
  should	
  you	
  do	
  to	
  prioritize	
  the	
  contacts?	
  
a. Engage	
  the	
  fastest	
  contact	
  
b. Engage	
  the	
  hostile	
  contact	
  
c. Engage	
  the	
  closest	
  contact	
  
d. It	
  makes	
  no	
  difference	
  in	
  what	
  order	
  you	
  engage	
  the	
  contacts	
  
	
  
19. On	
  the	
  average,	
  approximately	
  how	
  many	
  contacts	
  pop-­‐up	
  during	
  each	
  practice	
  
trial?	
  
a. 1	
  
b. 3	
  
c. 6	
  
d. 9	
  
	
  
20. Which	
  of	
  the	
  following	
  would	
  be	
  the	
  most	
  effective	
  strategy	
  for	
  defending	
  your	
  
Outer	
  Defensive	
  Perimeter?	
  
a. Zoom-­‐Out	
  to	
  128	
  nm,	
  locate	
  the	
  Marker	
  Contacts,	
  and	
  check	
  the	
  Speed	
  of	
  
contacts	
  near	
  the	
  Outer	
  Perimeter	
  
b. Zoom-­‐Out	
  to	
  256	
  nm,	
  locate	
  the	
  Marker	
  Contacts,	
  and	
  check	
  the	
  Speed	
  
of	
  contacts	
  near	
  the	
  Outer	
  Perimeter	
  
c. Zoom-­‐Out	
  to	
  128	
  nm,	
  locate	
  a	
  Hostile	
  Air	
  Contact,	
  and	
  check	
  the	
  Speed	
  of	
  
contacts	
  near	
  the	
  Outer	
  Perimeter	
  
d. Zoom-­‐Out	
  to	
  256	
  nm,	
  locate	
  a	
  Hostile	
  Air	
  Contact,	
  and	
  check	
  the	
  Speed	
  of	
  
contacts	
  near	
  the	
  Outer	
  Perimeter	
  
	
  
	
  

166	
  

	
  

	
  
	
  

	
  

21. If	
  all	
  penalty	
  intrusions	
  cost	
  -­‐100	
  points,	
  which	
  would	
  be	
  the	
  most	
  effective	
  
strategy?	
  
a. Do	
  not	
  allow	
  any	
  contacts	
  to	
  enter	
  your	
  Inner	
  Defensive	
  Perimeter,	
  even	
  if	
  it	
  
means	
  allowing	
  contacts	
  to	
  cross	
  your	
  Outer	
  Defensive	
  Perimeter	
  
b. Do	
  not	
  allow	
  any	
  contacts	
  to	
  enter	
  your	
  Outer	
  Defensive	
  Perimeter,	
  even	
  if	
  it	
  
means	
  allowing	
  contacts	
  to	
  cross	
  your	
  Inner	
  Defensive	
  Perimeter	
  
c. Defend	
  both	
  your	
  Inner	
  and	
  Outer	
  Defensive	
  Perimeters	
  
d. None	
  of	
  these	
  are	
  effective	
  strategies	
  
	
  
22. It	
  is	
  important	
  to	
  make	
  trade-­‐offs	
  between	
  contacts:	
  
a. That	
  are	
  Hostile	
  and	
  those	
  that	
  are	
  Peaceful	
  
b. Approaching	
  your	
  Inner	
  and	
  Outer	
  Perimeters	
  
c. That	
  are	
  Civilian	
  and	
  those	
  that	
  are	
  Military	
  
d. That	
  have	
  already	
  crossed	
  your	
  Inner	
  Defensive	
  Perimeter	
  and	
  those	
  that	
  are	
  
approaching	
  your	
  Outer	
  Defensive	
  Perimeter	
  

167	
  

	
  
Individual	
  Differences	
  and	
  Demographic	
  Measures	
  Explained	
  
Cognitive	
  ability	
  was	
  gathered	
  through	
  asking	
  individuals	
  for	
  their	
  ACT,	
  SAT	
  and	
  
GPA	
  scores.	
  Participants	
  were	
  ensured	
  that	
  their	
  scores	
  will	
  remain	
  confidential	
  and	
  only	
  
be	
  used	
  for	
  research	
  purposes.	
  Previous	
  research	
  suggested	
  that	
  these	
  are	
  acceptable	
  
measurements	
  of	
  cognitive	
  ability	
  (e.g.,	
  Phillips	
  &	
  Gully,	
  1997)	
  and	
  are	
  known	
  to	
  be	
  highly	
  
reliable	
  (e.g.,	
  KR-­‐20	
  =	
  .96	
  for	
  the	
  ACT	
  composite	
  score;	
  American	
  College	
  Testing	
  Program,	
  
1989).	
  	
  Self-­‐reported	
  cognitive	
  ability	
  scores	
  have	
  also	
  been	
  found	
  to	
  have	
  high	
  correlations	
  
with	
  official	
  scores	
  and	
  are	
  therefore	
  considered	
  an	
  acceptable	
  option	
  for	
  gathering	
  this	
  
information	
  (.95;	
  Gully,	
  Payne,	
  Koles	
  &	
  Whiteman,	
  2002).	
  	
  
Individual	
  adaptability,	
  a	
  trait-­‐level,	
  was	
  creased	
  by	
  Ployhart	
  and	
  Bliese	
  (2006)	
  as	
  a	
  
55-­‐item	
  measure	
  (i.e.,	
  the	
  I-­‐ADAPT)	
  that	
  captures	
  individual	
  variability	
  in	
  adaptation,	
  
based	
  on	
  the	
  eight	
  dimensions	
  of	
  adaptive	
  performance	
  developed	
  by	
  Pulakos	
  and	
  
colleagues	
  (2000)	
  This	
  measure	
  uses	
  a	
  five-­‐point	
  Likert-­‐type	
  scale	
  ranging	
  from	
  “strongly	
  
disagree”	
  (1)	
  to	
  “strongly	
  agree”	
  (5)	
  and	
  was	
  given	
  to	
  individuals	
  as	
  soon	
  as	
  they	
  enter	
  the	
  
lab	
  on	
  Day	
  1.	
  Cronbach’s	
  alpha	
  shows	
  that	
  this	
  scale	
  was	
  reliable	
  (α	
  =	
  .827).	
  	
  
Trait	
  goal	
  orientation	
  was	
  measured	
  using	
  a	
  modified	
  version	
  of	
  the	
  13-­‐item	
  
measure	
  developed	
  by	
  VandeWalle	
  (1997).	
  The	
  measure	
  had	
  a	
  six-­‐point	
  Likert-­‐type	
  rating	
  
scale	
  with	
  the	
  range	
  of	
  “strongly	
  disagree”	
  (1)	
  to	
  “strongly	
  agree”	
  (6)	
  and	
  has	
  three	
  
subscales	
  with	
  the	
  following	
  reliabilities:	
  mastery	
  orientation,	
  performance-­‐prove,	
  and	
  
performance-­‐avoid	
  (VandeWalle,	
  1997).	
  Cronbach’s	
  alpha	
  suggests	
  that	
  the	
  overall	
  
measure	
  was	
  reliable	
  (α	
  =	
  .751).	
  	
  
State	
  goal	
  orientation	
  measured	
  learning	
  and	
  performance	
  orientations	
  with	
  regard	
  
to	
  the	
  specific	
  task	
  both	
  at	
  the	
  end	
  of	
  day	
  1	
  and	
  beginning	
  of	
  day	
  2.	
  The	
  measure	
  was	
  used	
  

	
  

168	
  

	
  
by	
  Bell	
  and	
  Kozlowski	
  (2008)	
  in	
  the	
  same	
  task	
  environment,	
  adapted	
  the	
  version	
  published	
  
by	
  Horvath,	
  Scheu,	
  and	
  DeShon	
  (2001),	
  consisted	
  of	
  15	
  items,	
  and	
  had	
  three	
  subscales	
  
(learning,	
  performance	
  prove	
  and	
  performance	
  avoidance)	
  using	
  a	
  five-­‐point	
  Likert-­‐type	
  
scale	
  ranging	
  from	
  “strongly	
  disagree”	
  (1)	
  to	
  “strongly	
  agree”	
  (5).	
  Cronbach’s	
  alpha	
  
suggests	
  that	
  both	
  times	
  this	
  scale	
  was	
  administered,	
  it	
  was	
  reliable	
  (α	
  =	
  .811,	
  day	
  1	
  and	
  
.835,	
  day	
  2).	
  
	
  
Demographics	
  Questionnaire	
  and	
  Cognitive	
  Ability	
  
Please	
  provide	
  as	
  much	
  of	
  the	
  following	
  information	
  as	
  is	
  applicable.	
  It	
  is	
  important	
  to	
  
understand	
  that	
  these	
  scores	
  will	
  be	
  kept	
  confidential	
  and	
  used	
  only	
  for	
  research	
  purposes.	
  
If	
  you	
  do	
  not	
  remember	
  your	
  exam	
  scores,	
  please	
  put	
  a	
  zero	
  in	
  that	
  space.	
  
	
  
Gender:	
  _____	
   (M	
  /	
  F)	
  
	
  
	
  
	
  
	
  
College	
  GPA:	
  _________	
  
Age:	
  _____	
  

	
  

	
  

	
  

	
  

	
  

	
  

SAT	
  score:	
  ____________	
  

Year	
  in	
  College:	
  ____	
   	
  

	
  

	
  

	
  

	
  

ACT	
  score:	
  ____________	
  

Major:	
  __________________________________________________	
  
	
  
Trait	
  Goal	
  Orientation	
  
For	
  each	
  of	
  the	
  following	
  statements,	
  please	
  indicate	
  how	
  true	
  it	
  is	
  for	
  you	
  on	
  the	
  scale	
  
provided	
  below.	
  
	
  
	
  	
  	
  	
  	
  	
  1	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  2	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  3	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  4	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  5	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  6	
  
Strongly	
  	
  	
  	
  	
  	
  	
  	
  	
  Moderately	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Slightly	
  	
  	
  	
  	
  	
  	
  	
  	
  Slightly	
  	
  	
  	
  	
  	
  	
  	
  	
  Moderately	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Strongly	
  
Disagree	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Disagree	
  
	
  	
  Disagree	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Agree	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Agree	
  	
  
	
  	
  	
  	
  Agree	
  
	
  
Goal	
  Orientation	
  Learning	
  
1. I	
  am	
  willing	
  to	
  take	
  on	
  challenges	
  that	
  I	
  can	
  learn	
  a	
  lot	
  from.	
  
2. I	
  often	
  look	
  for	
  opportunities	
  to	
  develop	
  new	
  skills	
  and	
  knowledge.	
  
3. I	
  enjoy	
  challenging	
  and	
  difficult	
  activities	
  where	
  I’ll	
  learn	
  new	
  skills.	
  
4. For	
  me,	
  development	
  of	
  my	
  abilities	
  is	
  important	
  enough	
  to	
  take	
  risks.	
  
Goal	
  Orientation	
  Prove:	
  
1. I	
  prefer	
  to	
  do	
  things	
  that	
  require	
  a	
  high	
  level	
  of	
  ability	
  and	
  talent.	
  
2. I’m	
  concerned	
  with	
  showing	
  that	
  I	
  can	
  perform	
  better	
  than	
  my	
  peers.	
  
3. I	
  try	
  to	
  figure	
  out	
  what	
  it	
  takes	
  to	
  prove	
  my	
  ability	
  to	
  others.	
  
4. I	
  enjoy	
  it	
  when	
  others	
  are	
  aware	
  of	
  how	
  well	
  I	
  am	
  doing.	
  
5. I	
  prefer	
  to	
  participate	
  in	
  things	
  where	
  I	
  can	
  prove	
  my	
  ability	
  to	
  others.	
  
	
  

169	
  

	
  
Goal	
  Orientation	
  Avoidance:	
  
1. I	
  would	
  avoid	
  taking	
  on	
  a	
  new	
  task	
  if	
  there	
  was	
  a	
  chance	
  that	
  I	
  would	
  appear	
  rather	
  
incompetent	
  to	
  others.	
  
2. Avoiding	
  a	
  show	
  of	
  low	
  ability	
  is	
  more	
  important	
  to	
  me	
  than	
  learning	
  a	
  new	
  skill.	
  
3. I’m	
  concerned	
  about	
  taking	
  on	
  a	
  task	
  if	
  my	
  performance	
  would	
  reveal	
  that	
  I	
  had	
  low	
  
ability.	
  
4. I	
  prefer	
  to	
  avoid	
  situations	
  where	
  I	
  might	
  perform	
  poorly.	
  
	
  
	
  
	
  
	
  
State	
  Goal	
  Orientation	
  
For	
  each	
  of	
  the	
  following	
  statements,	
  please	
  indicate	
  how	
  true	
  it	
  is	
  for	
  you	
  with	
  regard	
  to	
  
how	
  your	
  approach	
  this	
  task	
  on	
  the	
  scale	
  provided	
  below.	
  
	
  
	
  	
  	
  	
  	
  	
  1	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  2	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  3	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  4	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  5	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  
Strongly	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Neither	
  Agree	
  	
  	
  
	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Strongly	
  
Disagree	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Disagree	
  
	
  	
  	
  Not	
  Disagree	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Agree	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Agree	
   	
  
	
  
Goal	
  Orientation	
  Learning	
  
1. I	
  prefer	
  to	
  work	
  on	
  aspects	
  of	
  this	
  task	
  that	
  force	
  me	
  to	
  learn	
  new	
  things.	
  
2. I	
  am	
  willing	
  to	
  work	
  on	
  challenging	
  aspects	
  of	
  this	
  task	
  that	
  I	
  can	
  learn	
  a	
  lot	
  from.	
  
3. The	
  opportunity	
  to	
  learn	
  new	
  things	
  about	
  this	
  task	
  is	
  important	
  to	
  me.	
  
4. The	
  opportunity	
  to	
  work	
  on	
  challenging	
  aspects	
  of	
  this	
  task	
  is	
  important	
  to	
  me.	
  
5. On	
  this	
  task,	
  my	
  goal	
  is	
  to	
  learn	
  the	
  task	
  as	
  well	
  as	
  I	
  can.	
  
Goal	
  Orientation	
  Prove:	
  
1. It	
  is	
  important	
  to	
  me	
  to	
  perform	
  better	
  than	
  others	
  in	
  this	
  task.	
  
2. It	
  is	
  important	
  to	
  me	
  to	
  impress	
  others	
  by	
  doing	
  a	
  good	
  job	
  on	
  this	
  task.	
  
3. I	
  was	
  the	
  experimenters	
  and	
  other	
  students	
  to	
  recognize	
  that	
  I	
  am	
  one	
  of	
  the	
  best	
  on	
  
this	
  task.	
  
4. I	
  want	
  to	
  show	
  myself	
  how	
  good	
  I	
  am	
  on	
  this	
  task.	
  
5. On	
  this	
  task,	
  my	
  goal	
  is	
  to	
  perform	
  well.	
  
Goal	
  Orientation	
  Avoidance:	
  
1. On	
  this	
  task,	
  I	
  would	
  like	
  to	
  hide	
  from	
  others	
  that	
  they	
  are	
  better	
  than	
  me.	
  
2. On	
  this	
  task,	
  I	
  would	
  like	
  to	
  avoid	
  situations	
  where	
  I	
  might	
  demonstrate	
  poor	
  
performance	
  to	
  myself.	
  
3. On	
  this	
  task,	
  I	
  would	
  like	
  to	
  avoid	
  discovering	
  that	
  others	
  are	
  better	
  than	
  me.	
  
4. I	
  am	
  reluctant	
  to	
  ask	
  questions	
  about	
  this	
  task	
  because	
  others	
  may	
  think	
  I’m	
  
incompetent.	
  
5. On	
  this	
  task,	
  my	
  goal	
  is	
  to	
  avoid	
  performing	
  poorly.	
  
	
  
	
  
	
  
	
  

	
  

170	
  

	
  
I-­‐ADAPT	
  
Please	
  use	
  the	
  scale	
  below	
  to	
  answer	
  this	
  survey	
  that	
  will	
  ask	
  you	
  a	
  number	
  of	
  questions	
  
about	
  your	
  preferences,	
  styles,	
  and	
  habits	
  at	
  work.	
  Read	
  each	
  statement	
  carefully.	
  There	
  
are	
  no	
  right	
  or	
  wrong	
  answers.	
  	
  
	
  
	
  
	
  	
  	
  	
  	
  	
  1	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  2	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  3	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  4	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  5	
  
Strongly	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Neither	
  Agree	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Strongly	
  
Disagree	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Disagree	
  
Nor	
  Disagree	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Agree	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Agree	
  
	
  
	
  
1. I	
  am	
  able	
  to	
  maintain	
  focus	
  during	
  emergencies	
  
2. I	
  enjoy	
  learning	
  about	
  cultures	
  other	
  than	
  my	
  own	
  
3.
4.
5.
6.
7.

I	
  usually	
  over-­‐react	
  to	
  stressful	
  news	
  
I	
  believe	
  it	
  is	
  important	
  to	
  be	
  flexible	
  in	
  dealing	
  with	
  others	
  
I	
  take	
  responsibility	
  for	
  acquiring	
  new	
  skills	
  
I	
  work	
  well	
  with	
  diverse	
  others	
  
I	
  tend	
  to	
  be	
  able	
  to	
  read	
  others	
  and	
  understand	
  how	
  they	
  are	
  feeling	
  at	
  any	
  
particular	
  moment	
  
8. I	
  am	
  adept	
  at	
  using	
  my	
  body	
  to	
  complete	
  relevant	
  tasks	
  
9. In	
  an	
  emergency	
  situation,	
  I	
  can	
  put	
  aside	
  emotional	
  feelings	
  to	
  handle	
  important	
  
tasks	
  
10. I	
  see	
  connections	
  between	
  seemingly	
  unrelated	
  information	
  
11. I	
  enjoy	
  learning	
  new	
  approaches	
  for	
  conducting	
  work	
  
12. I	
  think	
  clearly	
  in	
  times	
  of	
  urgency	
  
13. I	
  utilize	
  my	
  muscular	
  strength	
  well	
  
14. It	
  is	
  important	
  to	
  me	
  that	
  I	
  respect	
  others’	
  culture	
  
15. I	
  feel	
  unequipped	
  to	
  deal	
  with	
  too	
  much	
  stress	
  
16. I	
  am	
  good	
  at	
  developing	
  unique	
  analyses	
  for	
  complex	
  problems	
  
17. I	
  am	
  able	
  to	
  be	
  objective	
  during	
  emergencies	
  
18. My	
  insight	
  helps	
  me	
  to	
  work	
  effectively	
  with	
  others	
  
19. I	
  enjoy	
  the	
  variety	
  and	
  learning	
  experiences	
  that	
  come	
  from	
  working	
  with	
  people	
  of	
  
different	
  backgrounds	
  
20. I	
  can	
  only	
  work	
  in	
  an	
  orderly	
  environment	
  
21. I	
  am	
  easily	
  rattled	
  when	
  my	
  schedule	
  is	
  too	
  full	
  
22. I	
  usually	
  set	
  up	
  and	
  take	
  action	
  during	
  a	
  crisis	
  
23. I	
  need	
  for	
  things	
  to	
  be	
  ‘black	
  and	
  white’	
  

	
  

171	
  

	
  
24. I	
  am	
  an	
  innovative	
  person	
  
25. I	
  feel	
  comfortable	
  interacting	
  with	
  others	
  who	
  have	
  different	
  values	
  and	
  customs	
  
26. If	
  my	
  environment	
  is	
  not	
  comfortable	
  (e.g.,	
  cleanliness)	
  I	
  cannot	
  perform	
  well	
  
27. I	
  make	
  excellent	
  decisions	
  in	
  times	
  of	
  crisis	
  
28. I	
  become	
  frustrated	
  when	
  things	
  are	
  unpredictable	
  
29. I	
  am	
  able	
  to	
  make	
  effective	
  decisions	
  without	
  all	
  relevant	
  information	
  
30. I	
  am	
  an	
  open-­‐minded	
  person	
  in	
  dealing	
  with	
  others	
  
31. I	
  take	
  action	
  to	
  improve	
  work	
  performance	
  deficiencies	
  
32. I	
  usually	
  am	
  stressed	
  when	
  I	
  have	
  a	
  large	
  workload	
  
33. I	
  am	
  perceptive	
  of	
  others	
  and	
  use	
  that	
  knowledge	
  in	
  interactions	
  	
  
34. I	
  often	
  learn	
  new	
  information	
  and	
  skills	
  to	
  stay	
  at	
  the	
  forefront	
  of	
  my	
  profession	
  
35. I	
  often	
  cry	
  or	
  get	
  angry	
  when	
  I	
  am	
  under	
  a	
  great	
  deal	
  of	
  stress	
  
36. When	
  resources	
  are	
  insufficient,	
  I	
  thrive	
  on	
  developing	
  innovative	
  solutions	
  
37. I	
  am	
  able	
  to	
  look	
  at	
  problems	
  from	
  a	
  multitude	
  of	
  angles	
  
38. I	
  quickly	
  learn	
  new	
  methods	
  to	
  solve	
  problems	
  
39. I	
  ten	
  to	
  perform	
  best	
  in	
  stable	
  situations	
  and	
  environments	
  
40. When	
  something	
  unexpected	
  happens,	
  I	
  readily	
  change	
  hears	
  in	
  response	
  
41. I	
  would	
  quit	
  my	
  job	
  if	
  it	
  required	
  me	
  to	
  be	
  physically	
  stronger	
  
42. I	
  try	
  to	
  be	
  flexible	
  when	
  dealing	
  with	
  others	
  
43. I	
  can	
  adapt	
  to	
  changing	
  situations	
  
44. I	
  train	
  to	
  keep	
  my	
  work	
  skills	
  and	
  knowledge	
  current	
  
45. I	
  physically	
  push	
  myself	
  to	
  complete	
  important	
  tasks	
  
46. I	
  am	
  continually	
  learning	
  new	
  skills	
  for	
  my	
  job	
  
47. I	
  perform	
  well	
  in	
  uncertain	
  situations	
  
48. I	
  can	
  work	
  effectively	
  even	
  when	
  I	
  am	
  tired	
  
49. I	
  take	
  responsibility	
  for	
  staying	
  current	
  in	
  my	
  profession	
  
50. I	
  adapt	
  my	
  behavior	
  to	
  get	
  along	
  with	
  others	
  
51. I	
  cannot	
  work	
  well	
  if	
  it	
  is	
  too	
  hot	
  or	
  too	
  cold	
  
52. I	
  easily	
  respond	
  to	
  changing	
  conditions	
  
53. I	
  try	
  to	
  learn	
  new	
  skills	
  for	
  my	
  job	
  before	
  they	
  are	
  needed	
  
54. I	
  can	
  adjust	
  my	
  plans	
  to	
  changing	
  conditions	
  
55. I	
  keep	
  working	
  even	
  when	
  I	
  am	
  physically	
  exhausted	
  

	
  

172	
  

	
  
APPENDIX	
  E:	
  
IRB	
  Documentation	
  
	
  
Figure	
  15	
  
IRB	
  Informed	
  Consent	
  and	
  Approval	
  Documents	
  

	
  
	
  

173	
  

	
  
Figure	
  15	
  (cont’d)	
  

	
  
	
  

174	
  

	
  
Figure	
  15	
  (cont’d)	
  
	
  

	
  
	
  

175	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
REFERENCES

	
  

176	
  

	
  
REFERENCES	
  
	
  
	
  
	
  
Anderson,	
  J.	
  R.	
  (1983).	
  The	
  architecture	
  of	
  cognition.	
  Cambridge,	
  MA:	
  Harvard	
  University	
  
Press.	
  
	
  
Ashford,	
  S.	
  J.	
  &	
  Cummings,	
  L.	
  L.	
  (1983).	
  Feedback	
  as	
  an	
  individual	
  resource:	
  Personal	
  
strategies	
  of	
  creating	
  information.	
  Organizational	
  Behavior	
  and	
  Human	
  Performance,	
  
32,	
  370-­‐398.	
  
	
  
Baard,	
  S.	
  K.	
  (2013).	
  An	
  insight	
  into	
  adaptation:	
  Self-­‐regulatory	
  mechanisms	
  as	
  a	
  driver	
  of	
  
adaptive	
  performance	
  over	
  time.	
  (Master’s	
  Thesis,	
  Michigan	
  State	
  University,	
  United	
  
States,	
  MI).	
  Retrieved	
  from	
  ProQuest	
  Dissertations	
  and	
  Theses	
  database.	
  (UMI	
  No.	
  
1538915).	
  
	
  
Baard,	
  S.	
  K.,	
  Rench,	
  T.	
  A.	
  &	
  Kozlowski,	
  S.	
  W.	
  J.	
  (2014).	
  Performance	
  adaptation:	
  A	
  theoretical	
  
integration	
  and	
  review.	
  Journal	
  of	
  Management,	
  40(1),	
  44-­‐95.	
  
	
  
Bandura,	
  A.	
  (1986).	
  Social	
  foundations	
  of	
  thought	
  and	
  action:	
  A	
  social	
  cognitive	
  theory.	
  
Englewood	
  Cliffs,	
  NJ:	
  Prentice	
  Hall.	
  
	
  
Bandura,	
  A.	
  (1991).	
  Social	
  cognitive	
  theory	
  of	
  self-­‐regulation.	
  Organizational	
  Behavior	
  and	
  
Human	
  Decision	
  Processes,	
  50,	
  248-­‐287.	
  
	
  
Bandura,	
  A.	
  (1993).	
  Perceived	
  self-­‐efficacy	
  in	
  cognitive	
  development	
  and	
  functioning.	
  
Educational	
  Psychologies,	
  28(2),	
  117-­‐148.	
  
	
  
Bandura,	
  A.	
  &	
  Cervone,	
  D.	
  (1983).	
  Self-­‐evaluative	
  and	
  self-­‐efficacy	
  mechanisms	
  governing	
  
the	
  motivational	
  effects	
  of	
  goal	
  systems.	
  Journal	
  of	
  Personality	
  and	
  Social	
  Psychology,	
  
45(5),	
  1017-­‐1028.	
  	
  
	
  
Bandura,	
  A.	
  &	
  Locke,	
  E.	
  A.	
  (2003).	
  Negative	
  self-­‐efficacy	
  and	
  goal	
  effects	
  revisited.	
  Journal	
  of	
  
Applied	
  Psychology,	
  88(1),	
  87-­‐99.	
  
	
  
Beer,	
  M.,	
  Eisenstat,	
  R.	
  A.	
  &	
  Spector,	
  B.	
  (1990).	
  Why	
  change	
  programs	
  don’t	
  produce	
  change?,	
  
Harvard	
  Business	
  Review,	
  Nov-­‐Dec,	
  158-­‐166.	
  
	
  
Bell,	
  B.	
  S.	
  (2002).	
  An	
  examination	
  of	
  the	
  instructional,	
  motivational,	
  and	
  emotional	
  elements	
  
of	
  error	
  training.	
  (Doctoral	
  Dissertation,	
  Michigan	
  State	
  University,	
  United	
  States,	
  
MI).	
  Retrieved	
  from	
  ProQuest	
  Dissertations	
  and	
  Theses	
  database.	
  (UMI	
  No.	
  
3064202).	
  
	
  
Bell,	
  B.	
  S.,	
  &	
  Kozlowski,	
  S.	
  W.	
  J.	
  (2002a).	
  Adaptive	
  guidance:	
  Enhancing	
  self-­‐regulation,	
  
knowledge	
  and	
  performance	
  in	
  technology-­‐based	
  training.	
  Personnel	
  Psychology,	
  55:	
  
267-­‐306.	
  
	
  

177	
  

	
  
Bell,	
  B.	
  S.	
  &	
  Kozlowski,	
  S.	
  W.	
  J.	
  (2002b).	
  Goal	
  orientation	
  and	
  ability:	
  Interactive	
  effects	
  on	
  
self-­‐efficacy,	
  performance	
  and	
  knowledge.	
  Journal	
  of	
  Applied	
  Psychology,	
  87:	
  497-­‐
505.	
  
	
  
Bell,	
  B.	
  S.,	
  &	
  Kozlowski,	
  S.	
  W.	
  J.	
  (2008).	
  Active	
  learning:	
  Effects	
  of	
  core	
  training	
  design	
  
elements	
  on	
  self-­‐regulatory	
  processes,	
  learning,	
  and	
  adaptability.	
  Journal	
  of	
  Applied	
  
Psychology,	
  93:	
  296-­‐316.	
  
	
  
Bell,	
  B.	
  S.,	
  &	
  Kozlowski,	
  S.	
  W.	
  J.	
  (2010).	
  Toward	
  a	
  theory	
  of	
  learning	
  centered	
  training	
  
design:	
  An	
  integrative	
  framework	
  of	
  active	
  learning.	
  In	
  S.	
  W.	
  J.	
  Kozlowski	
  &	
  E.	
  Salas	
  
(Eds.),	
  Learning,	
  Training,	
  and	
  Development	
  in	
  Organizations	
  (pp.	
  261-­‐298).	
  New	
  
York,	
  NY:	
  Routledge	
  Academic.	
  
	
  
Blau,	
  G.	
  (1993).	
  Operationalizing	
  direction	
  and	
  level	
  of	
  effort	
  and	
  testing	
  their	
  relationships	
  
to	
  individual	
  job	
  performance.	
  Organizational	
  Behavior	
  and	
  Human	
  Decision	
  
Processes,	
  55(1),	
  152-­‐170.	
  
	
  
Bröder,	
  A.,	
  &	
  Schiffer,	
  S.	
  (2006).	
  Adaptive	
  flexibility	
  and	
  maladaptive	
  routines	
  in	
  selecting	
  
fast	
  and	
  frugal	
  decision	
  strategies.	
  Journal	
  of	
  Experimental	
  Psychology:	
  Learning,	
  
Memory,	
  and	
  Cognition,	
  32:	
  904-­‐918.	
  
	
  
Brown,	
  T.	
  C.	
  (2005).	
  Effectiveness	
  of	
  distal	
  and	
  proximal	
  goals	
  as	
  transfer-­‐of-­‐training	
  
interventions:	
  A	
  field	
  experiment.	
  Human	
  Resource	
  Development	
  Quarterly,	
  16:	
  369-­‐
387.	
  
	
  
Burke,	
  C.	
  S.,	
  Stagl,	
  K.	
  C.,	
  Salas,	
  E.,	
  Pierce,	
  L.	
  &Kendall,	
  D.	
  (2006).	
  Understanding	
  team	
  
adaptation:	
  A	
  conceptual	
  analysis	
  and	
  model.	
  	
  Journal	
  of	
  Applied	
  Psychology,	
  9:	
  1189-­‐
1207.	
  
	
  
Campbell,	
  J.	
  P.	
  (1990).	
  Modeling	
  the	
  performance	
  prediction	
  problem	
  in	
  industrial	
  and	
  
organizational	
  psychology.	
  In	
  M.	
  D.	
  Dunnette	
  &	
  L.	
  M.	
  Hough	
  (Eds.),	
  Handbook	
  of	
  
Industrial	
  and	
  Organizational	
  Psychology	
  (pp.	
  687-­‐732).	
  Palo	
  Alto,	
  CA:	
  Consulting	
  
Psychologists	
  Press,	
  Inc.	
  
	
  
Campbell,	
  J.P.	
  (2012).	
  Behavior,	
  performance,	
  and	
  effectiveness	
  in	
  the	
  21st	
  century.	
  In	
  S.	
  W.	
  
J.	
  Kozlowski	
  (Ed.)	
  The	
  Oxford	
  Handbook	
  of	
  Organizational	
  Psychology	
  (159-­‐195)	
  New	
  
York:	
  Oxford.	
  
	
  
Carver,	
  C.	
  S.,	
  &	
  Scheier,	
  M.	
  F.	
  (1982).	
  Control	
  theory:	
  A	
  useful	
  conceptual	
  framework	
  for	
  
personality-­‐social,	
  clinical,	
  and	
  health	
  psychology.	
  Psychological	
  Bulletin,	
  92(2),	
  111-­‐
135.	
  
	
  
Carver,	
  C.	
  S.,	
  &	
  Scheier,	
  M.	
  F.	
  (1998).	
  On	
  the	
  self-­‐regulation	
  of	
  behavior.	
  New	
  York:	
  
Cambridge	
  University	
  Press.	
  
	
  
Caldwell,	
  D.	
  F.,	
  &	
  O’Reilly,	
  C.	
  A.	
  (1982).	
  Boundary	
  spanning	
  and	
  individual	
  performance:	
  
	
  

178	
  

	
  
The	
  impact	
  of	
  self-­‐monitoring.	
  Journal	
  of	
  Applied	
  Psychology,	
  67(1),	
  124-­‐127.	
  
	
  
Cannon-­‐Bowers,	
  J.	
  A.,	
  Rhodenizer,	
  L.,	
  Salas,	
  E.	
  &	
  Bowers,	
  C.	
  A.	
  (1998).	
  A	
  framework	
  for	
  
understanding	
  pre-­‐practice	
  conditions	
  and	
  their	
  impact	
  on	
  learning.	
  Personnel	
  
Psychology,	
  51,	
  291-­‐320.	
  
	
  
Chen,	
  G.,	
  Thomas,	
  B.	
  &	
  Wallace,	
  J.	
  C.	
  (2005).	
  A	
  multilevel	
  examination	
  of	
  the	
  relationships	
  
among	
  training	
  outcomes,	
  mediating	
  regulatory	
  processes,	
  and	
  adaptive	
  
performance.	
  Journal	
  of	
  Applied	
  Psychology,	
  90:	
  827-­‐841.	
  
	
  
Clause,	
  C.	
  S.,	
  Delbridge,	
  K.,	
  Schmitt,	
  N.,	
  Chan,	
  D.,	
  &	
  Jennings,	
  D.	
  (2001).	
  Test	
  preparation	
  
activities	
  and	
  employment	
  test	
  performance.	
  Human	
  Performance,	
  14(2),	
  149-­‐167.	
  
	
  
Cohen,	
  J.	
  (1988),	
  Statistical	
  Power	
  Analysis	
  for	
  the	
  Behavioral	
  Sciences,	
  2nd	
  Edition.	
  
Hillsdale:	
  Lawrence	
  Erlbaum.	
  
	
  
Converse,	
  P.D.,	
  Piccone,	
  K.,	
  Lockamy,	
  C.	
  N.,	
  Miloslavic,	
  S.	
  A.,	
  Mysiak,	
  K.	
  &	
  Pathak,	
  J.	
  (2014).	
  
The	
  influence	
  of	
  perceived	
  accountability	
  and	
  outcome	
  interdependence	
  on	
  goals	
  
and	
  effort.	
  Journal	
  of	
  Applied	
  Social	
  Psychology,	
  44,	
  210-­‐219.	
  
	
  
Cron,	
  W.	
  L.,	
  Slocum,	
  J.	
  W.,	
  VandeWalle,	
  D.	
  &	
  Fu,	
  Q.	
  (2005).	
  The	
  role	
  of	
  goal	
  orientation	
  on	
  
negative	
  emotions	
  and	
  goal	
  setting	
  when	
  initial	
  performance	
  falls	
  short	
  of	
  one's	
  
performance	
  goal.	
  Human	
  Performance,	
  18(1),	
  55-­‐80.	
  
	
  
DeRue,	
  D.	
  S.,	
  Hollenbeck,	
  J.	
  R.,	
  Johnson,	
  M.	
  D.,	
  Ilgen,	
  D.	
  R.,	
  &	
  Jundt,	
  D.	
  K.	
  (2008).	
  How	
  
different	
  team	
  downsizing	
  approaches	
  influence	
  team-­‐level	
  adaptation	
  and	
  
performance.	
  Academy	
  of	
  Management	
  Journal,	
  51:	
  182-­‐196.	
  
	
  
DeShon,	
  R.	
  P.	
  (2012).	
  Multivariate	
  dynamics	
  in	
  organizational	
  science.	
  In	
  S.	
  W.	
  J.	
  Kozlowski	
  
(Ed.),	
  The	
  Oxford	
  handbook	
  of	
  organizational	
  psychology.	
  New	
  York:	
  Oxford	
  
University	
  Press.	
  
	
  
DeShon,	
  R.	
  P.,	
  Kozlowski,	
  S.	
  W.	
  J.,	
  Schmidt,	
  A.	
  M.,	
  Milner,	
  K.	
  R.,	
  &	
  Wiechmann,	
  D.	
  (2004).	
  A	
  
multiple-­‐goal,	
  multilevel	
  model	
  of	
  feedback	
  effects	
  on	
  the	
  regulation	
  of	
  individual	
  
and	
  team	
  performance.	
  Journal	
  of	
  Applied	
  Psychology,	
  89(6),	
  1035-­‐1056.	
  
	
  
Dormann,	
  T.,	
  &	
  Frese,	
  M.	
  (1994).	
  Error	
  management	
  training:	
  Replication	
  and	
  the	
  function	
  
of	
  exploratory	
  behavior.	
  International	
  Journal	
  of	
  Human-­‐Computer	
  Interaction,	
  6,	
  
365–372.	
  
	
  
Drach-­‐Zahavy,	
  A.,	
  &	
  Somech,	
  A.	
  (1999).	
  Constructive	
  thinking:	
  A	
  complex	
  coping	
  variable	
  
that	
  distinctively	
  influences	
  the	
  effectiveness	
  of	
  specific	
  difficult	
  goals.	
  Personality	
  
and	
  Individual	
  Differences,	
  27:	
  9-­‐984.	
  
	
  
Driskell,	
  J.	
  E.,	
  Goodwin,	
  G.	
  F.,	
  Salas,	
  E.,	
  &	
  O'Shea,	
  P.	
  G.	
  (2006).	
  What	
  makes	
  a	
  good	
  team	
  
player?	
  Personality	
  and	
  team	
  effectiveness.	
  Group	
  Dynamics:	
  Theory,	
  Research,	
  and	
  
	
  

179	
  

	
  
Practice,	
  10:	
  249-­‐271.	
  
	
  
Edmondson,	
  A.	
  (1999).	
  Psychological	
  safety	
  and	
  learning	
  behavior	
  in	
  work	
  teams.	
  
Administrative	
  Science	
  Quarterly,	
  44,	
  350-­‐383.	
  
	
  
Ertmer,	
  P.	
  A.,	
  &	
  Newby,	
  T.	
  J.	
  (1996).	
  The	
  expert	
  learner:	
  Strategic,	
  self-­‐regulated,	
  and	
  
reflective.	
  Instructional	
  Science,	
  24,	
  1-­‐24.	
  
	
  
Flavel,	
  J.	
  H.	
  (1979).	
  Metacognition	
  and	
  cognitive	
  monitoring:	
  A	
  new	
  area	
  of	
  cognitive-­‐
developmental	
  inquiry.	
  American	
  Psychology,	
  34(10),	
  906-­‐911.	
  
	
  
Ford,	
  J.	
  K.,	
  Smith,	
  E.	
  M.,	
  Weissbein,	
  D.	
  A.,	
  Gully,	
  S.	
  M.,	
  &	
  Salas,	
  E.	
  (1998).	
  Relationships	
  of	
  goal	
  
orientation,	
  metacognitive	
  activity,	
  and	
  practice	
  strategies	
  with	
  learning	
  outcomes	
  
and	
  transfer.	
  Journal	
  of	
  Applied	
  Psychology,	
  83:	
  218-­‐233.	
  
	
  
Fisher,	
  S.	
  L.	
  &	
  Ford,	
  J.	
  K.	
  (1998).	
  Differential	
  effects	
  of	
  learner	
  effort	
  and	
  goal	
  orientation	
  on	
  
two	
  learning	
  outcomes.	
  Personnel	
  Psychology,	
  51,	
  397-­‐420.	
  
	
  
Frese,	
  M.,	
  Brodbeck,	
  F.	
  C.,	
  Heinbokel,	
  T.,	
  Mooser,	
  C.,	
  Schleiffenbaum,	
  E.,	
  &	
  Thiemann,	
  P.	
  
(1991).	
  Errors	
  in	
  training	
  computer	
  skills:	
  On	
  the	
  positive	
  function	
  of	
  errors.	
  
Human–Computer	
  Interaction,	
  6:	
  77–93.	
  
	
  
Ghiselli,	
  E.	
  E.	
  (1951).	
  New	
  ideas	
  in	
  industrial	
  psychology.	
  Journal	
  of	
  Applied	
  Psychology,	
  
35(4),	
  229-­‐235.	
  
	
  
Good,	
  D.	
  &	
  Michel,	
  E.	
  J.	
  (2013).	
  Individual	
  ambidexterity:	
  Exploring	
  and	
  exploiting	
  in	
  
dynamic	
  contexts.	
  The	
  Journal	
  of	
  Psychology,	
  147(5),	
  435-­‐453.	
  
	
  
Griffin,	
  B.,	
  &	
  Hesketh,	
  B.	
  (2003).	
  Adaptable	
  behaviours	
  for	
  successful	
  work	
  and	
  career	
  
adjustment?	
  Australian	
  Journal	
  of	
  Psychology,	
  55:	
  65-­‐73.	
  
	
  
Griffin,	
  B.,	
  &	
  Hesketh,	
  B.	
  (2004).	
  Why	
  openness	
  to	
  experience	
  is	
  not	
  a	
  good	
  predictor	
  of	
  job	
  
performance.	
  International	
  Journal	
  of	
  Selection	
  and	
  Assessment,	
  12:	
  243-­‐251.	
  
	
  
Griffin,	
  B.	
  &	
  Hesketh,	
  B.	
  (2005).	
  Are	
  conscientious	
  workers	
  adaptable?	
  Australian	
  Journal	
  of	
  
management,	
  30:	
  245-­‐259.	
  
	
  
Gully,	
  S.	
  M.,	
  Payne,	
  S.	
  C.,	
  Koles,	
  K.	
  L.	
  K.,	
  &	
  Whiteman,	
  J.-­‐A.	
  K.	
  (2002).	
  The	
  impact	
  of	
  error	
  
training	
  and	
  individual	
  differences	
  on	
  training	
  outcomes:	
  An	
  attribute-­‐treatment	
  
interaction	
  perspective.	
  Journal	
  of	
  Applied	
  Psychology,	
  87(1),	
  143-­‐155.	
  
	
  
Hattie,	
  J.	
  &	
  Timperley,	
  H.	
  (2007).	
  The	
  power	
  of	
  feedback.	
  Review	
  of	
  Educational	
  Research,	
  
77(1),	
  81-­‐112.	
  
	
  
Holladay,	
  C.	
  L.	
  &	
  Quiñones,	
  M.	
  A.	
  (2003).	
  Practice	
  variability	
  and	
  transfer	
  of	
  training:	
  The	
  
role	
  of	
  self-­‐efficacy	
  generality.	
  Journal	
  of	
  Applied	
  Psychology,	
  88:	
  1094-­‐1103.	
  
	
  

180	
  

	
  
	
  
Horvath,	
  M.,	
  Scheu,	
  C.	
  R.,	
  &	
  DeShon,	
  R.	
  P.	
  (2001).	
  Goal	
  orientation:	
  Integration	
  theory	
  and	
  
measurement.	
  Paper	
  presented	
  at	
  the	
  16th	
  Annual	
  Conference	
  of	
  the	
  Society	
  for	
  
Industrial	
  and	
  Organizational	
  Psychology,	
  San	
  Diego:	
  CA.	
  
	
  
Ilies,	
  R.	
  (2003).	
  A	
  dynamic	
  multilevel	
  model	
  of	
  task	
  motivation	
  linking	
  personality,	
  affective	
  
reactions	
  to	
  feedback	
  and	
  self-­‐regulation.	
  (Unpublished	
  doctoral	
  dissertation.)	
  
University	
  of	
  Florida,	
  Gainesville,	
  FL.	
  
	
  
Ivancic,	
  K.	
  &	
  Hesketh,	
  B.	
  (2000).	
  Learning	
  from	
  errors	
  in	
  a	
  driving	
  simulation:	
  Effects	
  on	
  
driving	
  skill	
  and	
  self-­‐confidence.	
  Ergonomics,	
  43:	
  1966-­‐1984.	
  
	
  
Johnson,	
  M.	
  D.,	
  Humphrey,	
  S.	
  E.,	
  Ilgen,	
  D.	
  R.,	
  Jundt,	
  D.	
  &	
  Meyer,	
  C.	
  J.	
  (2006).	
  Cutthroat	
  
cooperation:	
  Asymmetrical	
  adaptation	
  to	
  changes	
  in	
  team	
  reward	
  structures.	
  
Academy	
  of	
  Management	
  Journal,	
  49:	
  103-­‐119.	
  
	
  
Jundt,	
  D.	
  (2009).	
  Adaptability	
  from	
  a	
  process	
  perspective:	
  Examining	
  the	
  effects	
  of	
  task	
  
change	
  type	
  and	
  a	
  metacognitive	
  intervention	
  on	
  adaptive	
  performance.	
  (Doctoral	
  
Dissertation,	
  Michigan	
  State	
  University,	
  United	
  States,	
  MI).	
  Retrieved	
  from	
  ProQuest	
  
Dissertations	
  and	
  Theses	
  database.	
  (UMI	
  No.	
  1432189).	
  
	
  
Kanfer,	
  R.	
  &	
  Ackerman,	
  P.	
  L.	
  (1989).	
  Motivation	
  and	
  cognitive	
  abilities:	
  An	
  
integrative/aptitude-­‐treatment	
  interaction	
  approach	
  to	
  skill	
  acquisition.	
  Journal	
  of	
  
Applied	
  Psychology	
  Monograph,	
  74(4),	
  657-­‐690.	
  
	
  
Karoly,	
  P.	
  (1993).	
  Mechanisms	
  of	
  self-­‐regulation:	
  A	
  systems	
  review.	
  Annual	
  Review	
  of	
  
Psychology,	
  44,	
  23-­‐52.	
  
	
  
Keith,	
  N.,	
  &	
  Frese,	
  M.	
  (2005).	
  Self-­‐regulation	
  in	
  error	
  management	
  training:	
  Emotion	
  
control	
  and	
  metacognition	
  as	
  mediators	
  of	
  performance	
  effects.	
  Journal	
  of	
  Applied	
  
Psychology,	
  90:	
  677-­‐691.	
  
	
  
Klein,	
  H.	
  J.	
  (1989).	
  An	
  integrated	
  control	
  theory	
  model	
  of	
  work	
  motivation.	
  The	
  Academy	
  of	
  
Management	
  Review,	
  14(2),	
  150-­‐172.	
  
	
  
Kluger,	
  A.	
  N.	
  &	
  DeNisi,	
  A.	
  (1996).	
  The	
  effects	
  of	
  feedback	
  interventions	
  on	
  performance:	
  A	
  
historical	
  review,	
  a	
  meta-­‐analysis,	
  and	
  a	
  preliminary	
  feedback	
  intervention	
  theory.	
  
Psychological	
  Bulletin,	
  119(2),	
  254-­‐284.	
  
	
  
Komarraju,	
  M.	
  and	
  Nadler,	
  D.	
  (2013).	
  Self-­‐efficacy	
  and	
  academic	
  achievement:	
  Why	
  do	
  
implicit	
  beliefs,	
  goals,	
  and	
  effort	
  regulation	
  matter?	
  Learning	
  and	
  Individual	
  
Differences,	
  25,	
  67-­‐72.	
  
	
  
Kozlowski,	
  S.	
  W.	
  J.,	
  &	
  Bell,	
  B.	
  S.	
  (2006).	
  Disentangling	
  achievement	
  orientation	
  and	
  goal	
  
setting:	
  Effects	
  on	
  self-­‐regulatory	
  processes.	
  Journal	
  of	
  Applied	
  Psychology,	
  91:	
  900-­‐
916.	
  
	
  

181	
  

	
  
	
  
Kozlowski,	
  S.	
  W.	
  J.,	
  Gully,	
  S.	
  M.,	
  Brown,	
  K.	
  G.,	
  Salas,	
  E.,	
  Smith,	
  E.	
  A.,	
  &	
  Nason,	
  E.	
  R.	
  (2001).	
  
Effects	
  of	
  training	
  goals	
  and	
  goal	
  orientation	
  traits	
  on	
  multi-­‐dimensional	
  training	
  
outcomes	
  and	
  performance	
  adaptability.	
  Organizational	
  Behavior	
  and	
  Human	
  
Decision	
  Processes,	
  85:	
  1-­‐31.	
  
	
  
Kozlowski,	
  S.	
  W.	
  J.,	
  Gully,	
  S.	
  M.,	
  Nason,	
  E.	
  R.,	
  &	
  Smith,	
  E.	
  M.	
  (1999).	
  Developing	
  adaptive	
  
teams:	
  A	
  theory	
  of	
  compilation	
  and	
  performance	
  across	
  levels	
  and	
  time.	
  In	
  D.	
  R.	
  
Ilgen	
  &	
  E.	
  D.	
  Pulakos	
  (Eds.),	
  The	
  changing	
  nature	
  of	
  work	
  performance:	
  Implications	
  
for	
  staffing,	
  personnel	
  actions,	
  and	
  development	
  (pp.	
  240-­‐292).	
  San	
  Francisco:	
  Jossey-­‐
Bass.	
  
	
  
Kozlowski,	
  S.	
  W.	
  J.,	
  Gully,	
  S.	
  M.,	
  Salas,	
  E.,	
  &	
  Cannon-­‐Bowers,	
  J.	
  A.	
  (1996).	
  Team	
  leadership	
  
and	
  development:	
  Theory,	
  principles,	
  and	
  guidelines	
  for	
  training	
  leaders	
  and	
  teams.	
  
In	
  M.	
  Beyerlein,	
  D.	
  Johnson,	
  &	
  S.	
  Beyerlein	
  (Eds.),	
  Advances	
  in	
  interdisciplinary	
  
studies	
  of	
  work	
  teams:	
  Team	
  leadership,	
  vol.	
  3:	
  251-­‐289.	
  Greenwich,	
  CT:	
  JAI.	
  
	
  
Kozlowski,	
  S.	
  W.	
  J.,	
  Toney,	
  R.	
  J.,	
  Mullins,	
  M.	
  E.,	
  Weissbein,	
  D.	
  A.	
  Brown,	
  K.	
  G.,	
  &	
  Bell,	
  B.	
  S.	
  
(2001).	
  Developing	
  adaptability:	
  A	
  theory	
  for	
  the	
  design	
  of	
  integrated-­‐embedded	
  
training	
  systems.	
  In	
  E.	
  Salas	
  (Ed.),	
  Advances	
  in	
  human	
  performance	
  and	
  cognitive	
  
engineering	
  research	
  (Vol.	
  1,	
  pp.	
  59-­‐123).	
  Amsterdam:	
  JAI/Elsevier	
  Science.	
  
	
  
Kozlowski,	
  S.	
  W.	
  J.,	
  Watola,	
  D.	
  J.,	
  Jensen,	
  J.	
  M.,	
  Kim,	
  B.	
  H.,	
  &	
  Botero,	
  I.	
  C.	
  (2009).	
  	
  Developing	
  
adaptive	
  teams:	
  A	
  theory	
  of	
  dynamic	
  team	
  leadership.	
  	
  In	
  E.	
  Salas,	
  G.	
  F.	
  Goodwin	
  &	
  C.	
  
S.	
  Burke	
  (Eds.),	
  Team	
  effectiveness	
  in	
  complex	
  organizations:	
  Cross-­‐disciplinary	
  
perspectives	
  and	
  approaches	
  (SIOP	
  Frontier	
  Series,	
  pp.	
  113-­‐155).	
  Mahwah,	
  NJ:	
  LEA.	
  
	
  
Landine,	
  J.	
  &	
  Stewart,	
  J.	
  (1998).	
  Relationship	
  between	
  metacognition,	
  motivation,	
  locus	
  of	
  
control,	
  self-­‐efficacy,	
  and	
  academic	
  achievement.	
  Canadian	
  Journal	
  of	
  Counseling,	
  
32(3),	
  200-­‐212.	
  
	
  
Lang,	
  J.	
  W.	
  B.	
  &	
  Bliese,	
  P.	
  D.	
  (2009).	
  General	
  mental	
  ability	
  and	
  two	
  types	
  of	
  adaptation	
  to	
  
unforeseen	
  change:	
  Applying	
  discontinuous	
  growth	
  models	
  to	
  the	
  task-­‐change	
  
paradigm.	
  Journal	
  of	
  Applied	
  Psychology,	
  94:	
  411-­‐428.	
  
	
  
Latham,	
  G.	
  P.	
  &	
  Locke,	
  E.	
  A.	
  (1991).	
  Self-­‐regulation	
  through	
  goal	
  setting.	
  Organizational	
  
Behavior	
  and	
  Human	
  Decision	
  Processes,	
  50,	
  212-­‐247.	
  
	
  
LePine,	
  J.	
  A.	
  (2003).	
  Team	
  adaptation	
  and	
  postchange	
  performance:	
  Effects	
  of	
  team	
  
composition	
  in	
  terms	
  of	
  members'	
  cognitive	
  ability	
  and	
  personality.	
  Journal	
  of	
  
Applied	
  Psychology,	
  88:	
  27-­‐39.	
  
	
  
LePine,	
  J.	
  A.	
  (2005).	
  Adaptation	
  of	
  teams	
  in	
  response	
  to	
  unforeseen	
  change:	
  Effects	
  of	
  goal	
  
difficulty	
  and	
  team	
  composition	
  in	
  terms	
  of	
  cognitive	
  ability	
  and	
  goal	
  orientation.	
  
Journal	
  of	
  Applied	
  Psychology,	
  90:	
  1153-­‐1167.	
  
	
  
	
  

182	
  

	
  
Lewin,	
  K.	
  (1958)	
  Group	
  decision	
  and	
  social	
  change,	
  in:	
  G.E.	
  Swanson,	
  T.M.	
  Newcomb	
  and	
  
E.L.	
  Nartley	
  (eds),	
  Readings	
  in	
  Social	
  Psychology,	
  pp.	
  197–211	
  (New	
  York:	
  Holt,	
  
Rhinehart	
  and	
  Winston).	
  
	
  
March,	
  J.	
  G.	
  (1991).	
  Exploration	
  and	
  exploitation	
  in	
  organizational	
  learning.	
  Organizational	
  
Science,	
  2(1),	
  71-­‐87.	
  
	
  
McGrath,	
  R.	
  G.	
  (2001).	
  Exploratory	
  learning,	
  innovative	
  capacity,	
  and	
  managerial	
  oversight.	
  
Academy	
  of	
  Management	
  Journal,	
  44(1),	
  118-­‐131.	
  
	
  
Mumford,	
  M.	
  D.,	
  Baughman,	
  W.	
  A.,	
  Threlfall,	
  K.	
  V.,	
  Uhlman,	
  C.	
  E.,	
  &	
  Costanza,	
  D.	
  P.	
  (1993).	
  
Personality,	
  adaptability,	
  and	
  performance:	
  Performance	
  on	
  well-­‐defined	
  and	
  ill-­‐
defined	
  problem-­‐solving	
  tasks.	
  Human	
  Performance,	
  6:	
  241-­‐285.	
  
	
  
Nadler,	
  D.	
  A.	
  (1988).	
  Organizational	
  frame	
  bending:	
  Types	
  of	
  change	
  in	
  the	
  complex	
  
organization.	
  In	
  R.	
  H.	
  Kilmann	
  &	
  T.	
  J.	
  Covin	
  (Eds.),	
  Corporate	
  Transformation:	
  
Revitalizing	
  Organizations	
  for	
  a	
  Competitive	
  World	
  (pp.	
  66-­‐84).	
  San	
  Francisco,	
  CA:	
  
Jossey-­‐Bass	
  Publishers.	
  
	
  
Nadler,	
  D.	
  A.	
  &	
  Tushman,	
  M.	
  L.	
  (1995).	
  Types	
  of	
  organizational	
  change:	
  From	
  incremental	
  
improvement	
  to	
  discontinuous	
  transformation.	
  In.	
  D.	
  A.	
  Nadler,	
  R.	
  B.	
  Shaw,	
  &	
  A.	
  E.	
  
Watson	
  (Eds.),	
  Discontinuous	
  Change:	
  Leading	
  Organizational	
  Transformation	
  (pp.	
  
14-­‐33).	
  San	
  Francisco,	
  CA:	
  Jossey-­‐Bass	
  Publishers.	
  
	
  
Nasin,	
  S.	
  &	
  Sushill	
  (2011).	
  Revisiting	
  organizational	
  change:	
  Exploring	
  the	
  paradox	
  of	
  
managing	
  continuity	
  and	
  change.	
  Journal	
  of	
  Change	
  Management,	
  11(2),	
  185-­‐
206.Niessen,	
  C.,	
  Swarowsky,	
  C.,	
  &	
  Leiz,	
  M.	
  (2010).	
  Age	
  and	
  adaptation	
  to	
  changes	
  in	
  
the	
  workplace.	
  Journal	
  of	
  Managerial	
  Psychology,	
  25:	
  356-­‐383.	
  
	
  
Phillips,	
  J.	
  M.	
  &	
  Gully,	
  S.	
  M.	
  (1997).	
  Role	
  of	
  goal	
  orientation,	
  ability,	
  need	
  for	
  achievement,	
  
and	
  locus	
  of	
  control	
  in	
  the	
  self-­‐efficacy	
  and	
  goal-­‐setting	
  process.	
  Journal	
  of	
  applied	
  
Psychology,	
  82(5),	
  792-­‐802.	
  
	
  
Pintrich,	
  P.	
  R.	
  (2000).	
  Multiple	
  goals,	
  multiple	
  pathways:	
  The	
  role	
  of	
  goal	
  orientation	
  in	
  
learning	
  and	
  achievement.	
  Journal	
  of	
  Educational	
  Psychology,	
  92(3),	
  544-­‐555.	
  
	
  
Pintrich,	
  P.	
  R.	
  &	
  De	
  Groot,	
  E.	
  V.	
  (1990).	
  Motivational	
  and	
  self-­‐regulated	
  learning	
  
components	
  of	
  classroom	
  academic	
  performance.	
  Journal	
  of	
  Educational	
  Psychology,	
  
82(1),	
  33-­‐40.	
  
	
  
Ployhart,	
  R.	
  E.	
  &	
  Bliese,	
  P.	
  D.	
  (2006).	
  Individual	
  ADAPTability	
  (I-­‐ADAPT)	
  theory:	
  
Conceptualizing	
  the	
  antecedents,	
  consequences,	
  and	
  measurement	
  of	
  individual	
  
differences	
  in	
  adaptability.	
  In	
  C.	
  S.	
  Burke,	
  L.	
  Pierce,	
  &	
  E.	
  Salas	
  (Eds.),	
  Understanding	
  
Adaptability:	
  A	
  Prerequisite	
  for	
  Effective	
  Performance	
  within	
  Complex	
  Environments	
  
(pp.	
  3-­‐39).	
  Elsevier	
  Science.	
  
	
  
	
  

183	
  

	
  
Porter,	
  C.	
  O.	
  L.	
  H.,	
  Webb,	
  J.	
  W.	
  &	
  Gogus,	
  C.	
  I.	
  (2010).	
  When	
  goal	
  orientations	
  collide:	
  Effects	
  
of	
  learning	
  and	
  performance	
  orientation	
  on	
  team	
  adaptability	
  in	
  response	
  to	
  
workload	
  imbalance.	
  Journal	
  of	
  Applied	
  Psychology,	
  95:	
  935-­‐943.	
  
	
  
Powers,	
  W.	
  T.	
  (1973).	
  Behavior:	
  The	
  control	
  of	
  perceptions.	
  Chicago:	
  Aldine.	
  
	
  
Pulakos,	
  E.	
  D.,	
  Arad,	
  S.,	
  Donovan,	
  M.	
  A.,	
  &	
  Plamondon,	
  K.	
  E.	
  (2000).	
  Adaptability	
  in	
  the	
  
workplace:	
  Development	
  of	
  a	
  taxonomy	
  of	
  adaptive	
  performance.	
  Journal	
  of	
  Applied	
  
Psychology,	
  85:	
  612-­‐624.	
  
	
  
Pulakos,	
  E.	
  D.,	
  Schmitt,	
  N.,	
  Dorsey,	
  D.	
  W.,	
  Arad,	
  S.,	
  Borman,	
  W.	
  C.,	
  &	
  Hedge,	
  J.	
  W.	
  (2002).	
  
Predicting	
  adaptive	
  performance:	
  Further	
  tests	
  of	
  a	
  model	
  of	
  adaptability.	
  Human	
  
Performance,	
  15:	
  299-­‐323.	
  
	
  
Quinn,	
  J.	
  B.	
  (1978).	
  Strategic	
  change:	
  logical	
  incrementalism,	
  Sloan	
  Management	
  Review,	
  
Fall,	
  7-­‐21.	
  
	
  
Randall,	
  K.	
  R.,	
  Resick,	
  C.	
  J.,	
  &	
  DeChurch,	
  L.	
  A.	
  (2011).	
  Building	
  team	
  adaptive	
  capacity:	
  The	
  
role	
  of	
  sensegiving	
  and	
  team	
  composition.	
  Journal	
  of	
  Applied	
  Psychology,	
  96(3),	
  535-­‐
540.	
  
	
  
Reger,	
  R.	
  K.,	
  Mullane,	
  L.	
  T.,	
  Gustafson,	
  L.	
  T.,	
  &	
  DeMarie,	
  S.	
  M.	
  (1994).	
  Creating	
  earthquakes	
  to	
  
change	
  organizational	
  mindsets.	
  Academy	
  of	
  Management	
  Executive,	
  8(4),	
  31-­‐43.	
  
	
  
Rosen,	
  M.	
  A.,	
  Bedwell,	
  W.	
  L.,	
  Wildman,	
  J.	
  L.,	
  Fritzsche,	
  B.	
  A.,	
  Salas,	
  E.,	
  &	
  Burke,	
  C.	
  S.	
  (2011).	
  
Managing	
  adaptive	
  performance	
  in	
  teams:	
  Guiding	
  principles	
  and	
  behavioral	
  
markers	
  for	
  measurement.	
  Human	
  Resource	
  Management	
  Review,	
  21:	
  207-­‐122.	
  
	
  
Schmidt,	
  A.	
  M.	
  &	
  DeShon,	
  R.	
  P.	
  (2010).	
  The	
  moderating	
  effects	
  of	
  performance	
  ambiguity	
  on	
  
the	
  relationship	
  between	
  self-­‐efficacy	
  and	
  performance.	
  
	
  
Schmidt,	
  A.	
  M.,	
  Dolis,	
  C.	
  M.	
  &	
  Tolli,	
  A.	
  P.	
  (2009).	
  A	
  matter	
  of	
  time:	
  Individual	
  differences,	
  
contextual	
  dynamics,	
  and	
  goal	
  progress	
  effects	
  on	
  multiple-­‐goal	
  self-­‐regulation.	
  
Journal	
  of	
  Applied	
  Psychology,	
  94(3),	
  692-­‐709.	
  
	
  
Selig,	
  J.	
  P.	
  &	
  Little,	
  T.	
  D.	
  (2012).	
  Autoregressive	
  and	
  cross-­‐lagged	
  panel	
  analysis	
  for	
  
longitudinal	
  data.	
  In	
  B.	
  Laursen,	
  T.	
  D.	
  Little,	
  &	
  N.	
  A.	
  Card	
  (Eds.),	
  Handbook	
  of	
  
Developmental	
  Research	
  Methods	
  (pp.	
  265-­‐278).	
  New	
  York,	
  NY:	
  Guilford	
  Press.	
  
	
  
Senge,	
  P.	
  M.	
  (1990).	
  The	
  leader’s	
  new	
  work:	
  building	
  learning	
  organizations,	
  Sloan	
  
Management	
  Review,	
  32(1),	
  7-­‐23.	
  
	
  
Seo,	
  M.-­‐G.	
  &	
  Ilies,	
  R.	
  (2009).	
  The	
  role	
  of	
  self-­‐efficacy,	
  goals,	
  and	
  affect	
  in	
  dynamic	
  
motivational	
  self-­‐regulation.	
  Organizational	
  Behavior	
  and	
  Human	
  Decision	
  Processes,	
  
109,	
  120-­‐133.	
  
	
  
	
  

184	
  

	
  
Sitzmann,	
  T.,	
  &	
  Ely,	
  K.	
  (2011).	
  A	
  meta-­‐analysis	
  of	
  self-­‐regulated	
  learning	
  in	
  work-­‐related	
  
training	
  and	
  educational	
  attainment:	
  What	
  we	
  know	
  and	
  where	
  we	
  need	
  to	
  go.	
  
Psychological	
  Bulletin,	
  137,	
  421-­‐442.	
  
	
  
Sitzmann,	
  T.	
  &	
  Yeo,	
  G.	
  (2013).	
  A	
  meta-­‐analytic	
  investigation	
  of	
  the	
  within-­‐person	
  self-­‐
efficacy	
  domain:	
  Is	
  self-­‐efficacy	
  a	
  product	
  of	
  past	
  performance	
  or	
  a	
  driver	
  of	
  future	
  
performance?	
  Personnel	
  Psychology,	
  66,	
  531-­‐568.	
  
	
  
Spiro,	
  R.	
  L.	
  &	
  Weitz,	
  B.	
  A.	
  (1990).	
  Adaptive	
  selling:	
  Conceptualization,	
  measurement,	
  and	
  
nomological	
  validity.	
  Journal	
  of	
  Marketing	
  Research,	
  27:	
  61-­‐69.	
  
	
  
Staddon,	
  J.	
  E.	
  R.	
  (1975).	
  Learning	
  as	
  adaptation.	
  In	
  W.	
  K.	
  Estes	
  (Ed.),	
  Handbook	
  of	
  Learning	
  
and	
  Cognitive	
  Processes,	
  vol.	
  2	
  (pp.	
  37-­‐98).	
  Hillsdale,	
  	
  NJ:	
  Erlbaum.	
  
	
  
Terreberry,	
  S.	
  (1968).	
  The	
  evolution	
  of	
  organizational	
  environments.	
  Administrative	
  Science	
  
Quarterly,	
  12(4),	
  590-­‐613.	
  
	
  
Thomas,	
  K.	
  M.	
  &	
  Mathieu,	
  J.	
  E.	
  (1994).	
  Role	
  of	
  causal	
  attributions	
  in	
  dynamic	
  self-­‐regulation	
  
and	
  goal	
  processes.	
  Journal	
  of	
  Applied	
  Psychology,	
  79(6),	
  812-­‐818.	
  
	
  
Tiffin,	
  J.	
  &	
  Lawshe,	
  C.	
  H.	
  (1942).	
  The	
  adaptability	
  test:	
  A	
  fifteen	
  minute	
  mental	
  alertness	
  test	
  
for	
  use	
  in	
  personnel	
  allocation.	
  Journal	
  of	
  Applied	
  Psychology,	
  26,	
  846-­‐849.	
  
	
  
Tolli,	
  A.	
  P.	
  &	
  Schmidt,	
  A.	
  M.	
  (2008).	
  The	
  role	
  of	
  feedback,	
  causal	
  attributions,	
  and	
  self-­‐
efficacy	
  in	
  goal	
  revision.	
  Journal	
  of	
  Applied	
  Psychology,	
  93(3),	
  692-­‐701.	
  
	
  
Trites,	
  D.	
  K.,	
  Kubala,	
  A.	
  L.,	
  &	
  Cobb,	
  B.	
  B.	
  (1959).	
  Development	
  and	
  validation	
  of	
  adaptability	
  
criteria.	
  Journal	
  of	
  Applied	
  Psychology,	
  43(1),	
  25-­‐30.	
  
	
  
Tsui,	
  A.	
  S.,	
  &	
  Ashford,	
  S.	
  J.	
  (1994).	
  Adaptive	
  self-­‐regulation:	
  A	
  process	
  view	
  of	
  managerial	
  
effectiveness.	
  Journal	
  of	
  Management,	
  20(1),	
  93-­‐121.	
  
	
  
Tushman,	
  M.	
  T.	
  &	
  Romanelli,	
  E.	
  (1985).	
  Organizational	
  evolution:	
  A	
  metamorphosis	
  model	
  
of	
  convergence	
  and	
  reorientation.	
  Research	
  in	
  Organizational	
  Behavior,	
  7,	
  171-­‐222.	
  
	
  
Vancouver,	
  J.	
  B.	
  &	
  Kendall,	
  L.	
  N.	
  (2006).	
  When	
  self-­‐efficacy	
  negatively	
  relates	
  to	
  motivation	
  
and	
  performance	
  in	
  a	
  learning	
  context.	
  Journal	
  of	
  Applied	
  Psychology,	
  91(5),	
  1146-­‐
1153.	
  
	
  
Vancouver,	
  J.	
  B.,	
  Thompson,	
  C.	
  M.,	
  Tischner,	
  E.	
  C.,	
  &	
  Putka,	
  D.	
  J.	
  (2002).	
  Two	
  studies	
  
examining	
  the	
  negative	
  effect	
  of	
  self-­‐efficacy	
  on	
  performance.	
  Journal	
  of	
  Applied	
  
Psychology,	
  87(3),	
  506-­‐516.	
  
	
  
Vancouver,	
  J.	
  B.,	
  Thompson,	
  C.	
  M.	
  &	
  Williams,	
  A.	
  A.	
  (2001).	
  The	
  changing	
  signs	
  in	
  the	
  
relationships	
  among	
  self-­‐efficacy,	
  personal	
  goals,	
  and	
  performance.	
  Journal	
  of	
  
Applied	
  Psychology,	
  86(4),	
  605-­‐620.	
  
	
  

185	
  

	
  
VandeWalle,	
  D.	
  (1997).	
  Development	
  and	
  validation	
  of	
  a	
  work	
  domain	
  goal	
  orientation	
  
instrument.	
  Educational	
  and	
  Psychological	
  Measurement,	
  57(6),	
  995-­‐1015.	
  
	
  
Veenman,	
  M.	
  V.	
  J.,	
  Van	
  Hout-­‐Wolters,	
  B.	
  H.	
  A.	
  M.,	
  &	
  Afflerbach,	
  P.	
  (2006).	
  Metacognition	
  and	
  
learning:	
  Conceptual	
  and	
  methodological	
  considerations,	
  Metacognition	
  Learning,	
  1,	
  
3-­‐14.	
  
	
  
Washburn,	
  D.	
  A.,	
  Smith,	
  J.	
  D.,	
  &	
  Taglialatela,	
  L.	
  A.	
  (2005).	
  Individual	
  differences	
  in	
  
metacognitive	
  responsiveness:	
  Cognitive	
  and	
  personality	
  correlates.	
  Journal	
  of	
  
General	
  Psychology,	
  132:	
  446-­‐461.	
  
	
  
White,	
  S.	
  S.,	
  Mueller-­‐Hanson,	
  R.	
  A.,	
  Dorsey,	
  D.	
  W.,	
  Pulakos,	
  E.	
  D.,	
  Wisecarver,	
  M.	
  M.,	
  Deagle,	
  E.	
  
A.,	
  &	
  Medini,	
  K.	
  G.	
  (2005).	
  Developing	
  adaptive	
  proficiency	
  in	
  Special	
  Forces	
  Officers.	
  
Research	
  Report	
  No.	
  1831,	
  U.S.	
  Army	
  Research	
  Institute	
  for	
  the	
  Behavioral	
  and	
  
Social	
  Sciences,	
  Arlington,	
  VA.	
  
	
  
Wiener,	
  N.	
  (1948).	
  Cybernetics:	
  Control	
  and	
  communication	
  in	
  the	
  animal	
  and	
  machine.	
  
Cambridge,	
  MA:	
  M.I.T.	
  Press.	
  
	
  
Woltz,	
  D.	
  J.,	
  Gardner,	
  M.	
  K.,	
  &	
  Gyll,	
  S.	
  P.	
  (2000).	
  The	
  role	
  of	
  attention	
  processes	
  in	
  near	
  
transfer	
  of	
  cognitive	
  skills.	
  Learning	
  and	
  Individual	
  Differences,	
  12:	
  209-­‐251.	
  
	
  
Wood	
  ,	
  R.	
  E.	
  (1986).	
  Task	
  complexity:	
  Definition	
  of	
  the	
  construct.	
  Organizational	
  Behavior	
  
and	
  Human	
  Decision	
  Processes,	
  37:	
  60-­‐82.	
  
	
  
Yeo,	
  G.	
  &	
  Neal,	
  A.	
  (2004).	
  A	
  multilevel	
  analysis	
  of	
  effort,	
  practice,	
  and	
  performance:	
  Effects	
  
of	
  ability,	
  conscientiousness,	
  and	
  goal	
  orientation.	
  Journal	
  of	
  Applied	
  Psychology,	
  
89(2),	
  231-­‐247.	
  
	
  
Yeo,	
  G.	
  &	
  Neal,	
  A.	
  (2008).	
  Subjective	
  cognitive	
  effort:	
  A	
  model	
  of	
  states,	
  traits,	
  and	
  time.	
  
Journal	
  of	
  Applied	
  Psychology,	
  93(30),	
  617-­‐631.	
  
	
  
Zaccaro,	
  S.	
  J.,	
  Banks,	
  D.,	
  Kiechel-­‐Koles,	
  L.,	
  Kemp,	
  C.,	
  &	
  Bader,	
  P.	
  (2009).	
  Leader	
  and	
  team	
  
adaptation:	
  The	
  influence	
  and	
  development	
  of	
  key	
  attributes	
  and	
  processes.	
  
Technical	
  Report	
  No.	
  1256,	
  U.S.	
  Army	
  Research	
  Institute	
  for	
  the	
  Behavioral	
  and	
  Social	
  
Sciences,	
  Arlington,	
  VA.	
  
	
  
Zimmerman,	
  B.	
  J.	
  (1989).	
  A	
  social	
  cognitive	
  view	
  of	
  self-­‐regulated	
  academic	
  learning.	
  
Journal	
  of	
  Educational	
  Psychology,	
  81(3),	
  329-­‐339.	
  
	
  
	
  
	
  

	
  

186