. l i.

«Noni; . 53!:
fir... 32. F
“iii. y.

avqtlz.
I’ll:

3..."...

an

. I?!

4.55 .1!an .15, .
yr 7.3...
'1» u“

gr!

1‘09.|
212......

.111
.. :15.
a) [1.6. u
.. if. 3i?
l..!..... .3}? .3 :
3;!)«7231. i
S .313...
iii...
.iixrii
Entity.“ .
334‘: ....
1:31:43: 3.....st
‘ Venn! .71.!»
‘ gt“), i-Tl'stt
. ii ill:

En.

STATE UNIVERSITY LIBRARIES

w withiiiminimum m w

W 3 1293 01417 3086

I

 

 

 

 

 

 

This is to certify that the
thesis entitled

THE ROLE OF COGNITIVE EFFORT
AND LEARNING OUTCOME DEMANDS
IN SKILL ACQUISITION AND LEARNING

presented by

Sandra Leigh Fisher

has been accepted towards fulfillment
of the requirements for

M.A.___ degree in isychology

Major professor

.4241» 9/
fl
Date—AuM

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

LIBRARY
Michigan State
University

 

 

 

PLACE DI RETURN BOX to roman this checkout“)!!! your record.
To AVOID FINES mum on or baton dot. dot.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

THE ROLE OF COGNITIVE EFFORT
AND LEARNING OUTCOME DEMANDS
IN SKILL ACQUISITION AND LEARNING

BY

Sandra Leigh Fisher

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF ARTS

Department of Psychology

1 995

ABSTRACT
THE ROLE OF COGNITIVE EFFORT
AND LEARNING OUTCOME DEMANDS
IN SKILL ACQUISITION AND LEARNING
By

Sandra Leigh Fisher

This research was designed to examine how learner awareness of the type of
Ieaming outcome, and the amount and type of cognitive effort used during Ieaming,
affect performance on knowledge and application tests. The construct of amount of
effort was expanded to include off-task attention and mental workload, as well as time
on task. Three cognitive Ieaming strategies; rehearsal, organizing and personalizing,
were examined. The impact of learning motivation and cognitive ability on effort and
performance was also studied. The results indicated that performance on the two
learning outcomes was positively related. Amount of effort was found to affect
performance on learning outcomes. None of the three hypothesized learning strategies
were related to performance, but the working of a sample problem during Ieaming was
related to application. Learning motivation affected amount of effort, but did not affect
the use of Ieaming strategies. Implications for further cognitive process and learning

research are discussed.

ACKNOWLEDGMENTS

I would like to thank the members of my thesis committee for guiding me
through the thesis process. Kevin Ford, my committee chair, provided support when I
needed it and pushed me when I needed it. Thanks also to Tynan for waiting until a
good time to arrive. Alison Barber endured the first thesis idea, and was still there to
provide ideas and support for the one that made it, along with guidance on polishing
and presenting my research. Rick Deshon was always there to answer questions, talk
through the logic Of my ideas, and provide a wealth of information. The only question I
ever stumped Rick on was about the St. Louis Blues.

Special recognition belongs to Jennifer Gibbings and Brenda lrrer for their hard
work during the data collection. i hope we fed you well enough to compensate for
hours of coding traces. Thanks also to Neal Schmitt for invaluable help with the pilot
test.

Most of all, I want to thank my husband Mike for his never-ending love and

support. if it wasn’t for him, I would be baking bread instead of finishing this thesis.

TABLE OF CONTENTS

LIST OF TABLES ................................................................................................ vi
LIST OF FIGURES ............................................................................................ viii
INTRODUCTION ................................................................................................. 1
Multidimensionality of Learning Outcomes .................................. 4

Motivation in Learning .................................................................. 6

Motivation and Effort .................................................................... 8

Measurement of Effort ................................................................. 9

Encoding and lnforrnation Processing in Learning ..................... 13

Encoding Specificity Research ....................................... 14

Transfer Appropriate Processing .................................... 16

Cognitive Processes in Encoding ................................... 19

Learning Strategies .................................................................... 21

Individual and Contextual Factors .............................................. 24

Learning Motivation ........................................................ 25

Awareness of Learning Outcomes ................................. 27

A Motivational Model of Learner Effort ....................................... 30

Hypotheses ................................................................................ 34

METHOD .......................................................................................................... 47
Participants ................................................................................ 47

Independent Variables ............................................................... 47

Cognitive Ability .............................................................. 47

Learning Motivation ........................................................ 47

Awareness of Learning Outcomes ................................. 48

Amount of Effort ............................................................. 48

Type of Effort ................................................................. 49

Dependent Variables ................................................................. 51

Knowledge Learning Outcome ....................................... 51

Application Learning Outcome ....................................... 52

Learning Task ............................................................................ 52

Experimental Procedure ............................................................ 54

Pilot Study .................................................................................. 55

RESULTS ........................................................................................................ 57

Adequacy of Measures .............................................................. 57

Hypothesis Testing .................................................................... 66
DISCUSSION ................................................................................................... 85

Summary of Findings ................................................................. 85

Study Limitations ....................................................................... 93

Future Research Opportunities .................................................. 96
APPENDIX A: Wonderiic Personnel Test ............................................ 100
APPENDIX B: Learning Orientation Scales ......................................... 103
APPENDIX C: Off-task Attention Scale ............................................... 104
APPENDIX D: Learning Strategy Scales ............................................. 106
APPENDIX E: Mental Workload Scale ................................................ 107
APPENDIX F: Stock Price Prediction Learning Task ........................... 108
APPENDIX G: Knowledge Learning Outcome ..................................... 113
APPENDIX H: Application Learning Outcome ..................................... 117
APPENDIX I: Pilot Study Results ....................................................... 121
LIST OF REFERENCES ................................................................................. 122

Table 1:

Table 2:

Table 3:

Table 4:

Table 5:

Table 6:

Table 7:

Table 8:

Table 9:

Table 10:
Table 11:
Table 12:
Table 13:
Table 14:

Table 15:

LIST OF TABLES

Factor Loadings for Learning Orientation Scales ....................... 59
Factor Loadings for Learning Strategy Scales ........................... 60
Factor Loadings for Off-task Attention and Mental

Workload Scales ........................................................................ 61
Scale lntercorrelations and Reliabilities ..................................... 64

Correlation Matrix for Learning Strategy Traces

and Self-reports ......................................................................... 65
Moderated Regression Analysis on Application Outcome ......... 68
Moderated Regression Analysis on Knowledge Outcome ......... 69
Regression Analysis Results on Off-task Attention .................... 71
Regression Analysis Results on TIme ........................................ 72
Regression Analysis Results on Mental Workload ..................... 73
Regression Analysis Results on Rehearsal ............................... 75
Regression Analysis Results on Organizing .............................. 76
Regression Analysis Results on Personalizing .......................... 77
Regression Analysis Results on Worked Sample Problem ........ 78

Mediated Regression Analysis Results on Knowledge of

Learning Outcome through Learning Strategies ........................ 81

vi

Table 16:

Table 17:

Table 18:

Table 19:

Mediated Regression Analysis Results on Knowledge of

Learning Outcome through Amount of Effort ............................. 82
Direct Effects Regression Analysis Results on Application
Learning Outcome ..................................................................... 83
Direct Effects Regression Analysis Results on Knowledge
Learning Outcome ..................................................................... 84

Prediction of Off-task attention from Learning Orientation ......... 87

vii

Figure 1:

Figure 2:

Figure 3:

Figure 4:

Figure 5:

LIST OF FIGURES

Conceptual Model of the Role of Effort in Learning and

Skill Acquisition .......................................................................... 35
Proposed Moderator Relationship .............................................. 35
Hypothesized Relationships Among Type of Effort,

Amount of Effort, and Learning Outcome Performance ............. 37
Antecedents of Learner Effort .................................................... 39
Hypothesized Relationships Among Learning Orientations

and Amount of Effort .................................................................. 41

viii

INTRODUCTION

Given the substantial investment in time and resources devoted to training and
education every year, psychologists in many disciplines have investigated ways to
improve individual Ieaming. Learning is defined as “a relatively permanent Change in
knowledge or skill produced by experience” (Weiss, 1990, p. 172). Leaming is
generally recognized to involve an interaction between individual characteristics, such
as ability, motivation, attention, and effort, and situational factors, such as the instructor
and the material to be learned (Gagné, 1984; Biggs, 1992).

Research investigating the role of attention and effort in Ieaming can be traced
back to the days of Ebbinghaus. Using nonsense syllables as material, Ebbinghaus
measured the amount of effort, or number of Ieaming trials, required to learn a set of
stimuli to a set criterion level. He then had subjects releam the material at a later date,
and calculated a savings score, which was the reduction in effort required to releam the
material (Ashcraft, 1993).

The effort put forth by subjects in Ebbinghaus’ experiments was rote
memorization, or pure, factual rehearsal. Subjects repeated the nonsense syllables
again and again until they remembered them. Cognitive psychologists have
investigated the I'OIO of rehearsal in memory, and discovered that it serves two primary
purposes; retention of material in short term memory, and the transfer of material from

short term memory into long term memory (Ashcraft, 1993). it has been generally

2
accepted that rehearsal is a useful way to transfer information into long term memory.

Craik and Lockhart (1972) attempted to distinguish between different types of
rehearsal through their depth of processing theory. They claimed that deep, or
meaningful, processing was the best way to remember learned material over time.
Shallow processing, similar to Ebbinghaus’ rehearsal, does not leave a durable memory
trace, according to Craik and Lockhart ( 1972). Thus, the depths of processing theory
proposed a ‘one best way' view of cognitive processing in Ieaming. Regardless of the
stimulus material and demands Of the outcome test, the levels of processing theory
predicts that deep, intentional processing produces the best retention.

Later research has indicated that there may not be one best way to team all
types of material in all situations. Morris, Bransford and Franks (1978) suggested the
notion of transfer appropriate processing. They posited that Ieaming outcomes, or test
conditions, must be considered when processing during Ieaming. If the desired
outcome is an understanding of a paragraph of text, then Craik and Lockhart’s idea of
deep processing may still be appropriate. However, if the desired Ieaming outcome is
to simply know how many words were in each sentence of the same paragraph, then
the traditional deep processing may not be the best mode of processing. Morris, et al.,
(1978) suggested that the type of processing that is meaningful differs for each
Ieaming task, and Ieaming outcomes must be designed to tap what was supposed to
be Ieamed.

Effort has been mentioned often in Ieaming and skill acquisition research, but
effort is Often put aside in favor of abilities or skills as the variable of primary interest.
Cognitive effort has typically been conceptualized as time on task, or attention devoted

to the task (Mitchell, Hopper, Daniels, Falvy & James, 1994; Kanfer & Ackerrnan,

3
1989). The construct of effort will be reviewed as it has been used in motivation and
Ieaming research. It is proposed that in addition to the amount of effort put forth by a
learner, the type of cognitive effort, or processing, used in a Ieaming situation will affect
performance on a Ieaming outcome measure. Theory and research on transfer
appropriate processing are reviewed. Individual factors (Ieaming orientation and
cognitive ability) and contextual influences (knowledge of the type of Ieaming outcome
measure) are proposed to affect the type and amount of effort used by the learner in
skill acquisition.

There are two main themes in this proposal. First, it is suggested that the
constmct of effort as it has been used in training and Ieaming research is incomplete.
Effort is typically Operationalized as time on task (Kanfer 8. Ackennan, 1989), or
occasionally perceptions of on-task attention (e.g. Paas, 1992). Measures of effort
should be used in conjunction with one another, as each measure contributes a unique
view of the amount of effort construct. In addition, researchers, instructors, and
Ieamers must consider not only the amount of effort allocated toward Ieaming, but also
the type of effort used. it is suggested that some types of effort are better suited to
particular Ieaming outcomes.

Second, it is proposed that evaluation procedures at the end of Ieaming events,
when made known, signal Ieamers to allocate their efforts toward acquiring the
knowledge and/or skills which will allow them to perform well on the evaluation. For
example, if the objective of a training program is for trainees to Ieam a skill, a
declarative knowledge test would be inappropriate to measure the Ieaming objectives.
Hubbard (1994) presented an example of how testing affects students” study

behaviors. First, students gather information from the syllabus, former students, and

4

the teacher concerning testing procedures. Students then adjust their study behaviors
accordingly. Hubbard (1994) suggested that students will use only the study skills
which are required for the expected testing procedures. If a college professor intends
students to develop skills that can be used on the job, he or she should orient
evaluation procedures to skills, rather than declarative knowledge as demonstrated on
a multiple choice exam.

Mutt—idimensionfly of Lea_r_nirrq Outcomes

Before attempting to determine how people learn, criteria must be specified so
“Ieaming” can be measured. Defining Ieaming as a relatively permanent change in
knowledge or skill does not directly suggest how one might measure whether or not any
Ieaming has taken place. When Ieaming is viewed as a multidimensional construct, the
specification of criteria becomes more clear. Kraiger, Ford and Salas (1993) presented
a framework which divides Ieaming outcomes into three categories; affective,
behavioral, and cognitive. Affective Ieaming outcomes consist of attitudes and
motivation which might be desired outcomes of training. Examples of this type of
Ieaming outcome include acceptance of diversity, and organizational commitment
(Kraiger, et al., 1993). Behavioral, or skill-based, outcomes consist of psychomotor
skills one could learn in training. Cognitive outcomes consist of verbal knowledge, as
well as higher order knowledge organization and cognitive strategies.

Anderson’s ACT‘ theory presents the acquisition of knowledge and skills as a
stage model. This theory posits that the dimensions of Ieaming are arranged
hierarchically. The first stage of the model involves the acquisition of declarative
knowledge. Declarative knowledge is knowledge about things, such as your mother’s

maiden name, or the channel Seinfeld is on. Procedural knowledge is knowledge of

5
how to do things, such as drive a car with a manual transmission, or write a
grammatically correct sentence. According to Anderson, one cannot learn procedural
skills without first having acquired the declarative basis for that skill. Thus, if the
ultimate Ieaming goal is skill-based. trainees must first acquire the relevant declarative
knowledge before applying that knowledge to proceduralization.

In the declarative stage of skill acquisition, task performance is slow and
effortful. The facts about performance must be held in working memory. The learner
must verbalize frequently during performance (Weiss, 1991). The second stage of skill
acquisition is knowledge compilation. During this phase, the Ieamer begins to build the
mles, or production systems, necessary to perform a task. Production systems are
arranged in “If: Then” form, and are organized hierarchically according to goals
(Anderson, 1982). When the Ieamer is presented with a goal, such as tying one’s
shoe, there are several sub-goals which must be accomplished in a particular order.
Over time, and with practice, the productions are arranged and combined to make
performance smooth and easy. When the Ieamer reaches the third stage of skill
acquisition, performance requires substantially fewer mental operations than
performance in the declarative stage.

Bloom (1956) presented a taxonomy of cognitive Ieaming outcomes which
contains six levels. The levels, ranging from the most concrete to the most complex,
are knowledge, comprehension, application, analysis, synthesis, and evaluation. The
taxonomy was originally developed to aid teachers in designing curriculum and
evaluating students. Knowledge is defined as remembering ideas or facts through
recall or recognition. Bloom recognized that knowledge is required to perform the more

complex objectives. However, the focus in this category is on isolated facts which can

6
be remembered separately. Comprehension is more complex, requiring students to
know what is being communicated in a situation, and make limited use of the material.
Limited use, in this situation, can include translation, interpretation, and extrapolation.
In contrast, application requires students to use an abstraction in a new situation.
Thus, the student must not only remember the concept, but use that concept to solve a
problem.

The distinction between knowledge and application is similar to that between
declarative and procedural knowledge. Both knowledge and declarative knowledge are
needed to develop the higher order stmctures. One important quality of true procedural
knowledge is that the Ieamer need not consciously access the declarative knowledge to
perform the behavior. In Bloom’s taxonomy (1956), no such limitation is placed on
application. Application is more likely to occur than proceduralization in the shorter time
periods of one class period or one training session. Regardless, the principles of
application and procedural knowledge are very similar. Both are higher level
knowledge structures, and both require the use of relevant factual knowledge.
Motivation {gum

Kanfer and Ackerrnan (1989) present a model of motivational processes within a
skill acquisition framework. They discuss motivation as a process in which both distal
and proximal motivational processes are used to allocate cognitive resources (i.e.
attention and effort). Distal motivation processes involve the choice to use one’s
resources for a particular task. At this first stage, the individual decides how many
resources to allocate for the task at hand. Proximal motivation processes determine the
distribution of attention and effort within a given task, once the individual has chosen to

engage in the task. Each individual possesses a limited amount of cognitive resources,

7
or ability, which can be distributed. These resources are allocated among three types
of behaviors; task related, self-regulation, and Off task. The motivational processes
that affect resource allocation include goals, incentives, individual personality
differences, and metacognitive knowledge. Kanfer and Ackerrnan (1989) demonstrated
that the use of explicit proximal goals in skill acquisition was associated with a decrease
in training performance. They contended that effort which was needed to learn the skill
was allocated toward goal monitoring.

There are several boundary conditions which must be considered in the
motivation-performance relationship. First, the level of skill proficiency affects the
amount of cognitive resources needed. Skill acquisition is generally considered to
consist of three phases. In the declarative stage, the Ieamer is becoming acquainted
with the task, and the demands of the task. The Ieamer must devote substantial
cognitive resources to Ieaming the task. In later phases of skill acquisition, knowledge
compilation and procedural knowledge, skills become more automatized as the Ieamer
performs the task more smoothly, and fewer cognitive resources are required to
perform the task.

Second, the resource-dependency of the task must be considered. If the task is
considered resource-dependent, performance is at least partially dependent on the
amount of attention which is allocated to the task. On a resource insensitive task,
however, attention allocation does not impact task performance. Such a task is said to
be data-limited, and performance is affected more by task characteristics than by the
availability of cognitive resources (Kanfer & Ackennan, 1989). Thus, both the resource
dependent task, and the early stages of skill acquisition require high amounts of

attention and effort.

8

Dweck and her colleagues (Dweck, 1986; Dweck & Leggett, 1988) also focus on
effort as a primary mechanism linking Ieaming motivation and performance. Dweck
and colleagues have suggested that students who are motivated by the task itself do
well because they have a healthy attitude towards effort, and can direct all effort
towards the task. Individuals who are motivated by the potential for reward or
recognition stemming from the Ieaming event tend to perform less well than task
motivated students. The performance oriented students do not direct all effort towards
the task, as they are concerned with protecting their ego. Thus, the students who are
able to direct maximum effort toward the Ieaming task generally perform better.

One limitation identified with the work of Kanfer and Ackennan (1989) is the
treatment of all on-task effort as being equal. Kanfer and Ackennan did not investigate
how the trainees Ieamed. They focused primarily on goal manipulations and ability as
determinants of training performance. They refer to variations in on-task effort as only
increases or decreases, not as changes in strategy or type of effort. Perhaps the less
successful performers were not approaching the task in the most useful manner. The
next section investigates the construct of effort, how effort has been measured, and
how the treatment of the construct has limited Ieaming research.

Motivation and Effort

Motivation research in Industrial and Organizational psychology has dealt with
three primary outcomes of motivational processes; direction of behavior, intensity of
action, and persistence of behaviors over time (Kanfer, 1990). Direction of behavior
typically refers to the choice of a behavior or course of action. However, direction can
also be considered within a particular course of action. For example, a student could

choose to study for an exam, but the student could then choose different study

9
strategies, such as memorizing a list of terms or paraphrasing class notes. Intensity of
action refers to how much effort is expended by the individual. For example, a student
could study for the exam for one hour, or six hours. Amount of effort, though, may not
directly lead to performance improvements. According to Kanfer, “motivational
processes may fail to affect performance because effort is misdirected - persons work
harder at the wrong things” (Kanfer, 1990, p. 81).

Persistence of behavior has been used less frequently in motivational research,
as it is a long-tenn result of motivation. Our student would display persistence if he
studied in a particular way, with a certain amount of effort, over the entire semester, or
his entire college career. Because persistence is viewed as a combination of direction
and intensity over time (Kanfer, 1991), persistence will not be explicitly considered in
this study.

In the Integrated Resource Model, motivational processes direct choices
involving the direction of effort and the intensity of effort (Kanfer & Ackennan, 1989).
Effort is directed in some combination of on-task, off-task, and self-regulatory activities.
Intensity is conceptualized as the amount of effort directed in any of these three
directions. In Kanfer and Ackerrnan’s model, direction is not considered within a
particular course of action. A more fine-grained analysis which looks into specific on-
task, off-task, and self-regulatory activities has potential to provide a greater
understanding of Ieaming processes. The present paper investigates measurement
and conceptualization issues surrounding on-task Ieaming activities.

_M_ea§urement of Effort
As suggested above, cognitive effort is comprised of two components: amount

of effort, and type of effort. Many researchers have operationalized effort as the

10
amount of time spent on a task (T erborg, 1977; Dweck, 1986), or perceived effort on
the part of the Ieamer (Paas, 1992; Wofford, 1990; Hart & Staveland, 1988). Any
measure of effort as actual time spent on the task, though, is contaminated. An
individual may appear to be working on a task, or thinking about a task, but his/her
attention may be focused elsewhere. Kanfer and Ackennan (1989) define effort as the
amount of attentional resources devoted to the task. Considering the importance of the
attentional component to the Integrated Resource Model, it seems useful to determine
if the trainee is focusing attention on the task at hand, along with the time spent on the
task. For example, a trainee could be working on a problem from 11:00 to noon, but if
he thinks about lunch every five minutes, he significantly reduces the actual attentional
effort devoted to the problem.

Paas and colleagues (Paas & Van Merrienboer, 1994; Paas, 1992) have
operationatized mental effort separately from time on task. According to Paas (1992),
mental effort is the amount of capacity that is allocated to instructional demands.
Mental effort was measured with a self-report scale on which subjects indicated the
perceived amount of mental effort devoted to the task. Trme on task did not appear to
be related to amount of perceived effort. For example, subjects who studied pre-solved
statistics problems spent significantly less time on the task than subjects who actually
solved statistics problems. The mean amount of perceived effort did not differ across
conditions (Paas, 1992). Fees did not measure the actual mental processes used, but
suggested that some mental processes may be more or less relevant to the teaming
task, and “less effort could be invested in more relevant Ieaming processes“ (Paas,

1992,p.433)

1 1

Hart and Staveland (1988) also measured amount of effort by subjective self-
report. They define mental workload, as the cost incurred by a person to achieve a
given performance level. It is a function of task properties, situational factors, and the
skills, behaviors and perceptions of the person. Although mental workload is typically
used to describe between-job workload requirements (9.9. Hancock & Caird, 1993), it
can also be used to describe between-individual workload expenditures on a given
task. Hart and Staveland (1988) suggest that self-report measures of mental workload
“may come closest to tapping the essence of mental workload, and provide the most
generally valid and sensitive indicator’ (Hart & Staveland, 1988, p. 141).

Hart and Staveland (1988) developed a self-report, multidimensional measure of
mental workload, the NASA Task Load Index (NASA-TLX). The measure includes
items which reference, for example, the mental demand, temporal demand, physical
demand, and fmstration level of a task. Ratings on individual dimensions were found to
correlate highly with global ratings of workload. The scale can be given in many forms,
including verbally, paper and pencil, or by computer without appreciably altering the
psychometric characteristics of the test. While between subject variability is often
considered problematic in human factors research (Hart & Staveland, 1988), it was
considered an interesting aspect of self-report ratings of mental workload. Wrth the
NASA-TLX, researchers can obtain stable, valid measures of the amount of effort
required by a particular task, for a given individual. The mental workload measure
addresses one deficiency in the measurement of effort as time on task.

Kanfer and Ackerrnan (1989) addressed another aspect of the construct of
amount of effort. They used a self-report measure of thought content which tapped into

the mental activity of the subjects. Subjects were asked to what extent they set goals

12
for themselves, compared their performance to others, or daydreamed. Thus, Kanfer
and Ackerrnan address another deficiency of the effort construct, as they combine the
traditional time on task (or number of trials), with information regarding on-task and off-
task mental effort.

Each measure of amount of effort discussed above, time on task, mental
workload, and on-task/off-task attention, captures a piece of the amount of effort
construct. The relationships among these measures should be explored. For example,
Paas (1992) has suggested that time on task is unrelated to perceived mental
workload. How is on-task/off-task attention related to time on task and workload?
Further, what are the relationships among these variables and Ieaming or task
performance?

Once the amount of on-task effort has been identified, it is important to discover
how the Ieamer spent that time. Terborg (1977) addressed the direction of effort in his
model of work performance. Direction of effort was Operationalized not through actions
of the worker, but through role definitions, or the worker’s understanding of the
appropriateness of various work related activities. However, most researchers do not
directly address the issue of the type Of on—task cognitive activities during teaming.

While llO psychologists have found that motivation and intensity of effort play
important roles in Ieaming and skill acquisition, many cognitive psychologists would
argue that the information processing which occurs during the teaming is more
important than the goals and motivation of the Ieamer (Anderson, 1983). There is
substantial evidence in the cognitive and instructional psychology Iiteratures that the

type of processing, or direction of effort within the Ieaming task, during the encoding

1 3
phase of teaming is vital. This literature is reviewed below, with the intention of linking
these cognitive processing concepts to on-task activity during teaming.
Encoding and Information Processing in Learning

Many cognitive psychologists have investigated the role of processing in the
acquisition of information and skills. The consensus view suggests that different types
of processing lead the Ieamer to attend to different aspects of the material.
Researchers must consider the processes involved in teaming - encoding, organization,
and retrieval. The accuracy of retrieval processes, such as those necessary for
performance on a teaming outcome measure, is dependent upon what information is
stored, how the information is stored, and how that information matches the
subsequent cues for retrieval (Lord & Maher, 1991). It is not just the information itself
that is important, but also the organization of the information in memory.

Much of the research concerning encoding and retrieval processes is based on
the encoding/retrieval paradigm of episodic memory research (T utving, 1983). This
paradigm proposes a 2X2 interaction between encoding condition and retrieval
condition. In experiments utilizing this paradigm, two encoding conditions are tested
with each of two retrieval conditions. The material used is typically the same for each
condition, but the focus in encoding and retrieval is varied. For example, a researcher
using one list of words could vary encoding condition by instructing subjects to write
down the first synonym they can think of for the target word (semantic) or count the
number of E's in that same target word (surface-features). Memory tests could be
implicit or explicit. Explicit tests involve conceptual, semantic information, while implicit
tests are data-driven, and involve physical properties of the stimuli. The usual

prediction of the encoding specificity hypothesis is that performance would be superior

14
in the "match" cells, where semantic encoding enhances performance on a free recall
or recognition test (Roediger, Weldon & Challis, 1989). A mismatch call, such as the
combination of semantic encoding and an implicit memory test, should result in poor
performance. Tulving (1983) contended that the use of this 2 X 2 design was the only
way to rule out alternative explanations, and fully test the encoding specificity
hypothesis.

Encoding Specific'gy Research
Earhard (1969), in an attempt to study the independence of memory traces,

discovered that subjects could not team a list of words in one manner, and then recall
them in a different manner. She presented the to-be-leamed words to subjects in a
serial fashion. Some Of the subjects were to recall the words in free recall or serial
conditions. Others were to recall them alphabetically. The subjects who were to recall
the words alphabetically performed poorly, unless they were told before presentation
they would be required to recall the words alphabetically. Thus, reorganization of
teamed words was an extremely difficult task, which became even more difficult as the
study time increased. Earhard concluded that it is beneficial to store items in memory
according to the retrieval system which will be used (Earhard, 1969).

In another study involving retrieval processes, Thomson and Tulving (1970) had
subjects team a list of words alone, or word pairs of weakly associated words. They
tested retrieval of the target words with these teamed cues, which they termed "weak,"
against a cue word which was typically strongly associated with the target word,
defined as the "strong" cue. Thomson and Tulving (1970) found that the strong cues
worked best when subjects did not team the word pairs, just a list of target words. The

weak cues facilitated performance better than the strong cues when the word pairs had

15

been Ieamed. The authors interpreted this effect as support of the encoding specificity
hypothesis, which states, " Specific encoding operations performed on what is
perceived determine what is stored, and what is stored determines what retrieval cues
are effective in providing access to what is stored” (T utving & Thomson, 1973, p. 369).
According to this principle, test performance will be facilitated to the degree that the test
stimulus conditions match the presentation, or Ieaming, stimulus conditions. When
there was no cue word teamed with the target words, the strong cues facilitated
retrieval because they were associated with the target words in everyday speech.
Since an experimental association was not teamed, the normal cue remained effective.

Tulving and Thomson (1973) further investigated the relationship between
encoding and retrieval processes in a study comparing recall and recognition of test
words. The typical result of such studies is that subjects perform better on recognition
tests than on recall tests. However, Tulving and Thomson found evidence for
substantial recognition failure. Subjects studied a set of target words accompanied by
a cue word. In the test, subjects could recognize the target word by itself only 24% of
the time. However, when given the cue word, the subjects could recall the words 63%
of the time (T ulving & Thomson, 1973). This result clearly contradicts the standard
notion that recognition is easier than recall in word list memory tests. The authors again
called upon the principle of encoding specificity to explain these results. Tulving and
Thomson hypothesized that if subjects had been asked to recognize entire word pairs,
as had been present in the encoding condition, instead of just the target word,
recognition performance would have been much better (T utving & Thomson, 1973).

The principle of encoding specificity has been demonstrated with pictures as

well as with words. Because of the sO-called pictorial superiority effect, a theory which

16

hypothesizes that pictures allow dual encoding of stimuli (imaginal and verbal), many
researchers believed pictures would lead to better performance than words on any
memory test (Weldon, Roediger & Challis, 1989). This hypothesis has been disproven
by several researchers. Weldon, Roediger and Challis (1989) showed that pictorial
encoding produces low performance on a word-based implicit memory test. Subjects
were shown a collection of pictures and words. They were then asked to complete
word fragment or word stem completion tests. Subjects were able to complete the
items which had been teamed as words significantly more often than words which had
been teamed as pictures. Watkins, Peynircoiglu and Brehms (1984) found similar
results. Subjects rehearsed a set of either pictures or words with the same semantic
meanings. They were then shown fragments of either pictures or words. Subjects who
had studied the pictures were able to identify the pictures more easily, whereas
subjects who had studied the words were able to identify words more easily. (Watkins,
Peynircoiglu & Brahms, 1984). Performance was best when the mode of study
matched the retrieval cue. These results are consistent with the principle of encoding
specificity. The pictorial superiority effect is similar to the "one best way" notion of
encoding which can be found in Craik and Lockhart's levels of processing idea. The
levels of processing theory claims that deeper, or more meaningful, processing during
encoding will result in better memory (Ashcraft, 1989).
Transfer Appropriate Processing

Morris, Bransford and Franks (1977) highlighted the importance of the type of
cognitive processing in the Ieaming event. This was the first study to use the term
“transfer appropriate processing.” Morris, et al., among others, were dissatisfied with

Craik and Lockhart’s (1972) levels of processing theory , which attempted to distinguish

17
between objectively defined meaningful and non-meaningful processing methods.
However, as noted by Morris, Bransford and Franks (1977), "many of the results
favoring the levels of processing claims may be due in large part to an inherent bias in
the way in which memory was tested (p. 521)." They believed that there was no one
best way to process information for Ieaming; rather, the best method of processing
would depend on the goal of the Ieaming, or the testing situation. The processing task
must still be meaningful, but within the specific Ieaming context. They found support
for their theory using a simple word task. Subjects performed better on a semantic test
when they had used a semantic acquisition task, and performed better on a rhyming
test when they had used a rhyme-focused acquisition task.

This series of experiments also showed that the rhyme encoding task allowed
superior performance over semantic encoding on the delayed rhyme test. This finding
questions the levels of processing idea that deeper processing allows for stronger
memory traces, hence longer retention. Thus, Morris, et. al., (1977), concluded that no
one teaming method is inherently superior in increasing the strength, accessibility, and
durability of memory traces. However, the different Ieaming methods can orient the
Ieamer towards different types of information. It then becomes very important to
encode to-be-leamed information in a way consistent with the ultimate Ieaming goal. As
demonstrated by both the Morris, et al., (1977) and Roediger, et al., (1989) studies, "the
value of particular acquisition activities must be defined relative to particular goals and
purposes" (Morris, et al., 1977, p. 528).

Morris, et al., (1977) also highlighted the importance of looking beyond simple
input retention as the representation of teaming. Depending on the teaming situation,

individuals must team from the inputs, rather than teaming the actual inputs. For

1 8
example, if the training is intended to teach assertiveness skills, it is not the actual
situation that is important, it is the concepts behind the sample situation. This is an
important distinction between the often used verbal Ieaming studies in cognitive and
instructional psychology, and the skill acquisition situations found in organizational
settings. Under the verbal teaming assumption that the outcome will be a test of the
retrieval of the list of inputs, the deep processing notion makes sense. When one
considers other types of Ieaming outcomes, one also must consider the viability of
other types of Ieaming processes.

Barnett, Di Vesta and Rogozinski (1981) investigated the purpose of notes in an
academic setting. Study experts usually recommend taking meaningful, or elaborated,
notes on lecture material as a way to better remember the material. It was found that
the students who elaborated on the material did worse on multiple choice tests
constructed by the teacher. Students had not teamed the material in the way in which
it was tested. When the students wrote their own items, they could answer those much
better at a later date if they had done the elaborative processing during note-taking.
Barnett, et al., (1981) claimed support for transfer appropriate processing. They also
found that the quality of the elaboration mattered. Students had to add conceptual
material, not just lengthen the sentences in the notes. This experiment points to the
content, or attentional function of the cognitive processes/Ieaming strategies. The
elaborations to the notes added material which was not covered on the test. On a
teacher constructed multiple choice test, this additional information interfered with the
retrieval of the original material. Similar results were reported by McKelvey and Lord
(1986), as elaborative note-taking improved memory accuracy only when matched with

similar retrieval situations.

19

Research in encoding and retrieval processes, such as transfer appropriate
processing, has highlighted the importance of considering various types of information
processing during teaming. If a Ieamer uses one type of processing during Ieaming,
that knowledge or skill may not be readily available under different retrieval, or test,
conditions. Schmidt and Bjork (1992) noted the importance of the overlap of cognitive
processes during training (acquisition) and processes necessary for test performance.
They suggested that effective training will maximize the overlap between the
processes. No single type of processing activities will work for all teaming tasks. The
following section examines various types of processing which are suggested to be
useful in Ieaming.
C_Oggiive Processesin Encoflg

The literature concerning cognitive processes during the encoding and
organization of information can be organized into two primary dimensions; a holistic,
Gestalt approach, and a specific, detail oriented, data driven approach. Many
psychologists have studied encoding processes in reading and arithmetic (e.g. Craik
and Lockhart, 1972; Das, 1988; Kirby, 1988, Marlon, 1988). Outside of the Ieaming
domain, Triesman uses a similar approach to describe encoding of letters (T riesman,
1986). Others have used comparable processes to describe the encoding of scenes
and pictures (Henderson, 1992; Biedennann, 1990). Regardless of the perceptual
subject, there is great consistency across approaches in the use of part-whole
distinctions in the encoding and organization of information. There are many different
labels for these dimensions (to. simultaneous/successive, holistic/atomistic), but they

are essentially instances of these two categories.

20

Das (1988) described these two types of processing as simultaneous and
sequential. In simultaneous encoding of information, separate pieces of information
are combined into a meaningful, relational format. In successive processing, each
piece of information is considered separately, and the information is placed into a
temporally ordered sequence. Das diagnosed the skill levels of subjects on each of
these two types of processing, and related their scores to performance on reading and
arithmetic tests. Unfortunately, type of processing was not manipulated, and Das could
not be sure which type of processing subjects were using during the performance tests.
However, students who scored highly on the reading test were skilled at both
simultaneous and sequential processing, and students who scored highly on the
arithmetic test tended to be skilled in simultaneous processing. Simultaneous
processing was strongly related to more advanced reading skills, such as
comprehension of conceptual relationships. Das (1988) concluded from these results
that sequential processing is less important for arithmetic skills, while both types of
processing are needed in reading.

Marion (1988) distinguished between encoding processes described as holistic
and atomistic approaches to teaming. tn the holistic approach, the Ieamer encodes the
material hierarchically, considering the material as a whole. Atomistic Ieamers encode
the material in a sequential fashion, focusing on the details (Marten & Saljo, 1984;
Marion, 1988). For example, a chapter in a textbook may outline a principle, and then
offer several instances of that principle. A Ieamer taking the holistic approach would
see the hierarchical relationships between the principle and the specific instances. The

Ieamer taking the atomistic approach could recall the principle and the instances, but

21
would view them on the same level, as a list of temporally ordered facts with no
apparent relationship (Marton, 1988).

These authors suggest that the two primary types of encoding processes are
demonstrated across many teaming situations. Comparing the ideas of Das and
Marten with the findings of the encoding-retrieval interaction paradigm leads us to
consider the retrieval situations, or types of teaming outcomes, in which these different
types of encoding processes might be appropriate. However, these general cognitive
processes are difficult to use in an applied setting such as organizational training.
These processes can better be thought of as constructs in teaming, which can be
operationally represented by Ieaming strategies (Schmeck, 1983; Das, 1988). The
literature on teaming strategies as an operationalization of cognitive encoding
processes is reviewed below.

Learning Strategies

Gagné, Briggs & Wager, (1992) define the Ieaming strategy as “an internal
process by which Ieamers select and modify their ways of attending, Ieaming,
remembering, and thinking” (Gagné, Briggs & Wager, 1992, p. 66). Three categories of
strategies directly related to Ieaming are rehearsal, elaboration, and organizing.
Rehearsal is a forced Ieaming, repetitious procedure. Elaboration requires the Ieamer
to associate the new material with other, familiar material. Organizing strategies require
the Ieamer to find similarities and themes within the new material. This strategy differs
from elaboration in that elaborative Ieaming links the new material to already familiar
material, while in organizing, links are found within the new material. There are also
monitoring strategies which relate to goal setting and other activities useful for meeting

teaming objectives. These monitoring strategies are not directly related to the

22
acquisition of knowledge or skills, and are similar to Kanfer and Ackerrnan’s (1986) self-
regulation category.

Gagné (1984) suggests that these categories Of Ieaming strategies are
differentially useful in the acquisition of various teaming outcomes, such as procedural
skills or declarative knowledge. For a knowledge outcome such as verbatim
reproduction of text, the Ieamer does not go through the stages of Ieaming that are
typical for a more complex, or application outcome (to. composition and automaticity).
The Ieamer could use the rehearsal strategy to acquire the declarative knowledge
(Gagné, 1984).

Schmeck defines a Ieaming strategy as a set of procedures for accomplishing
teaming. While the majority of Schmeck’s research has focused on teaming styles
rather than teaming strategies, Schmeck (1983) notes that there are situationally
specific factors of Ieaming processes, as well as general tendencies. He suggests that
a teaming style is a predisposition to use a particular strategy across time and teaming
domains. Schmeck’s (1988) classification of teaming strategies is similar to that of
Gagné, et al., (1992). Schmeck’s categories are conceptualizing, personalizing, and
memorizing. The conceptualizing strategy involves dealing with the abstractions in
material. Wrth this strategy, the Ieamer focuses on the meaning of the material, as in
Gagné’s organizing strategy. Personalizing strategies call for integrating new material
into personal experience, as with Gagné’s elaboration strategy. Schmeck’s memorizing
strategy is simply the rehearsal of information, focusing on the given attributes and
details, as in Gagné’s rehearsal strategy.

The three categories of cognitive strategies (Gagné, et at, 1992; Schmeck,

1988) represent the two types of cognitive processing outlined above. Data—driven,

23
detail oriented, sequential processing is represented by the rehearsal/memorization
strategy. Learners using this strategy attend to the specific information present in the
learning material, and direct their effort towards encoding the material as it is
presented. Holistic, simultaneous processing is represented by the integrative,
conceptualizing, personalizing strategies. Learners using these strategies are not
bound by the material presented. They direct their effort toward understanding how the
material fits in the bigger picture, either with previously teamed information, or within
itself. This structure is similar to the structure needed for proceduralization of skills.
The focus is on Ieaming relations among bits of information, rather than the bits of
information by themselves.

The bulk of the teaming strategy research has been conducted in the
instructional psychology domain. Thus, the setting of the Ieaming is traditionally the
classroom, and the Ieaming task is verbal. tn teaming style research, there is usually
not a specific Ieaming task; the researcher asks participants to respond to a
questionnaire asking about typical teaming habits (Biggs, 1993). Thus, Ieaming
strategies are differentiated from teaming styles with a temporal dimension. The
teaming style is a tendency to use a particular strategy across many situations
(Schmeck, 1983, 1988; Weinstein, 1986).

Many of the authors in the educational psychology literature do not
acknowledge the importance of the task specificity of Ieaming strategies. This
disregard for task demands in educational psychology could be a result of an
assumption of the experimental task; teaming the content of a written passage. For
example, Marton and Saljo (1976) addressed the issue of defining the Ieaming content

space, as what is teamed is just as important, if not more important, than how much is

24
Ieamed. However, they attempted to define the content-oriented teaming space
hierarchically. Marion and Saljo (1976) concluded that “deep” processing is better than
“shallow” processing, as the top of the hierarchy was the overall meaning of the
passage, not specific details. Some researchers have used problem solving tasks (e.g.
Dweck, 1986), but most have instructed subjects to read a passage, and report the
content of that passage in one form or another.

Levin (1986) emphasized the “multiple objectives principle,” where there is not
one best overall teaming strategy. Similar to the transfer appropriate processing
argument of Bransford, et al., (1977), Levin suggests the ‘best’ Ieaming strategy
depends on the demands of the task. Investigation of Ieaming strategies in the training
domain requires the consideration of task characteristics, and Levin’s multiple
objectives principle. Different Ieaming objectives require different Ieaming strategies.

Learning strategies have been categorized into three main types; rehearsal,
personalizing, and organizing (Gagné, et al., 1992; Schmeck, 1988). An individual can
use any combination of these strategies in a given Ieaming situation. However, no
“one best” Ieaming strategy has been identified across teaming situations. The best
strategy for a given situation depends on the teaming Objectives (Levin, 1986; Gagné,
1986). Unfortunately, Ieamers are often unaware of the specific Ieaming objectives.
Such lack of clarity may hinder the selection of an appropriate Ieaming strategy. In the
following section, factors which affect the Ieamer's selection of a teaming strategy are
investigated.

Individual and Contextual Factors
This section considers the factors leading to individual differences in the

allocation of effort and use of different types of effort. Researchers have found

25

evidence of roles for motivation, Ieaming styles, and other personality constructs in the
Ieaming process (Biggs, 1992; Schmeck, 1982; Dweck, 1986; Kanfer and Ackerrnan,
1989; Kanfer, 1990). Several researchers have called for greater attention towards an
interaction perspective on Ieaming (Gagné, 1984; Biggs, 1993), which addresses
influences from multiple sources on the Ieaming process. Biggs (1993) also advocates
this interactive perspective, cautioning researchers about the dangers of isolating
teaming in a ‘vacuum,’ as context aids in the explanation of the effects of Ieamer
motivation. Greater specificity in the description of this interaction between the Ieamer
and his or her environment will be useful (Gagné, 1984).

In the section below, aspects of both the Ieamer and his/her environment will be
reviewed. One individual difference variable which impacts the teaming process is
teaming motivation. Contextually, the impact of the Ieaming orientation and the

teamers’ knowledge of the type of assessment will be investigated.

Learning Motivation

Researchers have identified a variety of motivational states related to the
Ieaming process. Dweck and her colleagues (Dweck, 1986; Dweck and Leggett, 1988)
have identified two primary teaming motivations; performance and mastery. Wrth a
mastery motivation, the Ieamer is dedicated to increasing his/her competence with the
task. The mastery Ieamer is motivated to team the task, regardless of the amount of
effort required, or how he or she may look on the interim performance tasks. The
mastery Ieamer is motivated to actually team the task, and acquire new knowledge or
skills. Similarly, Biggs (1993) describes a Ieamer who takes a deep approach to

teaming. This Ieamer focuses on the meaning of the task, with the desire to perform

26
the task properly and maximize understanding. Thus, the mastery goal and deep
approach are task-focused.

tn the performance goal situation, the Ieamer is motivated to perform well. The
performance goal Ieamer is concemed about appearing competent at all times (Dweck,
1986). Thus, the performance oriented Ieamer is concerned with non-task functions.
Biggs (1993) discusses two types of non-task Ieaming motives. The first is the
achieving orientation, which is very similar to Dweck’s performance goat. Students or
trainees who take this approach to teaming are interested in gaining recognition from
performance. Biggs’ second approach is surface teaming, where the Ieamer attempts
to satisfice the task demands. The Ieamer who takes this approach will seek to use
minimal time and effort on the task, but still meet the requirements of the Ieaming
situation. Biggs’ achieving and surface teaming orientations share a focus on off-task
Ieaming motivations. Wrth these orientations, the Ieamer is not attempting to satisfy
his/her desire to team, but rather attain external rewards of recognition and/or
evaluation.

These approaches to Ieaming motivation can be summarized into two
categories; task/mastery and ego/social (Farr, Hofman 8 Ringenbach, 1993; Meece,
Blumenfetd 8. Hoyle, 1988). Task/mastery includes the task-focused orientations;
Dweck’s mastery goal, and Biggs’ deep Ieaming approach. The ego/social motivation
encompasses Dweck’s performance goal, and Biggs’ surface and achieving
orientations.

Kanfer and Ackerrnan’s (1989) resource allocation model is useful to
hypothesize how these teaming motivations may affect the effort dedicated to the task,

and thus the Ieaming outcomes. Learners with an ego/social motivation allocate

27

resources towards ego maintenance and extemal recognition as well as skill
acquisition. With the task/mastery motivation, all resources are allocated towards skill
acquisition/teaming. A second distinction between the two goal types is the perception
of the role of effort in teaming. The ego/social Ieamer believes that effort and ability
are inversely related (Dweck & Leggett, 1988). If you have to try hard to team
something, you must have low ability. The task/mastery Ieamer believes that effort and
ability are positively related. If you really try hard, you can increase your ability. In
terms of cognitive resource allocation, the ego/social Ieamer is at a double
disadvantage; not only are cognitive resources removed from the task for ego
maintenance purposes, but effort is also decreased because of its perceived
relationship with ability (Button, et al., 1994).
Awareness of Learning Outcomes

Learning strategies may be initiated either by intemal factors such as Ieaming
motivation, or by extemal factors in the teaming environment (Rigney, 1978).
Awareness of the type of Ieaming outcome should serve as a signaling factor to the
Ieamer. It is an aspect of the task environment which provides relevant information
which can direct the Ieamer towards the appropriate Ieaming strategy. Situational
demands of the teaming environment have been demonstrated to influence Ieaming
behaviors (Meece, et al., 1988). In an ambiguous environment, the Ieamer will not
have such information, and will rely more heavily on internal factors such as motivation
to select the type and amount of effort.

The awareness of the Ieaming outcome may function in much the same manner
as advance organizers. As stated by Ausubel (1968), ‘...the principal function of the

organizer is to bridge the gap between what the Ieamer already knows, and what he

28
needs to know before he can successfully team the task at hand” (p. 148). Advance
organizers typically introduce the Ieamer to the content of the to-be-Ieamed material.
Knowledge of the Ieaming outcome can orient the Ieamer towards the process
necessary to learn the material.

Gagné, et al., (1992) provide an example of how the teaming outcome can
guide the selection and use of a Ieaming strategy. Suppose the primary teaming
objective is for trainees to understand the concept of an electric circuit. Two possible
outcome measures of this objective are 1) to state verbally what an electric circuit is,
and 2) to assemble an electric circuit. Each outcome measure gives some indication of
the trainee’s understanding of an electric circuit. However, the specific learning
objective will require the use of different Ieaming strategies (Gagné, et al., 1992). The
verbal statement of “what a circuit is’ would require only declarative knowledge, which
could be acquired by rote memorization. The assembling Of a circuit would require
application, or the use of rules concerning current flow, which could better be acquired
by organization or conceptualization strategies.

Learning strategies are a useful operationalization of the encoding processes
utilized in Ieaming. Rate memorization or rehearsal strategies represent data driven,
sequential processing. Personalizing or conceptualizing strategies represent holistic,
simultaneous processing. These strategies are differentially useful for acquiring
different types of teaming outcomes. As suggested by Gagné, et al., (1992), informing
the Ieamer of the objective allows him/her to select “particular strategies appropriate to

the teaming task and its expected outcome” (p. 189).

29
Summary: The Role of Effort in Lea_rpipg

Kanfer, Ackerman and colleagues (Kanfer, 1990; Kanfer 8. Ackennan, 1989;
Kanfer, et al., 1994) have expanded our understanding of the role of effort in skill
acquisition. The Integrative Resource Model considers cognitive ability and motivation
simultaneously, and examines their combined effects on skill acquisition and skill
performance. The model posits that cognitive ability limits the amount of effort which
can be devoted to any one task, and that proximal motivational processes are used to
allocate effort between on-task, off-task, and regulatory processes (Kanfer, 1990). The
model suggests that the allocation of attentional effort to regulatory processes such as
goal setting early in skill acquisition decreases performance, because goal setting
reduces the amount of effort which can be devoted to on-task processes. However,
empirical support for the direct role of attentional effort in the acquisition Of knowledge
and skill is lacking in studies testing this model.

There is evidence that the use of goal manipulations in early skill acquisition
reduces complex task performance (Kanfer & Ackerrnan, 1989). Kanfer and Ackerrnan
(1989) also provide evidence to link cognitive ability and task performance. Few explicit
links have been made, though, between attentional effort and performance. In their
third experiment, Kanfer and Ackerrnan (1989) manipulated the type of initial teaming
of the trainees by providing trainees with procedural or declarative part-Ieaming tasks.
The stated intention of this manipulation was to alter the attentional requirements of the
training. While the trainees in the declarative part-task condition did have increased
knowledge of the rules of the Air Traffic Controller (ATC) task, the training condition did
not have a direct effect on either the on-task or Off-task attention of the trainees.

Kanfer, et al., (1994) reported a significant relationship between goats and attention,

30
and a significant relationship between goals and performance. They reported no
evidence directly relating attention to performance.

Thus, something in the ATC task studies other than the amount of attentional
effort directed towards the task must have affected the teaming outcomes. The
transfer appropriate processing literature (e.g. Morris, et al., 1977) suggests that the
type of effort applied to the teaming task affects performance on Ieaming outcomes.
Kanfer and Ackerrnan have looked at the distribution of effort across general categories
of on-task, off-task, and self regulatory processes. it is the author’s belief that the
investigation of effort in skill acquisition and teaming needs to go beyond general
categories, and look into the “black box” of on-task effort.

In Kanfer and Ackerman’s ATC training task, trainees who had teamed with the
declarative part-task condition performed better on the knowledge outcome than did
trainees in the procedural part-task condition. This result is consistent with the themes
of retrieval congruent encoding processes. Research from Tulving (1983), and Morris,
Bransford and Franks (1977) has demonstrated the importance of matching encoding
and retrieval processes. This paradigm provides suggestions as to why the type of
cognitive effort used in Ieaming is important. Two trainees who put forth the same
amount of effort, but use different types of encoding processes or Ieaming strategies,
may perform differently on measures of teaming. Similarly, one trainee may perform
differently on two measures of Ieaming because of the type of effort used during
Ieaming.

A Motivational Model of Lea_mer Effort
The model of the role of effort in skill acquisition and Ieaming proposed here

incorporates elements of Kanfer and Ackennan’s (1989) Integrated Resource Model

31
with evidence from cognitive and instructional psychology concerning encoding-retrieval
interaction. The conceptual model is presented in Figure 1. Learner effort during the
teaming process plays a primary role in the model, as both the amount of effort and
type of effort are examined. In the instructional psychology literature, there have been
several renditions of a systematic model of teaming which includes various
representations of the teaming environment, Ieamer characteristics, and task demands
(e.g. Das, 1988; Biggs, 1993).

Tire primary Ieamer characteristic in the model is teaming motivation. The
Ieaming environment is represented in the model by Ieamer knowledge of the type of
Ieaming outcome. Finally, the type of teaming outcome is suggested to be a task
demand characteristic which influences Ieamer success.

Learner Effort

The two components of effort in the model are 1) amount of effort and 2) type of
effort. Amount of effort is defined as the amount of cognitive, or attentional, resources
devoted to the teaming task. The type of effort is defined as the form of cognitive
processing used during the encoding stage of the Ieaming task. Both the amount and
type of effort used by the Ieamer are proposed to affect performance on teaming
outcomes. Greater amounts of effort devoted to Ieaming are suggested to lead to
better performance on a Ieaming outcome. The type of effort suggested to lead to
better performance on a teaming outcome is dependent on the type of Ieaming
outcome. As highlighted by Morris, et al., (1977) and Schmidt and Bjork (1992), the
cognitive processing used during teaming should match the type of processing required

by the teaming outcome measure.

32
Learning Outcomes
Two types of teaming outcomes are considered in the model. A knowledge
outcome requires the Ieamer to know things, such as facts or isolated pieces of
information. An application outcome requires the Ieamer to do things. This distinction
is based on several typologies of training and Ieaming outcomes (Bloom, 1956; Gagné,
et al., 1992; Kraiger, Ford & Salas, 1993), as well as Anderson’s (1982) ACT“ theory.

Antecegents of Learner Effort

 

Biggs (1993) has developed a systems model of classroom instruction which
includes environmental variables such as the teacher, the teaming materials and the
method of evaluation, and student variables such as interests and motivation. Both
environmental variables and student variables are hypothesized to lead to changes in
task processing, such as the types of cognitive processes used. Similarly, Das (1988)
has suggested that the type of processing used in teaming depends on task demands,
Ieamer preferences, and an interaction between the two. In the current model, one
internal student factor (teaming motivation) and one environmental factor (knowledge of
learning outcome) will be considered.

Learning Motivation Kanfer and Ackennan (1989) suggest that motivational
processes drive the allocation of attentional effort in skill acquisition. Thus, the primary
learner or trainee factor in this model is teaming motivation, which represents an
individual’s approach towards teaming. Wrth a task/mastery motivation, Ieamers strive
to increase their understanding of something new or their competence at a particular
activity (Dweck, 1986; Button, et al., 1994). With an ego/social motivation, Ieamers

strive to protect their ago by performing well on Ieaming indicators. Thus, the

33
task/mastery motivation is intemalty focused, while the ego/social motivation focuses
on extemal issues such as appearances.

In the proposed model, a task/mastery motivation is suggested to direct the
Ieamer towards strategies which are related to a deeper understanding Of new material,
such as the personalizing or organizing strategies. The task/mastery motivation is also
suggested to cause the Ieamer to devote greater effort to the Ieaming task.
Attemately, the ego/social motivation is suggested to direct the Ieamer towards
strategies which will allow the Ieamer to perform best on the Ieaming assessment test.
The ego/social motivation will also cause the Ieamer to devote less effort to the
Ieaming task, as the Ieamer will have increased off-task thoughts. A desire to
demonstrate competence through minimal time on task will also reduce the amount of
effort put forth by the ego/social Ieamer (Dweck, 1986; Button, et al., 1994).

A_wa_reness of Learning Outcomes The awareness of the type of teaming
outcome is expected to affect the amount and type of effort put forth by the Ieamer.
Biggs (1992) suggests that assessment methods affect Ieamer processing. Students’
perceptions of the forthcoming assessment procedures, accurate or not, will alter
students” approaches to teaming. Gagné, et al., (1992) suggest that the specification
of Ieaming objectives allows the instructor to match the instructional approach to the
desired teaming outcome. The same should hold true for the Ieamer; the specification
of the desired Ieaming outcome, and the method of measuring that outcome, allows the
Ieamer to adjust his/her approach to Ieaming.

D’Ydewalle and Swerts (1980) instructed students to study short history texts,
and informed them they would take either a multiple choice test or a test with open-

ended questions. They found that students who took the type of test they had

34

expected performed better than students who received the unexpected test.
D'Ydewalle and Swerts (1980) hypothesized that students who had expected the open-
ended questions studied “more intensively” than students who had expected the
multiple choice test. While both types of tests in this study were forms of declarative
knowledge, the findings lend support to the proposed links between awareness of
Ieaming outcomes, effort, and performance.
mgheses

The hypotheses in this section have been derived from the model presented in
Figure 1. In the first section of hypotheses, a moderating role for type of effort is
proposed. It is hypothesized that the relationship between amount of effort and
learning outcome performance is moderated by the type of Ieaming effort utilized (see
Figure 2). The second section of hypotheses details the individual links in the model
which affect the amount and type Of effort . This operational model of effort in Ieaming
and skill acquisition is presented in Figure 3. Third, the mediator relationships implicit in
the model are discussed. Finally, the role of cognitive ability in the model is discussed.

Moderating role of type of effort

Research indicates that more time on the task and greater on-task attention
lead to higher performance on a Ieaming outcome (e.g. d'Ydewalle 8. Swerts, 1980).
However, this relationship should be moderated by type of effort (see Figure 2).
According to Kanfer (1990), motivational processes can fail to affect performance if an
individual misdirects his/her effort. In other words, not all effort is equal. A Ieamer
must utilize the right type of effort. In the current model, both amount of effort and type

of effort are posited to affect performance on teaming outcome tasks. The use of the

35

 

 

 

 

 

 

 

 

 

Learning

4 Outcome

 

 

 

Learning
Motivation
\’ Effort
-Amount
-Type
Awareness
of
Learning
Outcomes

 

 

Performance

 

 

 

Figure 1: Conceptual Model Of the Role of Effort in Learning and Skill Acquisition

 

Amount of Effort

0 Time on Task

. Off task attention
. Mental Workload

 

 

 

l
Type of Effort I,
o Rehearsal
. Personalizing
o Organizing

 

 

 

giggle; Proposed Moderator Relationship

V

 

Leaming Outcome

Performance
. Knowledge
0 Application

 

 

 

36
appropriate, or congruent, type of effort is expected to lead to a more successful
Ieaming outcome.

In Anderson’s (1982) ACT“ framework, rehearsal Ieaming strategies facilitate
the acquisition of declarative knowledge. Facts are studied and retained. Further
processing is required to facilitate the acquisition of procedural knowledge. The facts,
or declarative knowledge, must be compiled into propositions, or knowledge
representations which focus on the relationships between facts. The use of rehearsal
Ieaming strategies facilitates performance on a declarative teaming outcome. For a
procedural outcome, the Ieamer must go beyond rehearsal strategies, and incorporate
personalizing and/or organizing strategies as well (see Figure 3). It is suggested that
the use of rehearsal strategies allows the Ieamer to acquire declarative knowledge, but
one or both of the more integrative strategies must also be used to acquire procedural
knowledge, or the ability to apply the knowledge. More specifically, high performance
on an application outcome requires the use of personalizing and/or organizing
strategies. In any Ieaming situation, a Ieamer may use any of the three Ieaming
strategies. Only strategies congment with the Ieaming outcome measure are expected
to impact performance on the outcome measure.

Wrthin type of effort, amount of effort is suggested to impact the Ieaming
outcome. When a Ieamer uses a rehearsal strategy to prepare for a knowledge test,
he or she must devote some amount of effort to the learning task. Users of Ieaming
strategies that are congruent with the Ieaming outcome, who exert greater effort
towards teaming, will perform better than those who exert little effort. Users of
incongruent Ieaming strategies will not perform well on the teaming outcomes,

regardless of the amount of effort.

37

Rehearsal Strategies Integrative/Personalizing Strategies
LO. LO.
Perfor- Perfor- , - '
mance mance . x ' '
a I o
o I ‘

 

 

 

 

 

ILO HI LO HI
Amount of Effort Amount of Effort
Knowledge Outcome Application Outcome

 

 

 

 

Figure 3: Hypothesized Relationships Among Type of Effort, Amount of
Effort, and Learning Outcome Performance

38

H1: Type of effort is hypothesized to moderate the relationship between
amount Of effort, as measured by time on task and on-task attention measures,
and performance on teaming outcomes. Specifically, amount of effort will be
strongly related to performance when a congruent strategy is used (i.e.
rehearsal strategies for a knowledge test, or personalizing/organizing strategies
for an application test). When an incongruent strategy is used (i.e. only
rehearsal strategies for an application test), amount of effort will have little
impact on performance. When personalizing/organizing strategies are used for
a knowledge test, amount Of effort will have a moderate impact on outcome

performance.

Antecedents of Learner Effort
The following section presents hypotheses regarding the effects of Ieaming motivation
and awareness of Ieaming outcomes on both amount and type of effort. The links
among these variables are displayed in Figure 4.
Link I: Learning motivation —> amount of effort

The model posits that teaming motivation impacts the amount of effort devoted
to the teaming task. Trainee Ieaming motivation leads to decisions of how much time
will be devoted to a task such as skill acquisition (Kanfer, 1990; Kanfer 8. Ackerrnan,
1989). Motivation also affects the distribution of attention between on task and off task
activities (Button, et al., 1994). Since an individual’s attentional resources are limited
(Kanfer 8: Ackerman, 1989), each Ieamer makes choices concerning how the resources

are allocated. A high ego/social motivation will lead to decreased effort because

 

 

Learning Motivation
0 Task/mastery
orientation
. Ego/social

orientation

 

 

 

Knowledge of
Learning Outcome
0 Knowledge

0 Application

0 Ambiguous

 

 

 

39

Figure 4: Antecedents of Learner Effort

 

Amount of Effort

0 Time on Task

. Off-task Attention
0 Mental Workload

 

 

 

Type of Effort

0 Rehearsal

o Personalizing
o Organizing

 

 

 

 

 

40

attentional resources are diverted to ego-protection processes; Ieamers focus on task
difficulty and normative evaluations of their performance (Fan, et al., 1993). In
addition, a high ego/social motivation is associated with a state notion of ability (Dweck,
1986). White task/mastery motivated individuals believe ability can be improved
through effort, high ego/social individuals tend to believe that high effort is an indication
of tow ability. Therefore, Ieamers with a high ego/social motivation will try to protect the
ego by minimizing the amount of effort devoted to the task (see Figure 5).

H23: Learners with a high task/mastery motivation will be likely to devote

high amounts of on-task effort, as measured by time on task, off-task attention,

and mental workload. Task/mastery orientation will have a direct, positive effect

on amount of effort.

H2b: Learners with a high ego/social motivation will be likely to devote very little

on-task effort, as measured by time on task, off-task attention, and mental

workload. Individuals with a high ego/social motivation are predicted to have

high off-task attention. Ego/social motivation will have a direct, negative effect

on amount of effort.

H2c: Learners with low task/mastery and low ego/social motivations are

predicted to devote little on-task effort, as measured by time on task, off-task

attention, and mental workload. These individuals will not have the positive

motivation of the high task/mastery individuals, but neither will they experience

the distracting effects of high off-task attention.

H2d: Learners with high task/mastery and high ego/social motivations will be

likely to devote a moderate amount of on-task effort, as measured by time on

41

 

 

 

EsklmanenLorientation
Ego/social Low High
Orientation
Low effort High effort
Low
High Very low effort Moderate effort

 

 

 

 

Figure 5. Hypothesized relationships among Ieaming orientations and
amount of effort.

42
task, Off-task attention, and mental workload. These Ieamers will devote
substantial effort to the task, but will continue to devote resources to ego
maintenance. Thus, the net result will be more effort than either of the low
task/mastery categories, but not as much as those high in task/mastery
orientation and low in ego/social.
Link 2: Leaming motivation -> type of effort

Meece, et al., (1988) have suggested that teaming motivation can directly
influence the selection Of teaming strategies. Individuals with a high ego/social
motivation are primarily concerned with doing well on the Ieaming outcome task, and
should be more likely to select a strategy that will lead to high performance on the
outcome. Wrth a declarative Ieaming outcome, the ego/social motivated Ieamer will
use rehearsal strategies. With an application outcome, the ego/social motivated
Ieamer will use primarily personalizing and organizing strategies.

The learner who is motivated primarily by task/mastery, regardless of the
awareness of a declarative task, will process deeply because of the overriding
motivation to really team the material. In the ambiguous outcome situation, ego/social
motivated Ieamers should use a strategy at which they are competent to protect the
ego. Schmeck (1988) suggests that mastery Ieamers are more likely to use
conceptualizing and personalizing strategies.

H3a: Task/mastery motivated Ieamers are predicted to use organizing or

personalizing strategies. There will be a direct, positive relationship between

task/mastery motivation and organizing and personalizing strategies.

H3b: Learners with a high ego/social motivation are predicted to adapt to the

demands of the outcome task. V\fith a knowledge outcome, they would focus on

43
rehearsal. Vtfith an application outcome, they would use a personalizing or
organizing strategy. Thus, there is no direct relationship predicted between the

teaming strategy and Ieaming motivation for high ego/social Ieamers.

Li_n_k 3: Aaareness of outcome -—Jamount of effort

Procedural knowledge requires the acquisition of initial declarative knowledge,
and then the compilation of that knowledge into proposition form (Anderson, 1983).
Thus, acquisition of declarative knowledge should require less effort than acquisition of
application type knowledge. From Bloom’s (1956) perspective, an application outcome
is more complex than a knowledge outcome. Thus, it could be expected to require
greater effort to attain the higher level outcome.

H4a: Awareness of an application outcome is predicted to require greater

investment of on-task effort, as measured by time, off-task attention, and mental

workload.

H4b: The awareness of a knowledge outcome is predicted to require less effort

investment, as the Ieamer must only go through one stage of acquisition.

Lirrk 4: Awareness of outcome --> type of effort

It has been suggested that prior knowledge of Ieaming outcomes will allow the
Ieamer to utilize a teaming strategy, or type of effort, which is congruent with the
Ieaming outcome. (Barnett, et al., 1981; Gagné, 1984). Awareness of an application or
knowledge test at the end of the training session should allow the Ieamer to orient his
or her teaming strategies toward that outcome. The Ieamer should use rehearsal

strategies to acquire the basic, factual knowledge, as well as an organizing or

44

personal'rzing approach to create the links between pieces of information. While
mastery oriented Ieamers are predicted to atvvays use organizing and personalizing
strategies, they may use these complex strategies to a greater extent in the face of an
application test. They may also use more rehearsal strategies than they would
normally use in the face of a knowledge test.

H5a: Awareness of a knowledge test is predicted to be associated with Ieamer

strategies of rote memorization and rehearsal.

H5b: Awareness of an application test at the end of the training session will be

associated with the use of rehearsal strategies, as well as an integrative

strategy and/or a personalizing strategy.

Mediator Relationshjpa

In the organizing model (Figure 1), teaming motivation and awareness of
Ieaming outcomes are depicted as affecting outcome performance only through their
effects on amount Of effort and type of effort. Thus, the effort constmcts serve as
mediators in the conceptual model. However, type of effort is also proposed to
moderate the relationship between amount of effort and outcome performance.
Initially, only the mediation role of amount of effort will be tested. The mediation role for
type of effort will only be tested if the moderator tests fail. The model presented in
Figure 1 likely contains errors of specification (James & Brett, 1984). There are
unmeasured variables which potentially are relevant to the model, such as previous
experience with, and interest in, the subject area. However, the inclusion of these

additional variables is beyond the scope of the present study. Thus, the mediation

45

relationships must be investigated in an exploratory manner, and results interpreted
cautiously (James & Brett, 1984).
Cpgnitive abil'y

While cognitive ability has not been specifically addressed in the above model
or hypotheses, it surely plays a role in the Ieaming process. To eliminate one source of
specification enor, cognitive ability will be included in this research. Kanfer and
Ackerrnan’s resource allocation model suggests that cognitive ability limits the amount
of effort available for any given task. Kanfer and Ackerrnan (1989) found a relationship
between ability level and allocation of effort. Specifically, they found that low ability
subjects devoted more effort to off-task activities such as spontaneous goal setting and
negative self-thoughts. Kanfer and Ackerrnan have also suggested that cognitive
ability is a significant factor in performance during eariy stages of skill acquisition.

Other researchers have suggested a relationship between cognitive ability and
cognitive flexibility. This link would indicate that high ability individuals are more likely
to adapt to the situation and use an appropriate Ieaming strategy. In a review of the
advance organizer literature, Mayer (1979) concluded that high ability Ieamers did not
benefit from the use of an advance organizer because they were already good at
assimilative Ieaming, which was deemed most appropriate for the tasks involved in the
advance organizer literature. Thus, Mayer suggested that high ability Ieamers are
better able to adapt to the demands of the task, and use a more appropriate Ieaming
strategy.

However, other researchers (Dweck 8. Leggett, 1988; Meece, et al., 1988)
have found no relationship between cognitive ability and teaming goat orientation. In

addition, cognitive ability is not related to the experimentally manipulated variable,

46
awareness of Ieaming outcomes, in this study. It is clear that cognitive ability should
play some role in the proposed model. However, it could affect several different
relationships in the model. Thus, this research will examine relationships between
cognitive ability and amount of effort, type of effort, and Ieaming outcome performance
in an exploratory manner. Cognitive ability will be covaried out of the regression

analyses.

METHOD
Participants

The participants in this study were 121 undergraduate students recruited from
the Psychology Department subject pool at Michigan State University. The use of 121
subjects allows for power of .80, with eight independent variables, and alpha level of
.05, assuming a medium effect s'ze (Cohen, 1992). The inclusion of three interaction
terms in the moderator regression analyses substantially decreases the power of the
tests. Due to resource constraints, the sample size was held at 121, and significance
tests for the moderated regression will be interpreted cautiously. Effect sizes will also
be examined closely due to the low power for these tests.

Independent Variables

Cognitive Ability - The 50 item Wonderlic Personnel Test was used to measure
general cognitive ability (see Appendix A). The Wonderlic is an individually
administered pencil and paper test with a 12 minute time limit. Test-retest reliability
estimates for the Wonderlic range from 82-94, and the internal consistency reliability
(KR-20) is estimated at .88 (User's Manual for the WPT and SLE, 1992).

Learning Motivation - Two 8 item measures of Ieaming and performance goat
orientations developed by Button, Mathteu and Zajac (1994) were used to measure
Ieaming motivation (see Appendix B). Structural equation modeling evidence supports
the existence of two distinguishable dimensions, Ieaming goal orientation and

performance goat orientation, between which exists a non-significant correlation. The

47

48
Ieaming goal, or task/mastery orientation, scale has intemal consistency reliability
ranging from .79 - .85 (Button, et al., 1994). A sample item from the Ieaming goal
orientation scale is: The opportunity to team new things is important to me. The
performance goal, or ego/social, orientation scale has internal consistency reliability
ranging from .68 - .82. A sample item for this scale is: I like to work on tasks that I
have done well on in the past. A five-point rating scale ranging from (1) = “Strongly
Disagree” to (5) = “Strongly Agree” is used for both Learning Motivation measures.

Awareness of Learning Outcomes - Participants were randomly divided into
three groups regarding the awareness of the type of Ieaming outcome. One third of the
participants were informed at the beginning of the experiment that they would take a
multiple choice test after they teamed the materials. This was the knowledge, or
declarative, condition. One third of the participants were told they would take a test
consisting of ten problems. This was the application condition. The remaining third of
the participants were simply told they would take a short test on the materials. This
was the ambiguous condition, and served as a control.

Amount of Effort - The amount of effort devoted to the Ieaming task by the
participants was measured in three ways. First, a measure of time spent Ieaming (in
minutes) was taken. Second, a 13 item measure of Off-task Attention, adapted from
Kanfer, et. al., (1994) was used to measure the amount of off-task mental effort (see
Appendix C). Kanfer, et at.’s (1989) Off-task thoughts scale had six items concerning
mental activities such as daydreaming and loss of interest while performing the task.
This scale had an internal consistency reliability of .59. The 4 item Affective thoughts
scale contained items concerning negative self-evaluations during task performance.

This scale had an internal consistency of .78.

49

The Off-task thoughts and Affective scales were modified for the purposes of
this study. The two scales were combined to form the Off-task Attention scale. The
distinction between these scales is not relevant in this study, as affective thoughts can
be categorized as off-task effort; mental effort not directed explicitly towards teaming
the materials. Three additional items were written to reflect thoughts about both the
Ieaming task and the future performance task, for a total of 13 items. Increasing the
number of off-task attention items increased the internal consistency reliability of the
measure to .87. A sample item from the Off-task Attention scale is: I took ‘mental
breaks’ white teaming. A five-point rating scale ranging from (1) = “Never’ to (5) =
“Constantly” is used for the Off-task Attention measure.

The third measure of amount Of effort is mental workload. This measure was
adapted from the NASA-TLX scale (Hart & Staveland, 1988). A six item scale was
written to reflect relevant dimensions from the set often rating scales used by Hart and
Staveland. Items regarding physical effort were discarded, as were items which were
conceptually too similar to off-task attention. Relevant dimensions include mental
demand, perceived effort, mental fatigue, and an overall rating of perceived workload.
Test-retest reliability for the NASA-TLX has been estimated at .83 (Hart & Staveland,
1988). A five-point rating scale ranging from (1) = “Strongly Agree” 10(5) = “Strongly
Disagree” is used for the Mental Workload measure. A sample item for this scale is
“Learning the stock price prediction materials required a lot of mental activity.“

Type of Effort - The type of cognitive effort used in the Ieaming task by
participants was measured in two ways. First, participants responded to a 17 item
questionnaire concerning the Ieaming strategies they used during the experiment (see

Appendix D). Each strategy (rehearsal/memorizing, elaboration/personalizing, and

50
organizing/conceptualizing) is comprised of multiple tactics, or specific behaviors which
can be used to accomplish the Ieaming objective (Schmeck, 1988; Gagné, et al.,
1992). Trainees can accomplish each teaming strategy through a variety of specific
behaviors.

The Learning Strategy scale for this study was adapted from the Inventory of
Learning Processes (ILP), created by Schmeck (1983). Internal consistency estimates
for the ILP scales ranged from .58 for the Fact Retention scale to .82 for the Deep
Processing scale. Test-retest reliabilities for the ILP scales ranged from .79-.88
(Schmeck, 1983). The ILP includes many items directed toward the general study
habits and capabilities of students. These items were not used in this study. Only
items pertaining to the “on-line processing” (Biggs, 1993) used during Ieaming were
retained. Items from the ILP deep processing factor (Schmeck, 1983) represent the
organizing/conceptualizing strategy. Items from the methodical study factor represent
rehearsal, and items from the elaborative processing factor represent the
elaboration/personalizing strategy. These classifications are supported by Schmeck
(1988). Several additional items were added to the Learning Strategy scale based on
examples Of teaming strategies provided by Gagné, et al., (1992).

The Learning Strategy scale contains five items representing rehearsal, or
memorizing strategies. This category consists of verbal or mental repetition, with a
focus on specific details. A sample item is: ltried to remember exact words or phrases
used in the materials. There were six items representing each of the remaining two
Ieaming strategies. Activities in the elaboration/personalizing strategy include
paraphrasing and generating questions with answers. A sample item is: I created my

own examples. Organizing/conceptualizing tactics include organizing material into a

51
chart or diagram, or making lists of related ideas. A sample item is: I made lists of
associated ideas. A five-point rating scale ranging from (1) = “Never” to (5) =
“Constantly“ is used for the Learning Strategies measure. Learning strategies will be
treated as three distinct variables. It is possible that one Ieamer could use all three
strategies, or limit him/herself to just one of the strategies.

Second, participant Ieaming materials were examined for written evidence of
teaming strategies used by the participant. A similar method was used by Howard-
Rose and Wrnne (1994), in which traces, or written evidence of particular cognitions,
were coded into 10 components with an interrater agreement (Kappa coefficient) of .75.
The traces in this study were content coded into only three components; rehearsal,
elaboration, and organizing, using a three point scale ranging from 0 (no evidence of
this strategy) to 2 (strong evidence of this strategy). Traces were coded by the author
and two trained undergraduate students. Work done on the sample problem provided
in the teaming materials was also coded. If the participant had correctly completed the
problem, that was scored as a 2. Attempted but incorrect problems were scored as a 1,
while unattempted problems were scored as 0.

Dependent Vafiplas

Knowledge Learning Outcome - The knowledge Ieaming outcome is an 18 item
multiple choice test, with five options per item (see Appendix E). This test requires
participants to recognize the correct factual response to a question about stock prices
or general regression. These items were developed directly from the teaming
materials, and focus on facts which were found in the text of those materials. A sample

item from this measure is: The companies list their stocks on the following stock

52
exchange: a) NYSE b) OTC c) NASDAQ d) CBOT e) AMEX. The knowledge
outcome measure was scored in number of correct answers.

Application Learning Outcome - The application Ieaming outcome required
participants to predict stock prices of ten fictional companies, using the performance
data for each division of each company and the mles for determining the values for
beta weights for each term (see Appendix F). These rules forced participants to
choose between three possible beta weights for each term, depending on the value of
the performance term. Participants were provided with data for the quarterly
performance of three divisions of the companies. Participants were then required to
apply multiple regression procedures to estimate the future price of the stock. The
problems varied in difficulty. Some of the problems required use Of only one beta value
rule, some problems required use of two rules, and others required use of all three
rules. Several problems presented the participant with performance data from four
divisions, and helshe was to select the three correct divisions. The application outcome
was also scored in number of correct answers.

Learning Task

The Ieaming materials (see Appendix G) were based on the task used by
Eartey, Connoty and Ekegren (1989). Eartey, et al., (1989) asked participants to predict
stock prices for 100 fictional companies, given performance data for three divisions in
each company. Feedback was given to the participants, in the form of the correct stock
price, as they tried to improve their prediction accuracy. In the version of the Stock
Price Prediction task used in this study, participants were asked to team the prediction

method (multiple regression) prior to the Ieaming outcome test.

53

The task required participants to read a one page fictional description of how
investment counselors make stock price predictions for their clients. The second page
of Ieaming materials details how multiple regression could be used to predict stock
prices. Multiple regression is explained in non-technical terms, and a brief example is
provided. Both pages include many facts, some of which relate to the actual prediction
of stock prices, and others which are to be tested in the knowledge teaming outcome.

This task provided several desirable conditions for testing the hypotheses in
this study. First, the participants in this study, undergraduate psychology students,
were unlikely to be familiar with the content area of the task. They should not have
been able to answer the knowledge items or successfully predict stock prices from prior
knowledge. Thus, in accordance with the eariier discussion of boundary conditions for
skill acquisition tasks (see page 7), the Stock Prediction task should have placed the
participants in the early stages of teaming. Second, the stock prediction task allowed
for variance in both amount of effort and type of effort. Each participant was free to
use whatever methods of Ieaming helshe desired, and spend as much time on the
Ieaming segment as helshe desired. The task was considered to be resource-
dependent, where success on the Ieaming outcome task depended to some extent on
the effort allocated to the Ieaming portion of the task.

Third, there were clear distinctions between the knowledge and application
learning outcome measures. In the knowledge outcome, the participants recognized a
variety of simple facts concerning stock prediction and multiple regression. In the
application outcome, participants selected relevant information using a series of rules,
and performed multiple regression on performance data from fictional companies.

Thus, high performance on the application outcome required use of the rules found in

54
the teaming materials in a new situation. Stock price prediction is a higher order nrle as
described by Gagné, et at. (1992). Successful prediction requires a combination of
several simpler rules, to. the rules for a regression equation, beta weight selection and
relevant performance data. It is one thing to recognize the answer to a question about
the definition of a beta weight. It is quite another matter to know and apply the rules, or
relations among facts, to produce an accurate stock price prediction.

_E_xpe_rimental Procefla

Participants first took the Wonderlic Personnel Test, and the Learning
Orientation Questionnaire. Participants were then placed in rooms individually, and
were instructed to take as much time as they needed to team the Stock Prediction Task
materials. Isolating participants from one another during the teaming portion of the
experiment was intended to reduce the chances of contamination of amount of effort as
measured by minutes spent Ieaming. If the participants were in a group setting during
Ieaming, they could have taken cues from one another concerning time spent teaming.
Keeping participants separate at this point was intended to increase the between
subjects variance on amount of effort. Participants individually indicated when they felt
they had sufficiently teamed the material. The length of time elapsed during Ieaming,
in minutes, was recorded for each participant.

Upon completion of the Ieaming portion, subjects completed the Off-task
Attention measure, the Learning Strategies measure, and the Mental Workload
measure. Completion of the effort measures at this time was intended to allow the
participants to most easily and accurately report their thoughts and mental processes
during the teaming of the stock prediction materials. In addition, participant reactions to

the Learning Outcome measures could not affect participant motivation to respond

55
accurately to the effort measures. Following the effort measures, participants were

given the learning outcome measures.

Awareness of Learning Outcome Conatjtions
Participants were informed prior to beginning the task that upon completion of

the Ieaming segment, they would either 1) take an 18 item multiple choice test
(knowledge condition) 2) predict stock prices for 10 companies (application condition)
or 3) take a short test (ambiguous condition). All participants were given both tests;
the order of the tests depended on the experimental condition. In the knowledge and
application conditions, the announced test was be given first, followed immediately by
the other test. In the ambiguous instruction condition, the order of the tests was
randomly counterbalanced across the participants in that cell.
Pilot Study

A pilot study was conducted prior to the full study. Participants were 38
undergraduates recruited from the Psychology Department subject pool at Michigan
State University. The pilot study was conducted in three groups of approximately 10-15
subjects. All subjects in each group were assigned to the same experimental condition.
Subjects in the pilot study were not asked to complete the Wonderlic Personnel Test or
the Learning Motivation questionnaire. In addition, three subjects in each group were
randomly selected to participate in a structured interview after the Ieaming outcome
measures had been completed.

The pilot study was used primarily to investigate the strength of the Awareness
of Learning Outcome manipulation, and the degree to which the manipulation affects

participants” attitudes toward the second teaming outcome. It is possible that asking

56
participants to complete a second Ieaming outcome, one that is not the type of test for
which they had prepared, could create negative affect and reduce motivation to
perform well on the second Learning Outcome measure. Data collected in the post
experimental interviews indicated this was not the case. None of the interviewees
indicated any negative reactions to the two tests.

The pilot study was also used to evaluate the completeness of the Learning
Strategy questionnaire. It is possible that participants would use specific Ieaming
tactics which were not listed on the questionnaire. While there is an Open-ended
question on the Learning Strategy questionnaire, the author believes participants are
more likely to respond to an item listed on a questionnaire than write in their own
response. As a result of the pilot study, several items were dropped because of low
variability, and two additional items were written for the rehearsal subscate.

As a result of the pilot study, additional material was added to the task, and
items were added to the knowledge outcome. A fifth distracter was also added to each
item. The application outcome appeared to be an adequate measure, so no changes

were made. Results of the pilot study are presented in Appendix I.

RESULTS

The results of this study are reported in two parts. First, the adequacy of the
measures is examined. This includes factor analyses, reliability of the measures,
content coding of relevant variables, and lntercorrelations of the variables. Second, the
hypotheses of the study are tested using a series of univariate multiple regression
procedures.

Adegaacv of Meaaures

Factor Analyses. Factor analysis was performed on the measures of Ieaming
orientation and type of effort. The teaming orientation measure was hypothesized to
contain two distinct factors; task/mastery motivation and ego/social (performance)
motivation. A common factor analysis (Principal factors extraction, Oblimin rotation)
supported this structure. The eigenvalues and factor loadings are presented in Table
1. The results of the factor analysis support the hypothesized two factor structure of
the orientation measures. Factor 1 represents ego/social orientation, and Factor 2
represents task/mastery orientation. A third factor had an eigenvalue of less than 1.00
(.96), and was discarded.

The type of effort questionnaire ts hypothesized to measure three different
Ieaming strategies; rehearsal, personalizing, and organizing. Thus, three factors are
expected. No specific pattern of correlations is expected among the three factors. A

common factor analysis (Principal factors extraction, Oblimin rotation) supported this

57

58
stmcture. The eigenvalues and factor loadings are presented in Table 2. The results of
the factor analysis supports the three factor structure of the teaming strategies
questionnaire. Factor 1 primarily represents the personalizing strategy. One item from
the organizing strategy loads most highly on this factor, but none of the loadings for the
item are strong. Factor 2 represents the rehearsal strategy. Factor 3 represents the
organizing strategy. A fourth factor was discarded because the eigenvalue was less
than 1.00 (.88).

The two self-report measures of amount of effort, Mental Workload and Off-task
Attention, were factor analyzed. Given the strong correlation between the variables, it
is possible that these two measures are tapping into the same constnrct. A common
factor analysis (Principal factors extraction, Oblimin rotation) supported the existence of
two separate constructs. Eigenvalues and factor loadings are presented in Table 3.

The first factor has an eigenvalue of 6.1, and measures Off-task Attention. The
second factor has an eigenvalue of 1.97, and represents Mental Workload. A third
factor has an eigenvalue of 1.48. By Kaiser’s rule of retaining all factors with
eigenvalues over 1.00, this third factor should be retained. This third factor would
contain five items from the Off-task Attention scale; items 6, 7, 8, 9, and 10. These are
all items which deal with thoughts completely unrelated to the experimental setting.

The Off-task Attention scale was intended to capture any non-Ieaming related
thoughts. No distinction was made between test-related thoughts and daydreaming, for
instance. The factor analysis demonstrates that the Off-task Attention items are distinct
from the Mental Workload items. The two factor solution presented in Table 3 reduces
the number of negative factor loadings, and is a more parsimonious solution. Thus, the

two factor solution will be retained for the purposes of this research.

59
Table 1

Factor Loadings for Learning Orientation Scales

Factor 1 Factor 2
Eigenvalue 3.9 2.9
Leam9 .73 .03
Leam12 .72 -.12
Leam16 .71 .03
Leam14 .70 -.12
Leam1 1 .69 -.12
Leam10 .65 .03
Leam13 .63 -.14
Leam15 .50 .08
Leam5 -.02 .74
Leam6 .08 .68
Leam7 -.20 .67
Leam1 -.04 .65
Leam3 .25 .61
Leam8 .07 .57
Leam4 -.21 .48
Leam2 -.1 1 .32

Note:
Items Leam1 to Leam8 were intended to measure Task/mastery orientation, and items
Leam9 to Leam16 were intended to measure Ego/social orientation.

60

Table 2

Factor Loadings for Learning Strategy Scales

Factor 1 Factor 2 Factor 3

Eigenvalues 3.5 1 .88 1 .3

P6 .74 .05 -.02
P2 .73 .03 .23
P5 .72 .05 .03
P1 .52 -.01 -.09
P4 .48 -.1 1 -.32
P3 .45 -.09 -.25
O1 .27 .05 -.21
R5 -.05 .76 .03
R1 .03 .64 -.04
R2 -.13 .59 -.1 1
R4 .10 .53 .09
R3 .04 .48 -.04
03 -. 14 .14 -.87
O4 -.1 1 -.05 -.84
05 .15 .12 -.38
02 .12 .03 -.38
O6 .16 -.01 -.27

Note:

R = item from rehearsal scale
0 = item from organizing scale
P = item from personalizing scale

61
Table 3

Factor Loadings for Off-task Attention and Mental Workload Scales

Factor 1 Factor 2
Eigenvalue 6.1 1 .9
A12 .68 -.09
A7 .68 .19
A8 .67 -.12
A10 .66 .02
A9 .62 .06
A2 .57 .09
A1 .54 .04
A4 .53 -.17
A3 .51 -.06
A5 .48 -.01
A11 .48 -.39
A13 .46 -.19
A6 .30 -.19
WL6 -.01 .81
WL2 -.14 .76
WL3 .06 .70
WL4 -.17 .68
WL5 .14 .63
WL1 -.28 .57

Note:
A = item from off-task attention scale
W = item from mental workload scale

62

lntercorrelations and Reliabilities. Correlations among all independent and
dependent variables are presented in Table 4. Summed scale scores were used to
calculate correlations. All scales are coded such that a high score reflects a larger
amount of that construct. lntemal consistency reliabilities (coefficient alpha) for all
scales appear on the diagonal of the matrix. All reliabilities were in the acceptable
range (above .70) with the exception of the knowledge test. The reliability of this 18
item, dichotomousty scored test was .65. There is no reliability estimate for the
measure of time. The reliability for the cognitive ability measure, the Wonderlic, was
based on information from the test publisher (“Wonderlic Personnel Test,” 1992).

While it was expected that the two dependent variables, knowledge and
application tests, were distinct Ieaming outcomes, the two measures were positively
correlated (r = .36, p<.01). The reliability of the knowledge test was also less than
desirable (a=.65). The correlation corrected for this unreliability increases to .47. The
correlation was not significantly attenuated by partialltng out cognitive ability (r = .32,
p<.01).

Correlations among the amount of effort variables ranged from .02 to -.48. Off-
task attention and workload were most strongly related, white time and workload were
unrelated. Correlations among the type of effort variables ranged from .07 to .40.
Personalizing and organizing were most strongly related, white rehearsal and
personalizing were minimally related.

Cognitive ability was associated with few of the variables of interest. Cognitive
ability was negatively correlated with time spent Ieaming, and positively related to
mental workload, but was unrelated to Off-task attention. It was positively correlated

with performance on both the knowledge test (r = .27) and the application test (r = .23).

63
Cognitive ability was not related to task/mastery or ego/social orientation, nor was it
related to the use Of Ieaming strategies.

Content Coding. In addition to the self-report measures of teaming strategies,
four measures were coded from the participants” Ieaming materials. First, rehearsal,
organizing and personalizing strategies were coded from the markings made by the
participants on their materials. The intenater reliability of these measures with three
raters was acceptable (rehearsal .89, organizing .91, and personalizing .74). However,
these measures did not demonstrate acceptable construct validity. The coded
measures correlated much more highly among themselves than with the corresponding
self-report measures. The correlation matrix is presented in Table 5. In addition, the
patterns of correlations with other variables differ substantially. For example, the
knowledge test correlates positively with the self-reports of both organizing and
personalizing, and does not correlate with rehearsal. In contrast, the knowledge test is
moderately correlated with the rehearsal trace, and not significantly correlated with
organizing or personalizing. It appears that the trace measures may be tapping a
different constmct, such as tendency to write while teaming. Because of the lack of
evidence for validity as cognitive Ieaming strategies, these measures were not used in
further analyses.

The final measure of Ieaming strategies was the coding of completion and
accuracy of the sample problem. Working the sample problem was positively

associated with performance on the application test. It was not associated with any of

4.00.0 A“ mom_0 5880:1303 man NEE—0:32..”

 

5.2020 2.00: m0 4 m w A m m u m o 8 3 Am
A. .00: 0:03 afim +3 38

 

 

N. 2.0m. 0103 and 93 Lo :8

 

 

w. .230 3b #3 Lo .5 |
A. 02.84...» web 3: .wA. 53.. La 33
one:

 

 

m.<<o_.x_oma Nob. PS A: .NE. .0» ..Am fimdi

 

m._~m:mm8m_ 3.0 who but .3 .0m Au Jom 6V8

 

NOGDENSQ 8..» #3 .0w .3 .mm.. ..:r .._w .No. AN:

 

 

 

 

@0283..- BE tram awe» .No. .3 Inc. .8. .3 38.. 38

N5

o. <<018a #3 .3 .8 :3 .mo. 33 .MM. .8 .2 .8 .t

manic

3.5.92680 3N Num ..8.. .wm. .3. sum. .8. .3 .NA. he... .2 33
050030

2500:0020: Pa 98 now .8 .om saw. .3. in .8 .3 Lot .8. .m8
050030

 

ANOOQ>U==< mm} mgm be. .om ANN» .on Am» )3 now AS .5» .MN» .Mm. .l

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.3330. 003.333 $2329.. 60030.03 05:3 03 08332. .3 =6 $3030.. mac—scam ma :2 05.520 34 :30 0..
Saxon 330.0 “080.03. mm 53‘ 03 9.6.0 :03 303.509 90:3 0020.020: 8030.036 03 REESE? «633003 R =5
.8 .05..

Correlation Matrix for Learning Strategy Traces and Self-reports

65

Table 5

 

 

 

 

 

 

 

 

 

 

1 2 3 4 5 6 7
1. Organize
2. OTrace .25*
3. Personalize .40Mr .15
4. PTrace .48“ .57** .18*
5. Rehearsal .20* .06 .07 -.01
6. RTrace .17 .61” -.05 .44” .10
7. Knowledge .24* .14 .24* .15 .01 .19*
8. Application .05 .07 .17 -.1O -.12 -.00 .36*

 

 

 

 

 

 

 

 

n=121

** significant at .01
* significant at .05

 

 

66

the other Ieaming strategy measures, nor was it related to performance on the
knowledge test.
flypgthesis Testing

Moderated Regressions. Hypothesis 1 suggests that the type of effort will
moderate the relationship between amount of effort and Ieaming outcome
performance. Specifically, amount of effort was not expected to impact Ieaming unless
that effort was congruent with the cognitive demands of the Ieaming outcome.
Separate regressions were performed on each dependent variable; knowledge
outcome and application outcome. First, to control for the effects of cognitive ability on
performance, cognitive ability was entered as a covariate. Type of effort and amount of
effort were entered on the second and third steps. Type of effort consists of four
variables. Rehearsal, personalizing, and organizing strategies are treated
independently. Working the sample problem was included as a Ieaming strategy.
Amount of effort consists of three variables; time on task (measured in minutes), off-
task attention, and mental workload. Finally, the product terms of amount of effort (3
variables) and type of effort (4 variables) were entered, for a total of 6 direct effects and
8 product terms. Off-task attention was not included in the interaction terms.
Theoretically, off-task attention should have no effect on the strategies used. It is not
effort applied to the task. Thus, it could not interact with the strategies to affect the
outcomes.

For the application outcome, cognitive ability was entered in the first step, and
produced a significant R2 of .05 (p<.01). In the second step, the four Ieaming strategies
(type of effort) were entered, and produced a significant change in R2 of .17 (p < .01).

Only Sample resulted in a significant final standard'zed beta weight (.30, p< .01). In the

67
third step, the three amount of effort variables were entered, and produced a significant
change in R2 (.09, p<.01). Workload (.21, p < .05) and Off-task attention (-.15, p < .05)
had significant final standardized beta weights. In steps four and five, the interactions
were entered. The product terms account for an insignificant amount of unique
variance. It should be noted that the standardized beta weights for personalizing X
workload, and rehearsal X time, were significant at the .10 level. However, there is little
evidence for the proposed moderator relationship. The full results are presented in
Table 6. The total amount of variance accounted for was .39.

For the knowledge outcome, cognitive ability was entered in the first step, and
produced a significant R2 of .07. In the second step, the four Ieaming strategies (type of
effort) were entered, and produced a significant change in R2 (.09, p<.01). although
none of the beta weights were individually significant. In the third step, the three
amount of effort variables were entered, and produced a significant change in R2 (.19,
p<.01). Each of the three variables made a significant contribution. In steps four and
five, the interactions were entered. The product terms account for no unique variance.
Thus, there is no evidence for the proposed moderator relationship. The full results are
presented in Table 7. Hypothesis 1 is not supported regarding the knowledge outcome.
The total amount of variance accounted for was .39.

Antecedents of Effort. To investigate the antecedents of both amount of effort
and type of effort, regression analyses were performed using amount of effort and type
of effort as separate dependent variables. Cognitive ability was entered first in the
analyses as a covariate. Learning motivation and knowledge of Ieaming outcome were
entered on the second step. Learning motivation was treated as two independent

variables, task/mastery orientation and ego/social orientation. Knowledge of Ieaming

68

Table 6

Moderated Regression Analysis Results on Application Outcome

 

 

Variable Beta R R2 R2
Change
Step 1:
Cognitive Ability .13 .23 .05 .05“
Step 2:
Organize -.07
Personalize .1 1
Rehearsal -.09
Sample .30“ .47 .22 .17**
Step 3:
Workload .21*
Time .05
Off-task Atten. -.15" .56 .31 .09"
Step 4:
O X W .01
P x w -.97“
R X W -.14
S X W -.08 .59 .35 .04
Step 5:
O X T .42
P X T -.84
R X T 1.13*
S X T -.19 .62 .39 .04
n=121

Note: Standardized Beta weights are used

** significant at .01
* significant at .05
# significant at .10

69

Table 7

Moderated Regression Analysis Results on Knowledge Outcome

 

 

Variable Beta R R2 R2
Change
Step 1:
Cognitive Ability .36” .27 .07 .07**
Step 2:
Organize .04
Personalize .1 1
Rehearsal .07
Sample -.12 .41 .17 .09*
Step 3:
Workload .18*
Time .34*
Off-task Atten. -.22* .60 .37 .20“
Step 4:
O X W .06
P X W -.62
R X W .90
S X W .06 .62 .39 .02
Step 5:
O X T -.17
P X T -.21
R X T .44
S X T .12 .63 .39 .00
n=121

Note: Standardized Beta weights are used

** significant at .01
* significant at .05
# significant at .10

70

Hypotheses H2a-H2d suggest several relationships between motivation and
amount of effort. Specifically, a task/mastery orientation will be positively related to
amount of effort, while an ego/social orientation will be negatively related to amount of
effort. An interaction between the two orientations was also hypothesized, which
suggested that the matched cells (high-high or low-low) would result in more moderate
amounts of effort (see Figure 5).

Mastery orientation and workload are positively correlated (r =.21), and a
negative relationship was found between mastery orientation and off-task attention
(r = -.17). These two correlations support H2a. Correlations also indicate a positive
relationship between ego/social orientation and off-task attention (r =.34), supporting
H2b. Separate regression analyses with the amount of effort measures as dependent
variables indicated that an ego/social orientation was a significant predictor of off-task
attention (b=.33, p<.01) and mastery orientation was a significant predictor of workload
(b=.19, p<.05). The interaction terms were not significant for any of the three measures
of amount of effort. Thus, there were only direct effects of Ieaming orientation on
amount of effort.

The third set of hypotheses (H3a & H3b) suggests several relationships
between motivation and type of effort. Specifically, high task/mastery orientation will be
directly associated with the use of organizing and/or personalizing strategies, while
ego/social orientation will interact with the awareness of Ieaming outcome to predict
use of Ieaming strategies. Correlations between the variables indicate positive
relationships between performance orientation and rehearsal (r = .23) and mastery
orientation and personalizing (r = .29). A negative relationship was found between

ego/social orientation and personalizing (r = -.20).

71
Table 8

Regression Analysis Results on Off-task Attention

Variable Beta R R2 R2

 

Change
Step 1:
Cognitive Ability .01 .02 .00 .00

Step 2:
Performance
orientation (P) .33“ .34 .12 .12**

Step 3:
Mastety
orientation (M) -.15 .37 .14 .02#

Step 4:

Knowledge

Condition -.04

Application

Condition -.03 .37 .14 .00

Step 5:

PXM -.14 .37 .14 .00
n=121

** significant at .01

* significant at .05
# significant at .10

Variable

Regression Analysis Results on Time

Beta

72

Table 9

R

R2

R2

 

Step 1:
Cognitive Ability

Step 2:
Performance
orientation (P)

Step 3:
Mastery
orientation (M)

Step 4:
Knowledge
Condition
Application
Condition

Step 5:
P X M

n=121
** significant at .01

* significant at .05
# significant at .10

-.24*

-.07

.12

-.21*

-.12

-.24

.24

.27

.33

.33

.05

.06

.07

.11

.11

Change
.05*

.01

.02

.03

.00

Variable

Step 1:
Cognitive Ability

Step 2:
Performance
orientation (P)

Step 3:
Mastery
orientation (M)

Step 4:
Knowledge
Condition
Application
Condition

Step 5:
P X M

n=121
** significant at .01

* significant at .05
# significant at .10

73

Table 10

Regression Analysis Results on Workload

Beta

.16#

-.10

.18*

-.08

-.07

.65

.17

.22

.29

.30

.26

R2

.03

.05

.08

.09

.07

Change
.03*

.01

.03*

.01

.00

74
outcomes was treated as two van'ables, dummy coded for the analysis. The results of
these regressions are presented in Tables 8-14, and are discussed below.

Separate regression analyses with the type of effort measures as outcome
variables indicated that mastery orientation (b=.28, p<.01) and ego/social on'entation
(b= -.18, p<.01) significantly predicted personalizing, and the beta weights are in
opposite directions. Ego/social orientation also predicted use of rehearsal strategies
(b=.24, p<.01). These findings support Hypothesis 3a. Hypothesis 3b is not supported.
No interaction effects were found regarding ego/social orientation; only direct effects
were found. The full results are presented in Tables 11-14.

The fourth set of hypotheses (H4a and H4b) suggests several relationships
between awareness of Ieaming outcome and amount of effort. Awareness of a
knowledge outcome was expected to be negatively associated with amount of effort,
while awareness of an application outcome was expected to be positively associated
with amount of effort. Separate regression analyses with the amount of effort
measures as outcome variables indicated that the knowledge test condition significantly
predicted time spent studying (b= -.23, p<.01), providing partial support for H4b. No
other regression terms were significant.

The fifth set of hypotheses (H5a and H5b) suggests several relationships
between knowledge of Ieaming outcomes and type of effort. Awareness of a
knowledge outcome was expected to be associated with the use of rehearsal
strategies, while awareness of an application outcome was expected to be associated
with the use of personalizing or organizing strategies. Separate regression analyses
with the type of effort measures as outcome van'ables indicated that awareness of the

knowledge condition was weakly associated with the use of rehearsal strategies (r =.17,

Variable

75

Table 11

Regression Analysis Results on Rehearsal

Beta

R

R2

R2
Change

 

Step 1:
Cognitive Ability

Step 2:
Performance
orientation (P)

Step 3:
Mastery
orientation (M)

Step 4:
Knowledge
Condition
Application
Condition

Step 5:
P X M

n=121

** significant at .01
* significant at .05
# significant at .10

-.14

.24“

.13

.17#

-.01

.46

.11

.26

.30

.34

.35

.01

.07

.07

.12

.12

.01

.06“

.02

.03

.00

Variable

76
Table 12

Regression Analysis Results on Organizing

Beta R R2 R2
Change

 

Step 1:
Cognitive Ability

Step 2:
Performance
orientation (P)

Step 3:
Mastery
orientation (M)

Step 4:
Knowledge
Condition
Application
Condition

Step 5:

P X M

n=121

** significant at .01

* significant at .05
# significant at .10

-.11 .08 .01 .01

.05 .09 .01 .00

.15 .18 .03 .02#

-.04

-.08 .19 .04 .00

77
Table 13

Regression Analysis Results on Personalizing

 

 

Variable Beta R R2 R2
Change

Step 1:
Cognitive Ability -.01 .00 .00 .00
Step 2:
Performance
orientation (P) -.18* .20 .04 .04*
Step 3:
Mastery
orientation (M) .28** .34 .11 .07**
Step 4:
Knowledge
Condition .07
Application
Condition .04 .34 .12 .00
Step 5:
P X M 1.09 .36 .13 .01
n=121

** significant at .01
* significant at .05
# significant at .10

78

Table 14

Regression Analysis Results on Worked Sample Problem

Variable

Step 1:
Cognitive Ability

Step 2:
Performance
onentation (P)

Step 3:
Mastery
orientation (M)

Step 4:
Knowledge
Condition
Application
Condition

Step 5:
P X M

n=121
** significant at .01

* significant at .05
# significant at .10

Beta

.21*

.01

-.08

-.10

.03

.45

R

.21

.21

.22

.25

.26

R2

.04

..04

.05

.06

.07

R2
Change

.04*

.00

.01

.01

.00

79
p<.10), but the change in R2 at this step was not significant. Awareness of the
application test was not associated with the use of any strategy. Hypothesis 5a is
somewhat supported, but H5b was not supported.

Mediator Analyses. As there was no support for the moderator relationship,
two mediator analyses were performed. Both the role of amount of effort and type of
effort were tested as mediating the relationship between Ieaming motivation and
performance. Mediator tests were appropriate only for the relationships involving the
knowledge test, as the application test was not correlated with either motivation scale
or the awareness of Ieaming outcome. The awareness of Ieaming outcomes was
included only in the mediator analysis involving amount of effort, as there was no direct
relationship between awareness of outcomes and type of effort.

The mediator relationships were tested with hierarchical regression analyses
using the knowledge test as the dependent variable. Correlations (see Table 4)
demonstrate the existence of relationships among the Ieaming orientations and the
knowledge outcome. Relationships were also demonstrated between the knowledge
outcome and all three measures of amount of effort, as well as organizing and
personalizing strategies. However, in the mediated regression analysis, there was no
evidence for mediator relationships. Wrth cognitive ability controlled for, significant
direct effects were found for the Ieaming orientations after entering effort into the
equation. A small effect was also found for knowledge of declarative condition.
Results of the mediated regression analyses are presented in Tables 15 and 16.

Full regression eguation. Given the lack of support for either the mediated or
moderated models, regression analyses were performed which included only direct

effects for all variables, with the application and knowledge tests as the dependent

80
variables. In the final equation, working the sample (b=.29, p<.01) and workload
(b=.23, p<.05) significantly predicted performance on the application outcome.
Knowledge of the application outcome was not a significant predictor, but the beta
weight was in the predicted direction (b=.15, p = .11). Results of the regression are
presented in Table 17. The total variance accounted for in the application outcome
was .33.

For the knowledge outcome, cognitive ability (b=.33, p<.01), time (b=.33, p< .01)
and task orientation (b=.19, p<.05) were significant predictors. Again, the beta weight
for experimental condition was not significant, but was in the predicted direction (b=.15,
p=.10). Several other variables nearly reached significance in the predicted direction,
including off-task attention (b=-.16, p<.10) and workload (b=.17, p<.10). Results of the

regression are presented in Table 18. The total variance accounted for was .42.

81

Table 15

Mediated Regression Analysis Results on
Knowledge Learning Outcome through Leaming Strategies

 

 

Variable Beta R R2 R2
Change

Step 1:
Cognitive Ability .27“ .27 .07 .07“
Step 2:
Organize .21*
Personalize .04
Rehearsal .01
Sample .05 .41 .17 .09“
Step 3:
Mastery Orient. .28**
Perf. Orient. -.17* .52 .27 .10“
n=121

Note: Final standardized Beta weights are presented.

** significant at .01
* significant at .05
# significant at .10

82

Table 16

Mediath Regression Analysis Results on
Knowledge Learning Outcome through Amount of Effort

 

 

Variable Beta R R2 R2
Change

Step 1:
Cognitive Ability .30“ .27 .07 .07“
Step 2:
Workload .16*
Time .33“
Off-task Atten. -.16* .57 .33 .26**
Step 3:
Mastery Orient. .22“
Perf. Orient. -.07 .62 .38 .05“
Step 4:
Knowledge
condition .16*
Application
condition -.03 .64 .41 .03”
n=121

Note: Final standardized Beta weights are presented.

** significant at .01
* significant at .05
# significant at .10

83
Table 17

Direct Effects Regression Analysis Results on
Application Learning Outcome

 

 

Variable Beta R R2 R2
Change
Step 1:
Cognitive Ability .15“ .23 .05 .05*
Step 2:
Knowledge
Condition .09
Application
Condition .15 .27 .07 .02
Step 3:
Mastery alient. -.02
Perf. orient. .03 .29 .08 .01
Step 4:
Organize -.O7
Personalize .12
Rehearsal -.10
Sample .29“ .49 .24 .16“
Step 5:
Workload .23*
Tlme .08
Off-task Atten. -.15 .58 .33 .09“
=121

Note: Final standard’zed Beta weights are presented.

** significant at .01
* significant at .05

# significant at .10

84

Table 18

Direct Effects Regression Analysis Results on

Knowledge Learning Outcome

 

 

Variable Beta R R2 R2
Change
Step 1:
Cognitive Ability .33“ .27 .07 .07"
Step 2:
Knowledge
Condition .15“
Application
Condition -.02 .31 .09 .02
Step 3:
Mastery orient. .19*
Perf. orient. -.07 .48 .23 .13*
Step 4:
Organize .06
Personalize .05
Rehearsal .03
Sample -.06 .53 .28 .06“
Step 5:
Workload .17“
Tlme .33“
Off-task Atten. -.16" .65 .42 .14"
n=121

Note: Final standardized Beta weights are presented.

** significant at .01
* significant at .05
# significant at .10

DISCUSSION

The purpose of the present study was to examine the role of both amount of
effort and type of cognitive effort in Ieaming. The effect of Ieamer awareness of the
type of Ieaming outcome on effort was also investigated. The discussion is divided into
three sections. First, the findings of the study are presented. As several of the
hypothesized relationships were not supported, alternative explanations are explored.
Second, limitations of the study are discussed, as well as the possible impact of those
limitations on the results. Finally, opportunities for further study of issues oonceming
effort in Ieaming are explored.

Summary of Findings

Learning Orientation. The data support the notion that task/mastery orientation
and performance, or ego/social, orientation are two distinct constructs, rather than
endpoints on a continuum. The correlation between the two variables was non-
significant (r = -.10). In the factor analysis, all items loaded as hypothesized on two
factors.

The hypotheses regarding Ieaming orientation stated that high task/mastery
orientation would lead to greater effort and use of more complex Ieaming strategies. It
was suggested that high ego/social orientation would lead to less on-task effort and the
use of simpler Ieaming strategies. Hypotheses were also made concerning patterns of

effort for the interaction among these two variables. The study provided partial support

85

86
for the direct effects of Ieaming orientation on effort. Task/mastery orientation was
positively associated with perceived mental workload and personalizing strategies, and
negatively related to off-task attention. Ego/social orientation was positively associated
with off-task attention and use of rehearsal strategies, and negatively correlated with
personalizing. In general, the direct effects were consistent with the hypotheses. The
hypotheses concerning the interaction between the orientations were not supported.
Thus, the orientations affect effort independently of one another. However, using off-
task attention as an example, the patterns of effort predicted by the regression
equation are consistent with the hypothesis. Recall that a high value on the off-task
attention scale is interpreted as less effort. If an individual is high on task/mastery
orientation and high on ego/social orientation, the prediction for amount of off-task
attention is slightly positive. Low scores on both would result in a negative
standardized off-task attention score. With high task/mastery and low ego/social
orientation, the equation would predict a negative value. With high ego/social and low
task/mastery orientation, the equation would produce a high negative value (see Table
19). Regardless, the interaction term for this particular effect produced zero change in
R2.

As the hypothesized mediator relationships were not statistically significant, the
direct effects of Ieaming orientation on outcome performance can be examined.
Neither orientation was directly related to the application outcome. However,
task/mastery orientation was positively related to performance on the knowledge test,
and the ego/social orientation was negatively related to performance. Finally, the
results of the study agree with the work of Dweck and colleagues, as there were no

significant relationships between cognitive ability and Ieaming motivation.

87
Table 19

Prediction of Off-task attention from Learning Orientation

Regression equation (from Table 8): A = .33 (P) - .15 (M)

 

 

 

 

 

 

Low Mastery-Low Perf mmmely-Low Perf
A = .33 (-2) - .15 (-2) A = .33(-2) - .15(2)

A = -.66 - (-.3) A = -.66 - .30

A = -.36 A = --93

MPerf-Low Mastenr High Perf-High Masteg
A = .33(2) - .15 (-2) A = .33(2) - .15(2)

A = .66 - (-.30) A = .66 - .30

A = .96 A = .38

 

 

 

Note: Standardized regression weights are used. Variable values are 2 standard
deviations above and below the mean. Positive values indicate high off-task attention;
negative values indicate low off-task attention.

 

88

Mness of Leaming Outcome. The hypotheses regarding awareness of the
Ieaming outcome (knowledge vs. application) suggested that participants would adjust
their Ieaming strategies and amount of effort depending on the test they expected. The
data show limited support for the importance of informing Ieamers of the type of
Ieaming outcome they can expect. While none of the relationships were statistically
significant, a few of the hypothesized relationships were nearly significant.

Participants who were told they would receive a knowledge based outcome tended to
spend less time Ieaming (b = -.21, p < .05), however, the change in R2 associated with
this beta weight was non-significant. Participants who were told they would receive a
knowledge based outcome were also more likely to use rehearsal Ieaming strategies (b
= .17). In each case, awareness of a knowledge based outcome test was associated
with reduced effort on the part of the Ieamers. This effect should be further examined,
given the limitations of the study. Perhaps the two outcomes were too similar, as both
were pencil and paper. The lack of support for these hypotheses could also be a result
of study habits. Students at large universities are accustomed to taking multiple choice
tests. The participants may simply have used the same study strategies they would
use in a typical classroom Ieaming situation.

Amount of Effort. The data suggest many interesting relationships among
measures of amount of effort. These constructs are related, but are not identical. Time
was not highly related to the self-report measure of mental workload. Off-task attention
and perceived mental workload are inversely related. Participants who were thinking of

things other than the task while Ieaming felt the leaming required less effort. As

89
expected, time and off-task attention were not related. These relationships reinforce
the need to measure amount of effort in multiple ways.

Each of the measures provided valuable information regarding one or more
hypotheses. Amount of effort was hypothesized to be indirectly related to the Ieaming
outcomes. It was suggested that the amount of effort expended would only matter if
the Ieaming strategies used were congruent with the Ieaming outcome. For example,
regardless of the amount of effort devoted to Ieaming, the use of rehearsal strategies
should not affect the application outcome. However, the mediated model was not
supported. All three measures of amount of effort were directly related to performance
on the knowledge outcome. Time was positively related to performance on the
knowledge test, suggesting that time on task is an important factor in memorization. In
contrast, off-task attention was negatively related to both the knowledge and
application outcomes. This suggests that regardless of the Ieaming outcome,
concentration is important, and focusing on external issues detracts from Ieaming.
Similarly, workload was positively related to both tests. If participants felt they had
devoted a great deal of mental effort to the task, they tended to perform well on the
outcome tests. In general, participants who worked harder performed better on the
Ieaming outcomes. This result is consistent with the position of Kanfer and Ackennan
(1989), that mental effort, or the devotion of attentional resources, is related to Ieaming
and task performance.

Leaming Strategies. Learning strategies were hypothes'zed to positively affect
congruent Ieaming outcomes. Use of rehearsal strategies was expected to lead to high
performance on the knowledge outcome, while high performance on the application

outcome was expected to require the use of organizing or personalizing strategies.

90

These hypotheses were not supported. In the final regression equation, none of the
cognitive Ieaming strategies were significant predictors of the knowledge outcome,
although both personalizing and organizing strategies were positively correlated with
the knowledge outcome. Working the sample problem was a significant predictor of
performance on the application test. Rehearsal strategies may have been unrelated to
performance because of low variance (x = 17.9, sd = 3.4).

The constnlct validity of the Ieaming strategy measures must be considered.
The self-report measures correlated in the expected pattern; rehearsal and organizing
had a low to moderate, positive correlation (r = .20, p<.05); rehearsal and
personalizing were not correlated; and organizing and personalizing had a strong,
positive correlation (r = .40, p< .01). The trace codings of Ieaming strategies were
more strongly correlated with one another than with the corresponding self-report
measures, displaying little discriminant validity. Thus, the self-report measures of the
Ieaming strategies appear to have greater use in distinguishing among the three types
of strategies. The trace codings of the strategies may be measuring a different
construct, such as tendency to write while Ieaming. This is a separate Ieaming strategy
than the ones which were examined in this study. Writing while Ieaming may be
considered a specific Ieaming tactic, rather than a cognitive Ieaming strategy.

However, there were clearly between-person differences on the traces. Many of
the Ieaming materials were clean, while others were filled with writing. All three traces
were positively related to performance on the knowledge test, with the correlation with
rehearsal traces reaching significance (r = .19, p<.05). In addition, the organizing and
personalizing traces were positively related to task/mastery orientation. This result is

consistent with the notion that task/mastery oriented individuals will use the more

91
complex strategies. The traces were not related to performance on the application test.
Working the sample, which was the other Ieaming strategy coded from the Ieaming
materials, was a strong predictor of performance on the application test.

The data also point to the amount of effort required for the different types of
Ieaming strategies. The use of organizing strategies was associated with greater time
spent Ieaming, and less off-task attention. Personalizing was associated with less off-
task attention, and greater perceived workload. Not surprisingly, performance oriented
individuals tended to use more rehearsal strategies, and fewer personalizing strategies.
In contrast, mastery oriented individuals tended to use more personalizing strategies.
The mastery oriented individuals put forth more effort during Ieaming, and chose
strategies which were more complex.

The study did not provide a great deal of explanation for why participants used
a particular Ieaming strategy. The percentage of variance explained in these variables
was low, ranging from .04-.13. Only cognitive ability was a significant predictor of
working the sample problem (b = .26, p< .01). Much of the literature regarding Ieaming
strategies focuses on how to teach various strategies to students. Perhaps the
participants in this study did not know how to use strategies other than rehearsal, which
many of them used. Ability to use particular strategies was not measured.

Working Sample PrgbLems. This study has also highlights the importance of
working sample problems when faced with a performance or application test.
Participants who successfully worked the sample problem included in the Ieaming
materials tended to score much higher on the application outcome test. Working the
sample problem had no effect on knowledge test performance. More simply stated,

practice on the required behavior while Ieaming improves post-Ieaming performance.

92
Since the knowledge test required recognition of facts concerning regression and stock
price prediction, rather than use of the regression formula, one would not expect
practice on a sample problem to improve knowledge test scores.

This result concurs with Anderson’s most recent data on the proceduralization of
skills (Anderson 8 Fincham, 1994). Anderson suggests that examples can provide a
direct linkage to proceduralization, without ever Ieaming a declarative representation of
the concept. Additionally, Paas (1992) suggested that the type of sample problem
affects outcomes of Ieaming. He found that Ieaming from already worked or partially
worked sample problems was positively related to test performance. Test problems
ranged from identical to the sample problems but with different numbers, to more
complex, less structured problems than the sample problems. Unfortunately, in the
current study, participants who did not work the sample problem were not asked if they
had studied the solution to the problem on the next page.

Role of cogmt'le a_blmy_ Cognitive ability was positively related to performance
on both outcome measures. This result supports the notion of general cognitive ability -
individuals with more 9 will perform better on a range of cognitive tasks. In contrast to
Kanfer and Ackerman’s suggestion that cognitive ability is negatively related to off-task
mental activities, cognitive ability was not related to off-task attention. Cognitive ability
was negatively related to time, and positively related to mental workload. Individuals
with high cognitive ability, as measured by the Wonderlic, spent less time on the
Ieaming task, but perceived using more mental effort. However, time and workload
were not related in the sample as a whole. Perhaps because these high ability
individuals concentrated their effort in a shorter period of time, they perceived that it

required more mental effort.

93
Study Limitations

The limitations of the study are discussed in relation to: (1) participant
characteristics, (2) measures used, (3) low power to detect significant effects, and (4)
tradeoffs associated with cognitive research.

Participant characteristics. The participants in this study were recruited through
the undergraduate subject pool at a large university. Thus, the participants were
relatively homogeneous in age, race, and background. The motivation of the
participants, however, varied widely. Many proceeded through the experiment as
quickly as possible. Receiving credit for the experiment was not contingent on their
performance in the experiment. Some wrote on their materials that this was the most
boring thing they had ever done. Others were stimulated to learn more about the stock
market, while still others remained after the experiment for several minutes to argue
about their scores, or apologize for poor performance. Informing participants at the
beginning of the experiment that their tests would be scored was intended to increase
the motivation of all participants to perform well. Clearly, it was not effective for
everyone.

Toward the end of the experiment, it was discovered that an unknown number
of subjects were cheating by writing relevant information on the tables in their individual
rooms. It would have been interesting to include cheating behaviors as a variable in
the study, both to covary cheating out of the performance measures, and to relate
cheating to Ieaming orientation. It seems likely that participants high on ego/social
orientation would be more likely to cheat.

Measures Used. While it was expected that the two Ieaming outcome

measures, knowledge and application, were distinct Ieaming outcomes, the two

94
measures were positively correlated (r = .36, p<.01). The reliability of the knowledge
test was also less than desirable (ot=.65). The correlation corrected for this unreliability
increases to .47. The correlation was not significantly attenuated by partialling out
cognitive ability. Thus, the two measures may have been tapping the same construct,
to a degree. The two tests did cover the same general content area, and both were
paper and pencil tests. These similarities likely contributed to the correlation. In
support of the distinction between the tests, the patterns of correlations between the
tests and other variables were quite different. For example, the knowledge test was
positively correlated with task/mastery orientation, and negatively correlated with
ego/social orientation. The application test was not significantly correlated with either
orientation.

Several of the effort measures used were written for this study. The Learning
Strategy measures, for example, were based on Schmeck’s (1983) Inventory of
Learning Processes (ILP). These scales had not been widely pretested. Thus, there is
little evidence supporting construct validity of these scales. In addition, the internal
consistency reliability was somewhat low for the Rehearsal (.73) and Organizing (.71)
subscales.

Power to Detect Significant Effects. Given the number of independent variables
and interaction terms entered in some of the regression analyses, the number of
participants (n=121) was too small to provide adequate power for detecting significant
effects. Indeed, several effects were in the predicted direction, but missed the
traditional significance levels. One of these effects was the manipulation of awareness
of Ieaming outcome type. The awareness of the knowledge outcome displayed effects

in the predicted direction on several occasions, but generally failed to reach

95
significance. Similarly, none of the tests for interactions found significant results.
Unfortunately, the number of participants was limited by available resources.

Meoffs Associgt_e_d with Cognitive Resea__rch_._ The study may have been
affected by several trade-offs which were made during the planning process. One of
these trade-offs was the type of task used. A task with greater complexity may have
produced results more similar to the hypotheses. Perhaps participants were not
challenged enough by the task to use complex Ieaming strategies. Perhaps the task
did not allow the use of complex Ieaming strategies. Cognitive ability may have an
even greater impact on more complex tasks. However, a relatively simple task was
chosen to reduce the time required by the participants.

The nested nature of types of Ieaming outcomes makes it difficult to separate
the effects of the knowledge and application measures. lf knowledge is required for
application, perhaps it should be covalied out of analyses. Clearly distinct testing may
help to separate the effects. A trade-off was made in the current study to measure
both types of outcomes using paper and pencil tests.

A third problem inherent in the study of cognitive processes is the lack of clear
definitions in the literature. While several concepts were consistent across authors and
studies, often the name and intricacies of the constructs varied widely. For instance,
the difference between examples of Ieaming styles, Ieaming strategies, and Ieaming
tactics was not always clear. This lack of consensus among previous researchers
increases the difficulty of writing scales which accurately reflect the nature of the

constnlct in question.

96
me Research Ogmrtunities

First, future research could address some of the limitations discussed above.
Regarding the Ieaming outcome measures, the knowledge test could be pencil and
paper, but the application test could be a role play or physical performance of a task. A
greater distinction between the tests may increase the effects of awareness of Ieaming
outcome on Ieamer effort. Learners may not distinguish between two pencil and paper
tests as they approach the Ieaming task.

Further work should be conducted exploring the construct validity of the
cognitive Ieaming strategy measures. Verbal protocol analysis during Ieaming could
give a more objective indication of what the Ieamers are actually doing. Fischer (1993)
analyzed videos of students in small group interaction to measure cognitive processes,
examining written transcripts of the Ieaming situation. Current self-report measures of
strategies rely on not only participants’ honesty, but also their accuracy. There is much
debate concerning how accurately people can use introspection to report their mental
processes. Adding to the difficulty of identifying one’s own mental processes is the
delay in completing the scales. In the current study, participants waited several
minutes before recording their processes. Mental processes may be more accurately
reported as they are occurring. Participants were not informed prior to the Ieaming that
they would be later asked to report their processing during Ieaming. Informing people
in advance may improve the accuracy of self-reported mental processes.

The trace codings of strategies used in the present study did not appear to
measure three different Ieaming strategies. However, the use of repetitious writing may
certainly be a valid Ieaming method. Three fairly general, cognitive Ieaming strategies

were examined in the study. Rather than attempting to measure the cognitive

97
processes behind Ieaming, a direct focus on the behaviors may be more fruitful. The
domain of specific Ieaming tactics, which comprise general Ieaming strategies
(Schmeck, 1983), should be further examined. The coding of whether or not the
participants had correctly completed the sample falls closer to a specific Ieaming tactic,
rather than a cognitive Ieaming strategy. Perhaps a focus on Ieaming behaviors would
be a useful direction for further research. A well-defined domain of Ieaming tactics may
lead to better specification of the general Ieaming strategies.

In addition to work on the measurement of Ieaming strategies, the ability to
select and use certain Ieaming strategies should be studied. The process used to
select Ieaming strategies may vary across Ieamers. The current study did not
substantially advance our understanding of this area. Selection of Ieaming strategies
has been considered a specific cognitive skill in itself, within the realm of metacognition
(Gagné, et al., 1992; Kraiger, Ford 8. Salas, 1993). The inclusion of various
metacognitive skills in future research may better predict which Ieaming strategies will
be used. Another avenue of research regarding strategy selection might be additional
individual characteristics. The Ieaming orientation measures and cognitive ability were
the best predictors of strategy use in this study. All Ieamers may not have access to
the same Ieaming strategies. Perhaps participants did not use personalizing strategies
because they were not aware it was an option. Schmeck (1983, 1988) has suggested
that people have different Ieaming styles which are consistent across situations.

Teaching Ieaming strategies could increase flexibility of strategy use across
situations. Schmeck (1988) suggests that students should be taught a range of
Ieaming tactics, as they will likely encounter a range of Ieaming outcomes. However,

students must also be taught when to use these strategies. Schmeck (1983) refers to a

98
training program which taught various strategies, but was not effective, as students did
not know when to apply these strategies. Relationships between Ieaming style and use
of specific Ieaming strategies could be investigated further.

The relationships among task/mastery orientation, ego/social orientation, and
additional Ieaming variables should be further examined. Interest in these constructs
has increased dramatically in l/O psychology in the past few years (e.g. Button, et al.,
1994; Boyle & Klimoski, 1995; Smith, Ford, Weissbein, & Gully, 1995). Smith, et al.,
(1995) discovered that Ieaming orientation was related to self-efficacy, which impacted
training performance. Boyle and Klimoski (1995) treated Ieaming orientation as both
state and trail variables. The trail orientation was measured in the current study, but
perhaps participants took on a different state orientation as a result of the specific task,
and this change went undetected. Similar to Ieaming style, the malleability of Ieaming
orientation should be studied. It may be that individuals who are more flexible in their
approach to Ieaming are more successful in Ieaming.

Finally, the long term effects of the amount and type of effort used during
Ieaming must be considered. High performance on a test at the end of a Ieaming
experience is irrelevant if that Ieaming is not transferred to other situations at other
times. It has been suggested that the same cognitive processes are not involved in
successful training performance and successful transfer of training (Schmidt & Bjork,
1992). An extension of the current study could consider the difference between the
Ieaming outcome at the end of training, and the test conditions present in the transfer
environment. It follows that cognitive processes required for the test at the end of

training should be as similar as possible to those processes required at transfer.

99
Future research on Ieaming motivation and Ieamer effort must address transfer of

Ieaming as well as initial Ieaming.

APPENDICES

 

100

APPENDIX A

Wonderlic Personnel Test
m

PERSONNEL TEST

FORM V

 

Social Security Number LLL-ZL-ICEEI

READ THIS PAGE CAREFULLY. DO EXACTLY AS YOU ARE TOLD.
DO NOT TURN OVER THIS PAGE UNTIL YOU ARE
INSTRUCT ED TO DO SO.

PROBLEMS MUST BE WORKED WITHOUT THE AID OF A CALCULATOR
OR OTHER PROBLEM-SOLVING DEVICE.

This is a test of problem solving abilty. it contains various types of questions. Below is a sample question correctly
filled in:

REAPistheopposlteof -
lobtaln. 2cheer. 3continue, 4exist. Sag .................................................

The correct answer is “sow”. (It is helpful to underlne the correct word.) The correct word is
numberedS. ThenwrltetheligureSinthebracketsattheendoftheline.

Answer the next sample question yourself. ‘ .j
Paper sells for23centsperpad. Whatwili4padscost? ............................................................ i
The correct answer Is 92¢. There. is nothing to underline so just place “92¢” in the brackets. U
Here is another example:

MINER MINOR— Do these words ,N
' 1 have similar meanings. 2 have contradictory meanings. 3 mean neither the same nor opposite? . . “"

 

The correct answer is “mean neither same nor opposite” which is number 3 so all you have to do
Is place a figure “3" in the brackets at the end of the line.

When the answer to a question is a letter or a number. put the letter or number in the brackets.
All letters should be printed.

This test contains 50 questions. It is unlikely that you will finish all of them. but do your best. After the examiner
tells you to begin, you will be given exactly 12 minutes to work as many as you can. Do not go so fast that you
make mistakes since you must try to get as many right as possible. The questions become Increasingly difficult.
sodonotsktpabout. Donotspendtoomuchtimeonanyoneproblem. Theexaminervnlnotansweranyquestlons
after the test begins.

Now, lay down your pencil and wait for the examiner to tell you to begin!

 

 

 

Donottum rhepageundlyOUONIOId‘OdO’a 946900724

 

u‘.

:ag'é.‘ {$1,- ‘-' "{‘y’ 3.. 73,-3.4. -,. . "

Z‘ , wum‘rm tmwmlepcrmmlrallne 6‘3”? , ._.. _

. KIM, e: ,3" “a; "'.“~"""‘...,z° i959ar. Wonderlic“fi;+° flaw N w h§$§N3~ 3er

. ':.. .MWWWmleIacalmN mmuwummm . 3‘11, .. ,J. 1".‘1’

hologrlgtbtybr‘tarsta'firm . panamquMBaMfi&fiéqammur

_ ‘electrenic‘lnicbuiicaL'W' Man'wilboutthepdorwrirtea’pcfnfisfoa MWmm 9311'
-5». 331:3?‘vmwmacauomawmmwmbtwatrium-mt wt .33

 

NI

 

 

101

 

 

congratulate»

10.
11.

12.
13.

14.
15.

16.

17.
18.
19.

. In the following set of words. which word is different from the others?

1 copper, 2 nickel. 3 aluminum. 4 wood. 5 bronze .................................................

. Which word below is related to bear as calf is to cow?

I chick. 2 cub. 3 fawn. 4 trout. 5 fox ..................................................................

. Most of the items below resemble each other. Which one is least like the others?

1 July. 2 February. 3 April. 4 Tuesday, 5 June .....................................................

. In 20 days a boy saved one dollar. What was his average daily saving? .........................................

HYPOCRITE HYSTERICAL—Do these words
I have similar meanings, 2 have contradictory meanings. 3 mean neither the same nor opposite?.

. Are the meanings of the following sentences: 1 similar, 2 contradictory, 3 neither similar nor

contradictory? Look before you leap. Think today and speak tomorrow. .......................................

. Assume the first 2 statements are true. Is the final one: 1 true, 2 false, 3 not certain? The flute.

lsintunewiththeharp. Theharplsintunewtththeviola. Theviolaisinttmewlththeflute. ..........

. In the following set of words, which word B different from the others?

1 beef. 2 mackerel, 3 veal, 4 bacon, 5 hot dog ......................................................

. Are the meanings of the following sentences: 1 similar, 2 contradictory, 3 neither similar nor

contradictory? Never look a gift horse in the mouth. You cannot make a silk purse out of a sow’s
ear ...................................................................................................................................
Most of the items below resemble each other. Which one B least like the others?

1 suspicion. 2 unbelief, 3 doubt, 4 resolve. 5 misgiving ............................................

SUPPORT is the opposite of
I maintain, 2 sustain. 3 cherish. 4 desert, 5 prop ...................................................

Assume thefirst2statementsmtrue. lsthefinalone: 1 true, 2 false. 3 notcertain?
These puppies are normal dog. All normal dogs are active. These puppies are active. ....................

How many of the five items listed below are exact dupficates of each other? ...................................
362 7363
62738 63738
918264 918264
1628357 1638357

Wire is 12.5 cents a foot. How many feet can you buy for a dollar? .............................................
DECEP'I'ION is the opposite of

1 falsehood. 2 trickery. 3 frankness. 4 finesse. 5 fabrication ......................................

Assume the first 2 statements m true. Is the final one: 1 true. 2 false. 3 not certain?
All red-headed boys like candy. Charles B red-headed. He likes candy. ........................................

A dealer bought some televisions for $2500. She sold them for $2900. making $50 on each
television. How many televisions were involved? .......................................................................
ABSURD ACCEDE—Do these words

1 have similar meanings, 2 have contradictory meanings, 3 mean neither the same nor opposite?.

Two of the following proverbs have similar meanings. Which ones are they? ...................................
;. \fiou catch more gu'wnh hhoney than with vinegar.
. e squ w ee e ease.
3. A fly toilet-swan hon m 9"
4. Sweet a sour w n we pay.
Too s urives as tardy as too slow.

20. In the followin set of words. which word B different from the others?
1 little. small, 3 tiny, 4 spacious, 5 precise .........................................................
21. ADORN is the op site of
1 garnish. ornament, 3 embellish, 4 bedeck, 5 deface ...........................................
22. Are the meanin of the following sentences: 1 similar, 2 contradictory. 3 neither similar nor
contradictory? ords are always bolder than deeds. Stab wounds heal, but bad words never. ...........
23. Two of the following proverbs have similar meanings. Which ones are they? ...................................
‘ 1. Once bitten. twice shy.
2. No one is happy all his life long.
3. Hitch your wagon to a star.
4. Fortune favors the brave.
5. All men have the same share of happiness.
24. A rectangular bin. completely filled. holds 640 cubic feet of grain. If the bin B 8 feet wide and 10 feet

long, how deep B it? ...........................................................................................................

. ANGIER is the opposite of
26.

2 vexation. 3 forbearance. 4 displeasure. 5 resentment ..................................

Assume the first 2 statements are true. Is the final one: 1 true, 2 false. 3 not certain?
These boys are normal children. All normal children are big eaters. These boys are big eaters. ..........

. A boy is 10 years old and his sister is twice as old. When the boy is 16 yem old. what will be the

age of his sister? .................................................................................................................

. Are the meanings of the following sentences: 1 similar. 2 contradictory. 3 neither similar nor

contradictory? All comedies are ended at marriage. The man who expecB comfort in this We must be
born deaf. dumb, and_blind. ............................................................. . ....................................

 

 

 

 

 

102

 

 

31.

32.

.‘K

. Which number in the following it of numbers represents the smallest amount?
.999 999 .9 1 2 98:83.? ................................................................................

. In the following set of words. which word is different from the others?

I odor, 2 scent. 3 sour. 4 spice. 5 fume .........................................................

. ABSCOND ABSENCE—Do these words

1 have similar meanings. 2 have contradictory meanings. 3 mean neither the same nor opposite?.

Fourhof the following 5 parts can be fitted together In such a way as to make a triangle. Which 4
are t ey. .....................................................................................................................

ii . W a /

RETREAT RETRIEVE—Do these words ‘
lhavesimllarmeanlngs. 2havecontradictorymeanings. 3meanneitherthesamenoropposlte?.

 

 

 

 

 

 

I

. Are the meanings of the following sentences: 1 similar, 2 contradictory. 3 neither similar

nor contradictory? A friend in need is a friend indeed. A faithful friend B a strong defense. .........

35. When the price of gasoline increased from 16.4 cents to 20.5 cents, what was the percent
increase in cost of the gasoline? ........................................................................................
36. APPEAL is the op site of
1 beseech, entreat. 3 request. 4 deny. 5 invoke .............................................
37. Two of the following proverbs have similar meanings. Which ones are they? .............................
1. Every effect becomes a cause.
2. The cautious seldom make mistakes
3. Two wll kin a Ion.
4.At ldcordBnotquicklybroken.
5. Waterfahrgdaybydaywearsthehardeststoneaway.
38. Suppou you arrange the following words so that make a complete sentence. If it B a true
sentence, mark (T) in the brackets; If false. put an in the brackets.
always Lightning follows thunder .........................................................................
39. A clock was exactly on time at noon on Monday. At 8 P.M. on Tuesday it was 64 seconds slow.
Atthatsamerate, how muchdiditloseln 16 hour? .............................................................
40. ENDURE B the opposite of
1 allow, 2 bear, 3 suffer. 4 sustain, 5 foil ........................................................
41.1fSV-bagsofseedcost $35.whatwill4‘/rbagscost? ...........................................................
42. Are the meanings of the following sentences: 1 similar. 2 con . 3 neither similar
nor contradictory? Politeness is excellent. but it does not pay the bill. B virtue. ..................
43

49.

. Which number in the folloxrb'lgNgé'oup of numbers represents the smallest amount?

222 .9 .73 2

.Ttheometricfigurecanbedividedbyastraightlineintotwopartswhichwillfittogetherina

certainwa tomakeaperfectsquare. DrawsuchalnebyjoinIngZofthenumbers. Thenwrlte
thesenum astheanswer. ..........................................................................................

 

ts
. Three individuals form a partnership and agree to divide the profits eq . X invests $6500. Y
invests $2000, and Z invests 31500. if the profits are 33000. how much does X receive than
if the profits \vere divided in proportion to the amount invested? .............................................
. What is the next number in this series? 16 4 1 .25 .....................................................
. Two of the following proverbs have similar meanings. Which ones are they? .............................

1. No one B wise all the time.
.AwordtothewlseBsufficient.

. The doors of wisdom as never shut.

. “TB wisdom sometimes to seem a fool.
. The geatest good is wisdom.

MPUN

. ADROIT ADEPT-Do these words

1 have similarmeanings. 2 have contradictory meanings, 3 mean neitherthe same noropposlte?.

In printing an article of 24.000 words. a printer decides to use two sizes of type. Using the larger
type, a printed page contains 1200 words. Using the smaller type, a page contains 1500
words. The article is allotted 17 full pages in a magazine. How many pages must be in the small

eeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee

. mmumber in the following series does not fit in with the pattern set by the others. What should

that number be? 1/1000 1/100 1/10 0 1/10 10 ..................................................

 

 

 

103

APPENDIX B

Learning Orientation Scales

This set of questions asks you to describe how you feel about each of the following
statements. Please use the scale shown below to make your ratings.

Strongly Strongly
Disagree Disagree Neutral Agree Agree
I

< I I I >
I I

(1) ('2) ('3) ('4) (5)

1. The opportunity to do challenging work is important to me.

I"

I do my best when I'm working on a fairly difficult task.
3. I try hard to improve on my past performance.

4. When I have difficulty solving a problem, I enjoy trying different approaches to see
which one will work.

5. The opportunity to Ieam new things is important to me.

6. The opportunity to extend the range of my abilities is important to me.

7. I prefer to work on tasks that force me to Ieam new things.

8. When I fail to complete a difficult task, I plan to try harder the next time.

9. The things I enjoy the most are the things I do the best.

10. I feel smart when I can do something better than most other people.

11. I like to be fairly confident that I can successfully perform a task before I attempt it.

12. I am happiest at work when I perform tasks on which I know that I won‘t make any
errors.

13. I feel smart when I do something without making any mistakes.
14. I prefer to do things that I can do well rather than things that I do poorly.

15. The opinions others have about how well I can do certain things are important to
me.

16. I like to work on tasks that I have done well on in the past.

 

104

APPENDIX C

Off-task Attention Scale

Please respond to the items below with the following scale:
1 = Never 2 = Seldom 3 = Occasionally 4 = Frequently 5 = Constantly

While I was learning the Stock Price Prediction Task material:

1. I thought about how much time I had spent Ieaming the material

A

 

 

 

 

 

 

1 2 3 4 5

Never Constantly
. I wondered about my performance compared with others

1 2 3 4 5

Never Constantly
. I wondered how well others have done on the test

1 2 3 4 5

Never Constantly
. I thought about how hard the material was to Ieam.

1 2 3 4 5

Never Constantly
. I wondered about how hard the test might be

1 2 3 4 5

Never Constantly
. I took “mental breaks’ while I was Ieaming

1 2 3 4 5

Never Constantly

105

7. I daydreamed while I was Ieaming

 

 

 

 

 

 

 

1 2 3 4 5

Never Constantly
8. I lost interest in Ieaming the material for short periods of time

1 2 3 4 5

Never Constantly
9. I thought about other things that l have to do today

1 2 3 4 5

Never Constantly
10. I let my mind wander while I was Ieaming the materials

1 2 3 4 5

Never Constantly
11. I became frustrated with my ability to Ieam the material

1 2 3 4 5

Never Constantly
12. I thought about how well or how poorly l was doing

1 2 3 4 5

Never Constantly
13. I got mad at myself while I was Ieaming the material

1 2 3 4 5

Never Constantly

106

APPENDIX D

Learning Strategy Scales
Please respond to the items below with the following scale:
1 = Never 2 = Seldom 3 = Occasionally 4 = Frequently 5 = Constantly
While I was Ieaming the Stock Price Prediction Task material:

1. I tried to memorize the facts.

 

2. I focused on remembering the details of the material. __

3. l repeated certain words or phrases to myself.

 

 

4. I repeated the regression formula to myself.
5. I tried to remember exact words or phrases used in the materials.

6. lcreated my own examples.

 

7. I related the regression equation to other formulas I know.

8. ltested myself on the material using my own questions.

9. I tried to express things in my own words.

 

10. I thought of similar concepts to which the material was related.
11. l leamed new ideas by associating them with words and ideas I already knew.
12. I tried to relate facts to other pieces of information found in the material.

13. I looked for conflicts or inconsistencies between pieces of information.

 

14. I made lists of associated ideas .

 

15. I made simple charts or diagrams to help relate ideas to one another.

 

16. I tried to organize the material.

17. I searched for general ideas in the material.

107

APPENDIX E

Mental Workload Scale

Please respond to the following items oonceming the stock price prediction materials
using the scale below.

1. I felt mentally tired and worn out after Ieaming the stock price prediction materials.

1 2 3 4 5

Strongly Agree Strongly disagree

 

2. Learning the stock prediction materials was a difficult and complex task.

1 2 3 4 5
Strongly Agree Strongly disagree

 

3. The overall mental workload I felt while Ieaming the stock prediction materials was
low.

 

 

1 2 3 4 5
Strongly Agree Strongly disagree
4. Learning the stock price prediction materials was easy.

1 2 3 4 5
Strongly Agree Strongly disagree

5. Learning the stock price prediction materials required a lot of mental activity.

 

1 2 3 4 5
Strongly Agree Strongly disagree

6. I had to work very hard to Ieam the stock price prediction materials.

1 2 3 4 5
Strongly Agree Strongly disagree

 

108

APPENDIX F

Stock Price Prediction Learning Task

Instructions (Application condition)

Take as much time as you need to Ieam the material on the following pages. You may
use any method you choose to Ieam the material. After you have Ieamed the material,
you will be asked to predict stock prices for ten companies. Later we will score your test
so you know how well you have done on the test.

Instructions (Knowledge condition)

Take as much time as you need to Ieam the material on the following pages. You may
use any method you choose to learn the material. After you have Ieamed the material,
you will be asked to take an 18 item multiple choice test on the material. Later we will
score your test so you know how well you have done on the test.

Instructions (Ambiguous condition)

Take as much time as you need to Ieam the material on the following pages. You may
use any method you choose to Ieam the material. After you have Ieamed the material,
you will be asked to demonstrate your Ieaming on a short test. Later we will score your
test so you know how well you have done on the test.

109

The following task focuses on how an investment counselor makes stock
recommendations to a client. During the course of this session, you will be
asked to learn how an investment counselor might make decisions about stock
values.

More specifically, you will learn about making estimates concerning the behavior
of the stock for a variety of similar large multi-national corporations based in the
United States. The current market value of each company’s stock as listed on
AMEX is $80 per share. The stock price for a given company will rarely fall -"
below $10 per share, or rise above $150 per share. In addition, the performance
of each company is independent. That is, the performance of company #10 does
not influence the performance of company #11.

Stock prices can be predicted from performance data about each company.
Three divisions of each company are considered in the prediction of stock I
prices; 1) marketing, 2) research and development, and 3) production. Each

division reports quarterly performance levels, measured in millions of dollars

gained or lost. A positive value reflects a profit, while a negative value reflects a

loss. Information concerning quarterly performance can be found in each

company's shareholder disclosure reports.

 

The shareholder disclosure reports also contain information such as the
previous annual dividend and the change in company profits. The previous
annual dividend gives an indication of how much money was paid to investors
during the previous year. The change in company profits indicates how much
money the company as a whole made during the previous year. The relationship
between the previous annual dividend is positive but low, usually a correlation of
.15.

The prediction of stock prices is somewhat uncertain. In a given year, the
average fluctuation of stock prices is $25. Investment counselors cannot always
perfectly predict the prices of stocks. In fact, they usually succeed only 65% of
the time. The stock market is affected by many factors outside of organizational
performance, such as interest rates, political events, and economic cycles.

Investment counselors also track the trends of stock prices. Short term stock
trends refer to the movement of the stock price over the past three months.
Long term stock trends refer to the general movement of stock prices over a
longer period of time, usually a year or more. These trends allow counselors to
give stock ratings. An A+ is the top stock rating, while a C- is the lowest
Possible rating.

110

Multiple regression is a statistical technique which can be used to predict one
number from a weighted combination of other numbers. The goal in multiple
regression is to reduce, or minimize, the errors one makes in prediction.

Multiple regression is based on the general linear model. Thus, the basic
equation for multiple regression is similar to the equation for a line:

y=a+b1x1+b2x2

y is the number you want to predict

a is the intercept term, or the place where the line crosses the y axis
b1 is the weight given to the first number that you know

x1 is the first number that you know

b: is the weight given to the second number that you know

x2 is the second number that you know

For example, you could predict the weight of children (y) from the number of
hours of television watched (x1) and their age (x2).

Assume that bl = 2, and b2 = 7. Assume also that a = 8, because that is the
average weight of a newborn. If a child spends 7 hours a week watching TV,
and is 10 years old, then the equation will look like this:

y=8+(2X7)+(7X10)
y = 92
Thus, one would predict that this child weighs 92 pounds.

Multiple regression can be used to predict stock prices from performance data,
and can be used with any number of x terms.

:> The b value for each term can change depending on the performance level of
each division. If the quarterly performance for one division is between 0 and
50, b=.5. If performance is between 51 and 100, b = .2. If performance is
between 101 and 150, b = .1. If the performance value is negative, use the
absolute value (remove the negative sign) to determine the b value.

:> The a value will always be the current price of the stock.

Two examples of the use of regression in predicting stock prices are on the next
page.

 

111

Example 1

At Bob’s Kreme Filling, Inc., the following quarterly performance data were
reported:

0 Marketing 20

a Research and
Development -60

0 Production 100

The regression equation for predicting the stock price of Bob’s Kreme Filling,
Inc. is:

y=80+(.5)20+(.2)-60+(_2) 100
y=80+10-12+20

y=98

Example 2:

At Mike’s International BrewPub, the following performance data were reported:

0 Marketing 70

0 Research and
Development 10

0 Production -20

Feel free to practice predicting the stock price for Mike’s lntemational BrewPub
in the space below. The completed regression equation is displayed on the next

page.

Feel free to look over any portion of the Stock Prediction materials until
you feel you have Ieamed the materials. When you are finished Ieaming
the material, please raise your hand and the experimenter will collect your
materials.

 

112

Answer to Example 2:

Marketing 70

Research and
Development 10

Production --20

y = 80 + (.2) 70 + (.5) 10 + (.5) -20

y=80+14+5-10

Y

89

 

113

APPENDIX G

Knowledge Learning Outcome

In this part of the session you will be asked to choose the correct answer to the
18 multiple choice items below. Please circle the correct answer.

1. The divisions of a company an investment counselor uses to predict stock
prices are:

a) marketing, sales, and human resources

b) marketing, research and development, sales, and production

c) finance, customer service, and research and development

d) marketing, research and development, and production

e) production, sales, and customer service

2. When evaluating how well a division has done, the investment counselor
looks at:

a) percentage of goal met

b) profit/loss

c) receivables

d) market share

e) stock ratings

3. Predicted stock prices generally range from:
a) $20 - $200

b) $10 - $150

c) $80 - $150

d) $10 - $80

e) $25 - $150

4. The investment counselors make predictions about:
a) chemical companies

b) large auto supply companies

c) a wide range of companies

d) large multi-national companies

e) large financial companies

Continue on to the next page

 

114

5. Performance is measured for each division in the companies:

a) yearly
b) monthly

c) weekly
d) quarterly
e) bimonthly

6. External influences on the stock market include:

3) seasons, weather, and mergers

b) interest rates, political events, and economic cycles.
c) Supreme Court decisions, acquisitions, and the Fed
d) major holidays, product cycles, and inflation

e) the global economy, elections, and downsizing

7. Multiple regression is based on:

a) the weighted geometric model
b) the general linear model

c) Euclidean geometry

d) vectors and angles

e) the factor analytic model

8. Investment counselors are correct about stock price predictions:

a) half the time

b) very rarely

c) two-thirds of the time
d) always

e) one-fourth of the time

9. The primary goal in multiple regression is to:

a) choose between several options

b) maximize the value of the stock price

c) reduce the amount of information needed
d) minimize errors in prediction

e) find the absolute value of performance

Continue on to the next page

115

10. The b value in a regression equation is:

a) the number you are trying to predict

b) the place where the line crosses the y axis

c) one of the numbers that you know

d) the weight given to one of the numbers that you know
e) the current value of the stock

11. In the first multiple regression example, we were trying to predict:

a) the weight of an average newborn

b) a child’s age

c) a child’s weight

d) the amount of time a child watched TV

e) the number of Tvvinkies eaten by a child each week

12. The companies list their stocks on the following stock exchange:

a) NYSE

b) OTC

c) NASDAQ
d) CBOT

e) AMEX

13. The current value of each stock is:

a)$50
b)$80
c)$10
d)$25
e)$100

14. Performance information for each firm can be found in:

a) the annual report

b) the Wall Street Journal

c) shareholder disclosure reports

d) Business Week

e) the business section of any newspaper

Continue on to the next page

116

15. The average annual fluctuation of stock prices is:

a)$10
b)$80
c)$25
d)$50
e)$15

16. The lowest possible stock rating is:

a) C
b) E
c) C-
d) F
e) D-

17. If the performance value for the research and development division of a
company is 50, the b value for that term is:

18. The relationship between change in company profits and previous annual
dividend is:

a) high and positive

b) moderate and negative
c) low and positive

d) low and negative

e) zero

Please raise your hand to let the experimenter know you are finished.

 

117

APPENDIX H

Application Learning Outcome

In this part of the session, you will play the role of an investment counselor.
After examining the performance of several divisions for each of the ten companies
listed below, you should estimate the price of the stock for that company.

Write your prediction in the box labeled “Predicted Stock Price.” You should try to
predict as close as possible to the actual stock price each time.

You will be provided with a calculator to assist in the math.
If you have any questions at this time, please feel free to ask the experimenter.

Please begin the test now.

 

 

Company 1:

Marketing 20
Research and Development 10
Production 40

 

 

 

Predicted Price

 

118

Company 2:

Marketing 110

Research and Development 120
Sales 70

Production 130

Predicted Price

 

Company 3:
Marketing 70

Research and Development 50
Production -30

Predicted Price

 

Company 4:

Marketing 140

Research and Development 110
Sales 50

Production 40

Predicted Price

 

 

 

 

 

 

 

 

 

 

 

 

 

119

Company 5:
Marketing 80

Research and Development 60
Production 90

Predicted Price

 

Company 6:

Marketing 140

Research and Development 80
Sales -20

Production -60

Predicted Price

 

Company 7:
Marketing 100

Research and Development 120
Production 120

Predicted Price

 

 

 

 

 

 

 

 

 

120

Company 8:
Marketing —1 10

Research and Development ~70
Production 40

Predicted Price

 

Company 9:

Marketing 20

Research and Development -70
Sales 40

Production 30

Predicted Price

 

Company 10:
Marketing -70

Research and Development 90
Production -80

Predicted Price

 

 

 

 

 

 

 

 

 

 

 

 

121

APPENDIX I

Pilot Study Results

The following are the item means and corrected item-scale correlations for the items on
the knowledge outcome measure used in the pilot. The coefficient alpha reliability for
the scale was .38. In an effort to correct the high means and low item-total correlations,
an additional distracter was added for each item. Three items were also added.

Ite_m t _rr._.
1. ‘L00 -—
2. 71 .06
3. .92 a22
4. .58 12
5. .95 11
6. .95 11
7. .92 11
8. .89 .12
9. .89 .02
10. .82 .08
11. .82 .40
12. .97 n25
13. .66 .32
14. .47 .29
15. .68 .37

The following are the item means and corrected item-scale correlations for the items on
the Ieaming strategy measures used in the pilot. As a result of these data, several
items on the rehearsal and organizing scales were rewritten.

Rehearsauot= .33) Organizing a =.60) Personalizinggot= .70)
Item x rt. Item x r;.. Item x r,..

 

3.97 .14 7 1.52 .71 13 3.29 .27
2.60 .13 8 4.31 -.01 14 2.52 .38
1.89 .04 9 1.81 .32 15 1.76 .63
3.92 .43 10 2.81 .19 16 1.55 .62
4.44 .00 11 2.58 .52 17 3.28 .37
2.81 .21 12 2.71 .45 18 3.86 .32

 

 

GUI-hOON-h

Note: The modified versions of all scales are presented in the Appendices.

LIST OF REFERENCES

LIST OF REFERENCES

Anderson, JR. (1982). Acquisition of cognitive skill. Psycholpgical Review, 89(4), 369-
406.

Anderson, J.R. (1983). aniiive Psychology and its Implications. New York: W.H.
Freeman and Company.

Anderson, JR. 8 Fincham, J.M. (1994). Acquisition of procedural skills from examples.
Journal of gperimentgl Psychology: Learning. Memorv a_nd Cogpition. @(6),
1 322-1 340.

Ashcraft, M.H. (1989). Human Memory and @gnition. HarperCollins Publishers.

Ausubel, DP. (1968). Educational psychology: A cognitive view. New York: Holt,
Rinehart, & Winston.

Baldwin, T.T., & Ford, J.K. (1988). Transfer of training: A review and directions for
future research. Personnel Psychology, fl, 63-105.

Barnett, J.E., DiVesta, F .J. and Rogozinski, J.T. (1981). What is Ieamed in note-taking?
Journal of Educational Psychology, _7_3_, 181 -1 92.

Biedermann, l. (1990). Higher level vision. In D.N. Osherson, S.M. Kosslyn, 8r J.M.
Hollerbach (Eds.), An invitation to anitive science. Volume 2: Visual anition
and action (pp. 41-72). Cambridge, MA: The MIT Press

Biggs, J. (1993). What do inventories of students’ Ieaming processes really measure?
A theoretical review and clarification. British Journal of Educational Psycholpgy,
Qt 3'19.

Bloom, BS. (1956). Taxonomy of Educational Obiectives: The Classification of
Educational Goals. Longmans, Green and Co: London.

 

Boyle, K.A. 8r Klimoski, R.J. (1995). The Role of Goal Orientation in a Training Context.
Paper presented at the Tenth Annual Meeting of the Society for Industrial and
Organizational Psychology. Orlando.

122

123

Button, S.B., Mathieu, J.E. 8r Zajac, BM. (1994). The Development and psychometric
evaluation of measures of Ieaming goal and performance goal orientation.
Manuscript under review, The Pennsylvania State University.

Cohen, J. (1992). A power primer. Psycholpgical Bulletin, 112(1), 155-159.

Craik, F.I.M. & Lockhart, RS. (1972). Levels of processing: A framework for memory
research. Journal of Verbal Learning and Verbal Behavior, 1_1, 671-684.

Das, JP. (1988). Simultaneous-successive processing and planning: Implications for
school Ieaming. In Schmeck, R.R. (Ed.), Learning Strategies and Learning
SMes. pp. 53-82. New York: Plenum Press.

Dweck, C. S. (1 986). Motivational processes affecting Ieaming. Americgn Psychologist,
41 (1 0), 1040-1048.

Dweck, CS, 8. Leggett, EL. (1988). A Social-cognitive approach to motivation and "
personality. Psychological Review, 95(2), 256-273.

 

D'Ydewalle, G., 8 Swerts, A. (1980). Motivational variables in knowledge acquisition:
Test expectancies. In R. Glaser and J. Lompscher (Eds.) Cpgnitive and
Motivational Aspects of Instruction. North-Holland: New York.

Earhard, M. (1969). Storage and retrieval of words encoded in memory. Journal of
Experimental Psychology. fl. 412-418.

Earley, P.C., Connoly, T., 8 Ekegren, G. (1989). Goals, strategy development, and task
performance: Some limits on the efficacy of goal setting. Journal of Applied

Psycholggx. 3(1), 24-33.

Farr, J.L., Hofman, DA, 8. Ringenbach, KL. (1993). Goal orientation and action control
theory: Implications for Industrial and Organizational Psychology. lntemational
Review of Industrial and Organizational Psychology, Vol. 8, 193- 232.

Fischer, HE. (1993). Framework for conducting empirical observations of learning
processes. Science Education, 21(2), 131-151.

Gagné, RM. (1984 April). Learning outcomes and their effects. American
Psycthgist, g (4), 377-385.

Gagné, R.M., Briggs, L.J., & Wager, W.W. (1992). Principles of Instructional Design.
Philadelphia: Harcourt Brace Jovanovich.

Hancock, P.A. 8r Caird, J.K. (1993). Experimental evaluation of a model of mental
workload. Human Factors 3_5_(3), 413-429.

 

Hart, S.G. & Staveland, LE. (1988). Development of NASA-TLX (Task Load Index):
Results of Empirical and Theoretical Research. In PA. Hancock and N.

124

Meshkati (Eds.), Human Menta_l Workload. Elsevier Science Publishers B.V.
(North- Holland): New York.

Henderson, J.M. (1992). Object identification in context: The visual processing of
natural scenes. Canadian Jouma_l of Psychtlogy, 49, 319-341.

Howard-Rose, D., & Winne, RH. (1994). Measuring component and sets of cognitive
processes in self-regulated Ieaming. Journal of Educational Psycholggy. §§(4),
591-604.

Hubbard, D.L. (1994). Can higher education Ieam from factories? Query Pmress,
fl(5). 93-97.

James, L.R. 8: Brett, J.M. (1984). Mediators, moderators, and tests for mediation.
Journal of Applied Psychology. §§(2), 307-321.

Kanfer, R. (1990) Motivation theory and Industrial and Organizational Psychology. In
M.D. Dunnette and L.M. Hough (Eds.), Handbook of Industrial and

Organizational Psycholggy, 2nd ed., (Vol. 1, pp. 75-170). Palo Alto, CA:
Consulting Psychologists Press.

Kanfer, R. & Ackerrnan, P.L. (1989). Motivation and cognitive abilities: An
integrative/aptitude-treatment interaction approach to skill acquisition. Journal
gt Applied Psychology, 74(4), 657-690.

Kanfer, R., Ackennan, P.L., Murtha, T.C., Dugdale, B., 8 Nelson, L. (1994). Goal
setting, conditions of practice, and task performance: A resource allocation
perspective. Journal of Applied Psychology, 79(6), 826-835.

Kirby, JR. (1988). Style, strategy, and skill in reading. In Schmeck, R.R. (Ed.),

Learning Strategies and Learning SMes. pp. 229-274. New York: Plenum
Press.

Kraiger, K. Ford, J.K., & Salas, E. (1993). Application of cognitive, skill-based, and
affective theories of Ieaming outcomes to new methods of training evaluation.

Journal of Applied Psychology, Z_8_, 311-328.

Levin, JR (1986). Four cognitive principles of Ieaming-strategy instruction.
Educational Psycholggist, _?._1(1&2), 3-17.

Lord, R.G., & Maher, K.J. (1991). Cognitive theory in Industrial and Organizational
Psychology. In M.D. Dunnette and L.M. Hough (Eds.), Handbook of Industrial
and Organizational Psychology. Vol. 2 (pp. 1-62), Palo Alto, CA: Consulting
Psychologists Press.

Marion, F. (1988). Describing and improving Ieaming. In Schmeck, R.R. (Ed.),
Learning Strategies and Learning Styles. pp. 53-82. New York: Plenum Press.

125

Marion, F., & Saljo, R. (1976). On qualitative differences in Ieaming: I - Outcome and
Process. British Joumal of Educationa_l Psychology, 51g, 4-11.

Mayer, RE. (1979). Twenty years of research on advance organizers: Assimilation
theory is still the best predictor of results. Instructiongl Science, 8, 133-167.

McKelvey, J.D., & Lord, RS. (1986, April). The effects of automatic and controlled
processing on rating accuracy. Paper presented at the annual meeting of the
Society for Industrial and Organizational Psychologit. Chicago.

Meece, J.L., Blumenfeld, PC, and Hoyle, RH. (1988). Students’ goal orientations and
cognitive engagement in classroom activities. Joumgl of Educational

PsEhOMy, 80 (4), 514-523.

Mitchell, T., Hopper, H., Daniels, 0., Falvy, J. 8. James, L. (1994). Predicting self-
efficacy and performance during skill acquisition. Journal of Applied
Psycholggy, Q, 506-517.

Morris, C.D., Bransford, J.D., & Franks, J.J. (1978). Level of processing versus

transfer appropriate processing. Journal of Verbal Learning and Verbal
Behavior, fl, 519-533.

Paas, F.G.W.C. (1992). Training strategies for attaining transfer of problem-solving skill
in statistics: A cognitive-load approach. Joumflf Educational Psycholpgy,
§4(4), 429-434.

Paas, F.G.W.C., & Van Menienboer, J.J.G. (1994). Variability of worked examples
and transfer of geometrical problem-solving skills: A cognitive load approach.
Journal of Eggcatiorfl Psychology, _8_§(1), 122-133.

Rigney, J.W. (1978). Learning strategies: A theoretical perspective. In H.F. O’Neil, Jr.
(Ed.) Learning Strategies. New York: Academic Press.

Roediger, H.L., Weldon, M.S., & Challis, B.H. (1989). Explaining dissociations
between implicit and explicit measures of retention: A processing account. In
H.L. Roediger and F .l.M. Craik (Eds.), ygrieties of Memory and Consciousness:
Essays in HOI‘lOl_I_I' of Endel Tulvflg. (pp. 3-41). Hillsdale, NJ: Lawrence
Earlbaum Associates.

Schmeck, RR. (1983). Learning styles of college students. In R.F. Dillon and RR.
Schmeck (Eds.), Individual Differences i_n Cogflon (Vol. 1). New York:
Academic Press.

Schmeck, RR. (1988). Individual differences and Ieaming strategies. In C.E.
Weinstein, E.T. Goetz, 8: PA. Alexander (Eds.) Learning and Study
Strategies: Issues in Assessment, Instrt_tction ang Evaflation (pp. 171-191).
San Diego: Academic Press.

126

Schmidt, RA. & Bjork, RA. (1992). New conceptualizations of practice: Common
principles in three paradigms suggest new concepts for training. Psycholggical
Science, 3(4), 207-217.

Smith, E. M., Ford, J.K. Weissbein, D. & Gully, SM. (1995). The influence of goal
orientation and metacognition on the Ieaming and transfer of complex skills.
Paper presented at the Tenth Annual Meeting of the Society for Industrial and
Organizational Psychology, Orlando.

Terborg, JR. (1977). Validation and extension of an individual differences model of
work performance. Organizational Behavior and Huma_n Performance, 18, 188-
216.

Thomson, D.M., 8r Tulving, E. (1970). Associative encoding and retrieval: Weak and
strong cues. Journal of E_xperimenta_l Psychology, Q, 255-262.

Triesman, A. (1986). Features and objects in visual processing. Scientific American,
2_55. 1148-125.

Tulving, E. (1983). Elements of Episodic Memogy. New York: Oxford University Press.

Tulving, E., 8. Thomson, 0M. (1973). Encoding specificity and retrieval processes in
episodic memory. Psycholggical Review, pp, 352-373.

User’s Manual for the Wonderlic Personnel Test and the Scholastic Level Exam.
(1992). Wonderlic Personnel Test, Inc. Libertyville, IL.

Watkins, M.J., Peynircoiglu, Z.F., & Brehms, DJ. (1984). Pictoral rehearsal. Memory
and Cognition, g, 553-557.

Weinstein, C. E. (1986). Why does a school psychologist need to know about Ieaming
strategies? Journal of School Psychology, 2_4(3), 257-265.

Weiss, HM. (1990). Learning theory and Industrial and Organizational Psychology. In
M.D. Dunnette and L.M. Hough (Eds.), Handbook of Industrial and
Organizational PsychologLVol. 1 (pp. 171-221), Palo Alto, CA: Consulting
Psychologists Press.

Weldon, M.S., Roediger, H.L. & Challis, B.H. (1989). The properties of retrieval cues
constrain the picture superiority effect. Memory a_nd Cognition. fl, 95-105.

Wofford, J.C. (1990). Experimental Analysis of a cognitive model of motivation. m
Journal of PsychoLOQx. MU). 87-101.

MICHIGAN STATE UNIV. LIBRARIES
IlllllllllllllllllllIll“Ill"lllllllllllllll
31293014173086