 

r,1q.

.1 ... 5.,

3' 1‘
“in.”

.~ 11.
H... b? ‘2
k 3. § .
.2.

.11». a v‘
. 11 § :
ital: .. . V .
.lIf.......l.,¥a..il... 315%..
I, . v- _ (.2 .
ck ..

n

'3

' f
If“;

, .l
a
\ its... zip-‘45
Nu. 14 {.9 no 4..
. .a $1.? 1. 2.
(Jazz .5... .
I :2.
.111“
. is...)
.ﬂqﬁihnalhu .1...
.‘nmuwuawwhrrmﬁvvkh
:\ .I .. 5 if. .
_..w:8..hu 323.2
3... . . M...”
AS . t
. V 7‘1!
. z

m. ,,

“I. It,
Jun! .35., .Hvﬂuah.rfr .

 

b . £559...- 9
w. Isvl
.3...)
..
.5

1.511 .4. 1.2.1.‘ -
xvi-2:35.11 .
II‘)‘ I“ ll

. . 3:! ...l
{.1351 . .ﬁwwowa.

I 5.41... _.

MES Eu“ . a

:2

“

 

2/57?

This is to certify that the
dissertation entitled

THE DIFFERENTIAL EFFECTS OF POSITION, VELOCITY, AND
ACCELERATION FEEDBACK ON MOTIVATION OVER TIME

presented by

Daniel Jacob Watola

has been accepted towards fulﬁllment
of the requirements for the

Doctoral degree in Psychology

 

 

W/WL

Major Pr essor’ Ignature
OZ; Mﬂ6% 50725”,

Date

 

MSU is an Afﬁnnative Action/Equal Opportunity Institution

 

LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 p:/ClRC/DateDue.indd-p.1

THE DIFFERENTIAL EFFECTS OF POSITION, VELOCITY, AND
ACCELERATION FEEDBACK ON MOTIVATION OVER TIME

By

Daniel Jacob Watola

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Psychology

2005

ABSTRACT

THE DIFFERENTIAL EFFECTS OF POSITION, VELOCITY, AND
ACCELERATION FEEDBACK ON MOTIVATION OVER TIME

By

Daniel Jacob Watola
This paper is concerned with the deﬁnition and application of position, velocity, and
acceleration performance information as feedback. Speciﬁcally, it examines individuals’
affective, behavioral, and cognitive reactions to feedback frames overtime and across
two contrasting performance proﬁles. Repeated measures MANCOVA supported a
performance proﬁle it time interaction for state positive affect, task self-efﬁcacy,
satisfaction with performance, and goal commitment. Simple effects analyses indicated
that participants’ indicators of task motivation increased over time in the accelerating
performance proﬁle, but decreased over time in the decelerating performance proﬁle. A

three-way interaction between feedback frame, performance proﬁle, and time was not

supported as hypothesized.

ACKNOWLEDGMENTS

This dissertation would not have been possible without the guidance of my advisor, Steve
Kozlowski, the wisdom of the I/O Psychology faculty at Michigan State University, the
support of my friends and family, and the generous funding and sponsorship of the Air
Force Institute of Technology and United States Air Force Academy. The views
expressed in this article are those of the author and do no reﬂect the ofﬁcial policy or

position of the United States Air Force, Department of Defense, or U. S. Government.

iii

TABLE OF CONTENTS

LIST OF TABLES .............................................................................................................. vi
LIST OF FIGURES ........................................................................................................... vii
INTRODUCTION ............................................................................................................... 1
LITERATURE REVIEW
The Role of Feedback in Theories of Self-Regulation ............................................ 4
Feedback Frames: Position, Velocity & Acceleration Information ....................... 11
Learning or Skill Acquisition Performance Proﬁles Over Time ........................... 22
Motivational Indicators .......................................................................................... 25
Research Hypotheses ............................................................................................. 28
METHOD
Design .................................................................................................................... 35
Sample .................................................................................................................... 35
Procedure ............................................................................................................... 36
Measures ................................................................................................................ 40
RESULTS
Manipulation Check ............................................................................................... 44
Interaction Hypotheses ........................................................................................... 45
Additional Analyses ............................................................................................... 49
DISCUSSION
Implications ............................................................................................................ 56
Limitations & Future Research .............................................................................. 57
Conclusions ............................................................................................................ 6O
APPENDICES
Tables ..................................................................................................................... 63
Figures .................................................................................................................... 76
Informed Consent Form ......................................................................................... 95
Authorization to Release Test Scores Form ........................................................... 97
Social Intuition Task Trials 1-12 ........................................................................... 99
Demographic and Individual Difference Measures ............................................. 112
F amiliarization Trial ............................................................................................ 1 14
Feedback Manipulation ........................................................................................ 116
Debrieﬁng Form ................................................................................................... 120
Motivational Measures ......................................................................................... 122

iv

REFERENCES ................................................................................................................ 1 25

LIST OF TABLES

Table 1. Means, standard deviations, zero-order correlations, and reliabilities of
the individual difference and reaction measures ................................................................ 64

Table 2a. Descriptive statistics for reaction measures by time and performance

proﬁle for participants receiving position feedback .......................................................... 67
Table 2b. Descriptive statistics for reaction measures by time and performance

proﬁle for participants receiving velocity feedback .......................................................... 68
Table 2c. Descriptive statistics for reaction measures by time and performance

proﬁle for participants receiving acceleration feedback .................................................... 69
Table 3. Results of the repeated measures multivariate analysis of covariance ............... 70
Table 4. Results of the repeated measures univariate analyses of covariance .................. 71

Table 5. Results of the simple effects univariate analyses for the performance
proﬁle x time interaction .................................................................................................... 73

Table 6. Results of the feedback frame x performance proﬁle x time interaction’s
simple effects univariate analyses for the task self-efﬁcacy measure ............................... 74

Table 7. Feedback ﬂames that produced signiﬁcantly higher motivational scores
by performance proﬁle ....................................................................................................... 75

LIST OF FIGURES

Figure 1. A nominal meta loop with an embedded action loop ........................................ 77
Figure 2. Lawrence et al.’s (2002) velocity feedback conditions ..................................... 78
Figure 3. An example of position-salient feedback .......................................................... 79
Figure 4. An example of velocity-salient feedback .......................................................... 80
Figure 5. An example of acceleration-salient feedback ................................................... 81

Figure 6. A decelerating learning curve characterized by rapid early performance
gains that slowly diminish over time ................................................................................. 82

Figure 7. An accelerating learning curve characterized by slow early performance

gains that rapidly increase over time ................................................................................. 83
Figure 8. Hypothesized performance proﬁle x time interaction ....................................... 84
Figure 9a. Hypothesized feedback ﬂame x performance proﬁle x time interaction

for position feedback .......................................................................................................... 85
Figure 9b. Hypothesized feedback ﬂame x performance proﬁle x time interaction

for velocity feedback .......................................................................................................... 86
Figure 9c. Hypothesized feedback ﬂame x performance proﬁle x time interaction

for acceleration feedback ................................................................................................... 87
Figure 10a. Performance proﬁle x time interaction for state positive affect .................... 88
Figure 10b. Performance proﬁle x time interaction for task self efﬁcacy ........................ 89
Figure 10c. Performance proﬁle x time interaction for self-satisfaction with

performance ....................................................................................................................... 90
Figure 10d. Performance proﬁle x time interaction for goal commitment ....................... 91

Figure 11 a. Feedback ﬂame x performance proﬁle x time interaction for task-self
efﬁcacy with position feedback ......................................................................................... 92

Figure I I b. Feedback ﬂame x performance proﬁle x time interaction for task-self
efﬁcacy with velocity feedback ......................................................................................... 93

vii

Figure 1 1c. Feedback ﬂame x performance proﬁle x time interaction for task-self
efﬁcacy with acceleration feedback ................................................................................... 94

viii

INTRODUCTION

In their review of the literature concerning the efﬁcacy of feedback interventions,
Kluger and DeNisi (1996) noted that organizational researchers have historically ignored
the variable relationship between feedback and performance which, in turn, has
contributed to the widely held assumption that providing feedback improves
performance. This assumption prompted Pritchard, Jones, Roth, Stuebing, and Eckberg
(1988) to proclaim, “The positive effect of feedback on performance has become one of
the most accepted principles in psychology” (p. 338).

While other researchers have been more cautious in their statements regarding the
efﬁcacy of feedback (e.g., Balcazar, Hopkins & Suarez, 1985; Ilgen, Taylor & Fisher,
1979; Latham & Locke, 1991), Kluger and DeNisi (1996) identiﬁed Ammons’ (1956)
review as an often-cited source for the feedback assumption. After reviewing the
research on feedback, Ammons had concluded that knowledge of performance positively
affects both learning and motivation. However, Kluger and DeNisi claim that Ammons’
conclusions are inappropriate considering their own review of his referenced studies
revealed contradictory evidence, inaccurate operationalizations, and substandard
methodologies. Indeed, their meta-analysis of feedback interventions found a sample-
size weighted mean effect size of 0.41, which suggests a moderately positive effect of
feedback on performance; however, they also noted that some 38% of the observed
effects were negative. Ultimately, they concluded that much more needs to be learned

about feedback, with a special emphasis on identifying the feedback property, task

characteristic, contextual, and individual difference moderators of the feedback-
performance relationship.

It is ironic that industrial and organizational psychologists do not know more
about the effects of providing performance feedback given the prominent role it plays in
theories and processes of performance appraisal, performance development, self-
regulated learning, and skill-acquisition. One result of this situation is that, even though
feedback can take many forms and be used for many purposes, there is no commonly-
accepted typology of performance feedback. Some researchers have proposed
differentiating among types of feedback according to its intended purpose or information
content. For example, Kluger and DeNisi (1996) distinguish between providing
knowledge of results (i.e., outcome or “what” information) and knowledge of
performance (i.e., process or “how” information), Kozlowski et a1. (2001) propose a
feedback typology that distinguishes between the informational (i.e., descriptive) and
interpretational (e.g., evaluative or attributional) properties of feedback, and Carver and
Scheier’s (1990; 1998) interpretation of Control Theory distinguishes between the use of
position feedback (i.e., discrepancy with respect to goal) and velocity feedback (i.e., rate
of change in discrepancy reduction with respect to goal). Without an organizing
ﬂamework for the study of feedback, it is no wonder that some types of feedback are
relatively unknown. For example, despite the introduction of velocity feedback by
Carver and Scheier in 1990, only a handful of studies have been published that explore
this concept and its related forms (i.e., position and acceleration feedback).

This paper is concerned with the deﬁnition and application of position, velocity,

and acceleration performance information as feedback. Speciﬁcally, this study will

examine individuals’ affective, behavioral, and cognitive reactions to performance
feedback presented in position-, velocity—, and acceleration-salient frames. By exploring
the effects of these feedback frames on different performance proﬁles over time, I expect
to ﬁnd that some ﬁames produce greater motivation gains than others in speciﬁc proﬁles
and at speciﬁc time periods during skill acquisition. Before introducing my research
hypotheses, I will ﬁrst review the role of feedback in self-regulation theory; conceptually
and operationally deﬁne position, velocity, and acceleration feedback ﬁames; present two
contrasting performance proﬁles (i.e., learning curves) that depict the process of self-
regulated learning or skill acquisition over time, and describe several outcome variables

that can be used to assess the effects of feedback in motivating performance.

LITERATURE REVIEW

The Role of Feedback in Theories of Self-Regulation

Feedback is a core component of many theories of self-regulation including Goal
Setting Theory (Locke & Latham, 1990; Latham & Locke, 1991), Social Cognitive
Theory (Bandura, 1986; 1991; Wood & Bandura, 1989), and Control Theory (Campion
& Lord, 1982; Carver & Scheier, 1982; 1990; 1998; Lord & Hanges, 1987). In the
following section, I will introduce each of these theories and discuss the role that
feedback plays in their explanation of self-regulatory behavior.

Goal Setting Theory. With well over one-thousand goal-related studies in
publication, Locke and Latham’s (1990) Theory of Goal Setting has emerged as “the
single most dominant theory in the ﬁeld [of organizational behavior]” (Mitchell &
Daniels, 2003, p. 231). At its core, the theory asserts that challenging, speciﬁc goals are
a proximal regulator of human behavior because people have “things” that they want to
achieve. Put simply, goal setting facilitates self-regulation because a goal speciﬁes a
standard, or deﬁnes an acceptable level of performance, for use in evaluating purposeful
activities. In order to evaluate a performance relative to a standard, performance
feedback must be available. When efforts produce outcomes and corresponding feedback
that indicate a failure to meet the standard, then these efforts are evaluated negatively and
prompt actions to improve subsequent performance. However, when efforts result in
outcomes and associated feedback that reveals that the standard has been achieved or
surpassed, then they are evaluated positively and can lead to a reduction in effort or the

adoption of a new, higher standard.

The beneﬁts of difﬁcult and speciﬁc goals have been demonstrated across
numerous studies employing different tasks, settings, subjects, countries, criteria, and
time spans (see Latham and Locke, 1990, for a review), and is reﬂected in the results of
several meta-analyses of the goal-performance relationship (e. g., Mento, Steel & Karren,
1987; Tubbs, 1986; Wright, 1990). However, this observed relationship is moderated by
several individual and situational variables such as ability (Locke, 1982), task complexity
(Wood, Mento & Locke, 1987), and the presence of feedback. Speciﬁcally, individuals
must have adequate knowledge of results (i.e., feedback) if they are to evaluate their own
progress toward their goals (Erez, 1977; Locke, Shaw, Saari & Latham, 1981). Bandura
and Cervone (1983) found that feedback was an essential component of goal systems
after observing signiﬁcant increases in participants’ performances on a strenuous physical
task, but only when both goals and performance feedback were provided. Thus, feedback
is not only necessary for self-regulation in Goal Setting Theory, it is also an important
means for leveraging performance.

Social Cognitive Theory. Social Cognitive Theory provides a framework for
understanding the interactions among human behavior, cognition, affect, and the
environment in self-regulation. Like Goal Setting Theory, Social Cognitive Theory is
capable of explaining the self-regulation of goal-directed behavior; however, it places
greater emphasis on a person’s cognitive and affective reactions to their performance and
their environment (Kanfer, 1993).

According to Bandura (1986; 1991), three psychological “subfunctions” are
employed for self-regulation: self-monitoring, self-evaluation, and self-reaction. Self-

monitoring involves the observation of those behaviors, enviromnental conditions, and

outcomes that are instrumental for goal attainment. These observations provide the
information required to set achievable goals and evaluate goal progress. Self-evaluation
entails making a judgment about goal progress. More precisely, it requires making a
comparison between a current state (i.e., performance) and a desired state (i.e., goal).
Finally, self-reaction refers to internal, affective responses to self-evaluation (i.e.,
satisfaction or dissatisfaction). When performance is substandard, self-reactions generate
feelings of dissatisfaction. However, when performance meets or surpasses the standard,
they generate feelings of satisfaction. In either case, the intensity of the self-reaction is
commensurate with the magnitude of the discrepancy between the performance and the
standard. Collectively, the self-monitoring, self-evaluation, and self-reaction
subfunctions provide the mechanism for self-regulation. Wood and Bandura (1989)
summarize this process best:
People seek self-satisfactions ﬁ'om fulﬁlling valued goals, and they are motivated
by discontent with substandard performances. Thus, discrepancies between
behavior and personal standards generate self-reactive inﬂuences, which serve as
motivators and guides for action designed to achieve desired results. (p. 366)
Thus, Social Cognitive Theory relies on feedback obtained during self-monitoring
to evaluate performance during self-evaluation. Speciﬁcally, people self-monitor their
performance and attend to environmental indicators of their performance during self-
monitoring in order to generate performance feedback. This feedback is then used during
self-evaluation to determine if performance is substandard, adequate, or more than

adequate with respect to the goal. Therefore, like Goal Setting Theory, Social Cognitive

Theory requires feedback to evaluate goal progress and motivate subsequent goal-
directed behaviors.

Control Theory. As with the preceding theories, Control Theory includes
feedback in its explanation of self-regulation. However, Carver and Scheier’s (1990;
1998) revised treatments of Control Theory distinguish among different types of
feedback. Speciﬁcally, they discuss the role of position feedback, introduce the concept
of velocity feedback, and allude to the possibility of acceleration feedback. As their
theory is foundational to this study’s discussion of feedback frames, I will review it in
greater detail.

Many of the concepts found in Control Theory are drawn from Cybernetics, the
science of feedback processes for the control and regulation of engineering systems (see
Weiner, 1948). At the heart of both Cybernetics and Control Theory is the negative
feedback loop, or discrepancy-reducing loop, the mechanism that affords system self-
regulation in the pursuit of a desired outcome or state. In one of its earliest
manifestations, Miller, Galanter, and Pribram (1960) referred to this loop as a Test-
Operate—Test-Exit (TOTE) unit. This unit “tests” for discrepancies between the system’s
desired state and the environment’s actual state, “operates” on the environment to reduce
apparent discrepancies, “tests” again for discrepancies, and “exits” the loop when
discrepancies are absent.

However, a more contemporary application of the discrepancy-reducing loop is
presented in Carver and Scheier’s (1982) initial explanation of Control Theory. Their
theory is based on a negative feedback loop consisting of four system components: an

input function, a reference value, a comparator, and an output function. The purpose of

the input function, or sensor, is to obtain information from the environment and introduce
it into the system. Put another way, the sensor “perceives” feedback from the external
environment to aid in self-regulation. The reference value is an internally-derived
standard by which the input function’s feedback is to be judged. This reference standard
is commonly referred to as a goal; an objective or state a self-regulating system is trying
to achieve or maintain (Lord & Hanges, 1987). As its namesake implies, the comparator
compares the sensed external information ﬁom the input function to the internal reference
standard (i.e., goal) held by the system in order to determine if the values are discrepant.
When a discrepancy exists, the output function, or effecter, initiates a response for the
purposes of altering the environment. That is, an effecter “behaves” in a particular
manner in order to generate feedback that is more consistent with the reference standard.
In subsequent work (i.e., Carver & Scheier, 1990; 1998), the authors expand and
revise their theory to discuss the role of affect as a motivational inﬂuence in self-
regulation. This is accomplished by proposing two nested negative feedback loops: an
inner “action loop” and an outer “meta loop” (see Figure 1 in Appendix B). The action
loop encompasses the discrepancy-reducing processes that take place in the negative
feedback loop as described previously. In this inner loop, self-regulation consists of a
monitoring function focused on the amount of discrepancy present. Thus, feedback in the
action loop is positional. That is, position feedback is used in the action loop to
determine the amount of discrepancy present at any given time. If discrepancies are
present, the action loop triggers behaviors intended to reduce the likelihood of future

discrepancies.

Although the outer meta loop is comprised of the same four components found in
the inner action loop, the meta loop performs a “meta monitoring” function that is
concerned with the rate of change in discrepancy reduction taking place in the action
loop. Carver and Scheier (1990; 1998) refer to this type of feedback used in the meta
loop as “velocity” because the rate of change in distance (i.e., the ﬁrst derivative of
distance) is called velocity in the physical sciences. So, while the action loop uses
positional feedback to identify discrepancies between the current state and some desired
state, the meta loop uses velocity feedback to determine if there is a discrepancy between
the current rate of discrepancy reduction and some “acceptable rate of behavioral
discrepancy reduction,” (p. 122) which serves as the meta loop’s reference value.
Furthermore, the action and meta loops generate different outcomes in the event of a
discrepancy. Whereas discrepancies in the action loop trigger discrepancy reducing
behaviors, Carver and Scheier suggest that discrepancies in the meta loop generate affect.

Affect generated in the meta loop serves to motivate the intensity of discrepancy
reducing behaviors (Carver & Scheier, 1990; 1998). Borrowing from Higgins’ (1987)
Self-Discrepancy Theory, Carver and Scheier classify affect as positive (i.e., elation or
joy), neutral, or negative (i.e., depression). When the rate of discrepancy reduction in the
meta loop is perceived to be consistent with the reference standard, affect is neutral and
the intensity of discrepancy reducing behaviors is unchanged. However, when the rate of
progress is slower than the standard, or when there is no progress whatsoever, affect is
negative and in proportion to the size of the discrepancy. The experience of depression
motivates the system to intensify its discrepancy reducing behaviors. By contrast, when

the rate of progress is faster than the standard, affect is positive and proportional to the

discrepancy. It is the experience of elation that motivates the system to scale back on
excessive discrepancy reducing behaviors.

Carver and Scheier (1990; 1998) also note that people are able to sense
acceleration, or changes in velocity (i.e., changes in the rate of change in discrepancy
reduction), and assert that the presence of acceleration leads to changes in affect.
Speciﬁcally, they propose that a sensed increase in velocity (i.e., positive acceleration)
results in a change in affect that is manifested as a shift toward a more positive
experience of elation. Conversely, a sensed decrease in velocity (i.e., negative
acceleration or deceleration) results in a change in affect characterized by a shift toward a
more negative experience of depression. While this explanation suggests the presence of
an “uber-loop” that encapsulates both the meta and action loops to monitor acceleration
feedback, Carver and Scheier do not speciﬁcally propose a third feedback control loop.
Instead, they suggest that it is possible to sense a quality such as acceleration without a
feedback loop.

Each of the theories discussed in this section provide an adequate explanation for
self-regulatory behavior. Indeed, the three theories have much in common as each
requires (1) a goal or standard that represents an objective or state that is desirable to
individuals, (2) individuals engage in behaviors that are expected to achieve their goals,
(3) feedback that indicates individuals’ progress toward their goals, and (4) individuals
evaluate their performance and take corrective action in the event of a discrepancy.
However, Control Theory is particularly interesting because it proposes that position,
velocity, and acceleration information are distinct forms of feedback that produce

different behavioral or affective responses during self-regulation. In the following

10

section, I will conceptually deﬁne position, velocity, and acceleration as they are
presented in the physical sciences, review the limited number of studies that have applied
these constructs in the social sciences, and discuss the appropriate translation of these
constructs for self-regulatory research.
Feedback Frames: Position, Velocity & Acceleration Information

Carver and Scheier’s (1990; 1998) conceptions of position, velocity, and
acceleration feedback are derived ﬁ'om the same properties discussed in the physical
sciences. Speciﬁcally, the Law of One-Dimensional Motion deﬁnes the position function
of an object as x(t) = l/2at2 + vt + c, the velocity function as v(t) = at + v, and the
acceleration function as a(t) = a (where a is the object’s acceleration, v is the object’s
velocity, and c is the object’s position at time t). Using basic calculus, it can be shown
that v(t) is the ﬁrst derivative of x(t) and indicates the rate of change in position or the
change in position over time. Furthermore, a(t) is the ﬁrst derivative of v(t), the second
derivative of x(t), and indicates the rate of change in velocity over time (i.e., the rate of
change in the rate of change in position over time). Graphically, position is often
portrayed as a point on an axis at a given time, velocity is depicted as a slope indicating
the rate of change in that position over time, and acceleration is illustrated by a curve
revealing the rate of change in velocity over time, whereby any line (i.e., slope) drawn
tangent to the curve is the velocity at that time.

Despite the popularity of Control Theory, the number of published studies
examining velocity or acceleration feedback can be counted on just one hand. Working
independently of Carver and Scheier (1990; 1998), Hsee and associates conducted a

series of three studies exploring the relationship between outcomes, ﬁ'amed in terms of

11

position, velocity, and acceleration, on individuals’ levels of satisfaction. In the ﬁrst
study, Hsee and Abelson (1991) proposed that individuals not only derive satisfaction
from the position of outcomes (e.g., a stock valued at $100), but also the velocity of
outcomes (e.g., a stock rapidly appreciating from $50 to $100). To test their hypothesis,
they presented individuals with pairs of hypothetical outcomes describing different
positions and velocities and asked them to select the outcome that they found more
satisfying. The data revealed that individuals preferred (1) a more positive outcome to a
less positive outcome, (2) an outcome preceded by a rapid rise to the same outcome
preceded by a slow rise, (3) an outcome preceded by a slow fall to the same outcome
preceded by a fast fall, (4) an outcome preceded by a small but fast rise to the same
outcome preceded by a large but slow rise, and (5) an outcome preceded by a large slow
fall to the same outcome preceded by a small fast fall. This study established that
outcome satisfaction is related to both the position and velocity of outcomes.
Furthermore, it demonstrated that individuals do recognize and respond to velocity
information in predictable ways. Speciﬁcally, more positive velocities are associated
with greater satisfaction, while more negative velocities are associated with less
satisfaction.

In a companion study, Hsee, Abelson and Salovey (1991) hypothesized that an
individual’s total satisfaction with an outcome was the sum of their position and velocity
satisfactions. To test this hypothesis, they obtained baseline satisfaction ratings of two
outcomes: a negative velocity resulting in a higher outcome position and a positive
velocity culminating in a lower outcome position. The authors then asked participants to

provide satisfaction ratings of additional outcome pairs in which either position or

12

velocity information was made salient. They found that ﬁ'arning outcomes in position-
and velocity-salient ways altered the relative contributions of position and velocity
satisfaction, respectively, to the total amount of reported satisfaction for the outcomes.
This study was signiﬁcant as it demonstrated that the framing of outcomes in terms of
position or velocity does affect individuals’ decisions, a conclusion that is consistent with
Kahneman and Tversky’s (197 9; 1984) research of the effect of positive or negative
semantic frames on decision-making in situations involving risk.

In their ﬁnal study of the outcome satisfaction relations, Hsee, Salovey and
Abelson (1994) proposed that individuals derive satisfaction from the acceleration of
outcomes as well as their position and velocity. After presenting pairs of outcome curves
with common displacements and velocities but different accelerations, they found that
participants were more satisﬁed with curves depicting more positive accelerations.
However, they recognized that all positive acceleration curves used in the study also
portrayed a positive tail-end velocity which constituted a potential confound in their
study. To clarify their results, they conducted a second study in which they provided
participants with outcome curves featuring varying displacements but identical velocities
at the tail ends of their curves, and asked then to provide satisfaction ratings continuously
as the curves were slowly drawn on a computer screen. The data revealed that
participants’ reported increasingly more positive satisfaction ratings when the tail ends of
the acceleration curves grew more positive. Apparently, in addition to position and
velocity information, the authors demonstrated that individuals are also responsive to

acceleration information.

13

Collectively, these studies by Hsee and colleagues demonstrate that people do
respond differently to similar outcomes ﬂamed in ways that make different aspects of a
given performance or outcome salient. However, in each case, the authors simply asked
participants to rate their level of satisfaction with various hypothetical outcomes.
Without the presence of a clear goal, it is difﬁcult to explain the observed results in terms
of Control Theory, as the comparator requires a reference value in order to evaluate
feedback and identify a discrepancy. Consequently, it is neither clear what referent
standard participants had in mind when making their satisfaction ratings of position-,
velocity-, or acceleration-framed outcomes, nor how these different feedback ﬂames
might inﬂuence their reactions when immersed in a goal-directed environment or activity.

To date, Lawrence, Carver and Scheier (2002) have published the only study that
incorporates velocity feedback and a behavioral goal in the discrepancy-reducing context
of Control Theory. In an apparent test of Control Theory’s meta-loop as an affect
generator, the authors designed an ambiguous performance task that requested
participants rely on their “social intuition” to identify which of 60 foreign words
presented in 6 trials of 10 items were paired with their English translations. In fact, the
foreign words were bogus and the task was sufﬁciently ambiguous that participants were
forced to rely on the experimenter-provided feedback to determine how they were
performing. The use of this type of task allowed the authors to plausibly manipulate
performance feedback and subsequently assess participants’ reactions to their feedback
(i.e., participants’ mood ratings obtained before trial 1 and after trial 6).

The participants’ stated action-loop goal was to maximize the number of correct

responses at the end of the trials and a guess rate of 50% was provided to serve as a meta-

14

loop reference value. Because Lawrence et a1. (2002) were not concerned with position
feedback or action loop processes in this study, they only brieﬂy discuss the role of
participants’ cumulative number of correct responses as an indication of action loop
discrepancy reduction (i.e., position). Instead, they focused on the provision of velocity
feedback, which they operationalized as the general slope of participants’ performance
over trials (i.e., the trend line ﬁt to a plot of the number of correct items per trial).

The authors assigned participants to one of ﬁve velocity feedback conditions (see
Figure 2). In the “high positive velocity” condition, participants were informed after
each trial that they had correctly identiﬁed 9, 8, 8, 7, 6, and 5 items, respectively. The
authors hypothesized that participants’ reported moods would decrease (i.e., become
more negative) from time 1 to time 2. In the “low positive velocity” condition,
participants were informed after each trial that they had correctly identiﬁed 1, 2, 2, 3, 4,
and 5 items, respectively, and the authors hypothesized that participants’ reported moods
would increase (i.e., become more positive) ﬂom time 1 to time 2. The middle three
velocity conditions were described as medium-high, intermediate, and medium-low
positive velocities; offered commensurate feedback; and were hypothesized to produce
smaller or negligible mood changes in a pattern consistent with the more extreme
conditions. Upon analyzing the data, Lawrence et a1. (2002) found that participants
receiving the increasingly more positive velocity feedback (i.e., those in the low positive
velocity condition) reported more positive mood change, and those receiving the
increasingly less positive velocity feedback (i.e., those in the high positive velocity

condition) reported more negative mood change.

15

However, Lawrence et al.’s (2002) operationalization of velocity feedback is
problematic. For example, the two upper trend lines portrayed in Figure 2 are clearly
downward sloping (i.e., negative), yet are supposed to represent high and medium-high
positive velocities, respectively. The authors assert that despite the decreasing
performance across trials that contribute to these negative slopes, the two conditions
actually exhibit positive velocities because participants are still approaching the action
loop goal of maximizing the total number of correct responses. In short, they explain that
all velocities in this study are positive, and negative velocities can only occur if
participants were to “lose” points for incorrect items.

This confusing discrepancy can be avoided by deﬁning velocity in a manner that
is consistent with the construct as it is applied in the physical sciences. That is, velocity
feedback should be operationally deﬁned as a rate of change in position (i.e., discrepancy
reduction) rather than a trend line or slope ﬁt to participants’ performance across trials.
Thus, if a participant receives 90% of the items correct in trial 1, then their position is 9
with respect to their action loop goal of maximizing the total number of correct responses
(i.e., a maximum value of 60), and their rate of change in discrepancy reduction (i.e.,
velocity) is positive 90%. Put simply, the percentage of items participants correctly
identify in a trial is their velocity because it represents their current rate of discrepancy
reduction (e. g., they are currently moving toward the goal value of 60 at a rate of 9 items
per trial).

This error in operationalization appears to be the result of confusing velocity
delivered as verbal feedback with velocity delivered as visual feedback. Recall that

velocity is often verbally discussed as a rate, but visually portrayed as a slope. Lawrence

16

et a1. (2002) have focused on the representation of velocity as a slope and discuss it in
terms of the “slopes of the success curves over trials” shown in Figure 2 (p. 793). The
result of this inappropriate operationalization of velocity is a negative slope-positive
velocity discrepancy. A more accurate and appropriate depiction of velocity is a slope
indicating the rate of discrepancy reduction. For example, if the rate of discrepancy
reduction is 90%, then a visual depiction of that rate would be a line with a slope of 9
(i.e., a line with a rise of 9 and a run of 1). In short, Figure 2 does not depict velocity
feedback as a slope; it merely depicts a trend in the verbally-delivered velocity feedback
plotted across trials. To plot each velocity feedback slope across trials would reveal a
pattern resembling an acceleration curve.

This problem contributes to two errors in Lawrence et al.’s (2002) study. First, it
led them to provide less than adequate velocity information to participants. Speciﬁcally,
after each trial, they intended to provide velocity feedback to participants, yet actually
reported the number of items each participant correctly identiﬁed. That is, they should
have provided the rate of change in discrepancy reduction (i.e., percent of items correct
as velocity-salient information), but instead provided the number of items correct (i.e.,
information that is more position- than velocity-salient). In this instance, there are only
ten items per trial so the number of items correct is easily transformed into a velocity
(e. g., 9 is 90%). However, we do not know if this information was transformed by
participants or if it was interpreted as position or velocity (or even acceleration)
information. This is not a trivial issue as Hsee et a1. (1991) found that by framing
outcomes in position-salient and velocity-salient ways, they could alter participants’

levels of satisfaction with similar outcomes. Thus, the ﬂaming of outcomes can lead

17

participants to respond to similar outcomes in different ways and researchers should take
care to ensure that position feedback is position-salient and velocity feedback is velocity-
salient.

The second error the velocity operationalization problem spawned concerns the
source of the mood ratings provided by participants. Because Lawrence et a1. (2002)
improperly operationalized velocity as the “slopes” displayed in Figure 2, they were
content to measure participants’ mood at baseline and at the end of trial 6. However, the
lack of clear velocity feedback (i.e., rate of discrepancy reduction) provided at the end of
each trial leaves the object of participants’ ratings in doubt. Did participants interpret
their feedback (i.e., number of items correct) as position or velocity feedback? Also, is it
possible that participants in the slowly declining and slowly rising performance
conditions (i.e., the high positive velocity and low positive velocity conditions pictured at
the top and bottom of Figure 2, respectively) derived some acceleration information ﬂom
their feedback? If so, this could also explain why the former group reported more
negative mood change and the latter group more positive mood change as such results are
consistent with Hsee et al. (1994). In short, given their problems in operationalizing
velocity, the immediate cause of Lawrence et al.’s observed results is not clear.

Perhaps the most recent study involving velocity feedback is an unpublished
conference presentation by Johnson and Lord (2005) concerning the joint effects of
performance goal discrepancies and progress rates on individuals’ motivation. In their
study, participants were asked to achieve an action loop performance goal and a meta
loop progress rate goal while engaged in an ambiguous performance task that required

them to learn artiﬁcial rules of grammar. The authors manipulated both the size of the

18

participants’ performance goal discrepancies (i.e., large or small) and their rates of
progress (i.e., slow or fast) when providing position and velocity feedback after each
trial. Motivational outcome variables including task satisfaction, task persistence, goal
commitment and success expectancy were assessed at the conclusion of the trials.
Analysis of the data revealed that both performance goal discrepancy (i.e., position
feedback) and progress rate (i.e., velocity feedback) were related to participants’
motivation: Speciﬁcally, small performance goal discrepancies predicted task
satisfaction, success expectancy, and goal commitment, while faster progress rates
predicted success expectancy. However, the most interesting ﬁnding of this study was a
signiﬁcant discrepancy x progress rate interaction for task satisfaction, success
expectancy, and goal commitment, which suggests that experiencing faster progress rates
may be capable of erasing the negative motivational effects of large performance goal
discrepancies.

In summary, despite its introduction nearly 25 years ago, just ﬁve studies have
explored the nature of velocity feedback, one study has addressed the use of acceleration
feedback, and only two of the four studies have employed velocity feedback in a
discrepancy-reducing context consistent with Control Theory. Consequently, not much is
known about how position, velocity, and acceleration feedback are related and how they
should be applied during goal-directed pursuits. One lesson to be learned ﬂom Lawrence
et a1. (2002) is that position, velocity, and acceleration feedback should be deﬁned in a
manner that is consistent with the use of these properties in the physical sciences. By
doing so, we are able to maintain the conceptual relationships between the constructs and

the utility of the metaphor. Given the importance of properly deﬁning and implementing

19

each feedback ﬂame, I will now provide conceptual and operational deﬁnitions and
present examples of verbal and graphical feedback for each ﬂame as they might be
applied in a study using an ambiguous performance task paradigm as found in Lawrence
et al. (2002) and Johnson and Lord (2005).

Position feedback. In the action loop, position feedback is conceptually deﬁned
as feedback that describes the current level of performance as referenced to the desired
level of performance. Therefore, an appropriate operational deﬁnition of position
feedback for an ambiguous task would be the cumulative number of correct responses.
Thus, at the conclusion of each trial, verbal position feedback should emphasize only the
current level of performance using position-salient information. For example, if the
stated action loop goal is to maximize the total number of responses, then appropriate
position feedback for a participant who correctly responds to 9 items after the ﬁrst trial is,
“thus far, you have correctly responded to a total of 9 items.” When multiple trials are
used, then the cumulative number of correct responses should be provided at the end of
each trial to indicate the current level of performance relative to the desired standard.
Similarly, the visual representation of position feedback should only emphasize position-
salient information (i.e., the current level of performance relative to the desired level of
performance). Figure 3 provides an example of position feedback for a participant
correctly responding to 9 items in their ﬁrst trial. Note that the use of a “thermometer” or
“tape” display draws attention to the position of the current level of performance (i.e., 9
correct responses thus far) relative to the desired level of performance (i.e., 100 correct

responses) as displayed on the y-axis.

20

Velocity feedback. Velocity feedback is conceptually deﬁned as the rate of
change in discrepancy reduction between the current level of performance and the desired
level of performance. As previously discussed, velocity feedback should be operationally
deﬁned as the percentage of items correct per trial in the context of an ambiguous
performance task. Therefore, verbal velocity feedback should emphasize velocity-salient
rate information (e.g., “You correctly responded to 90% of the items in this trial”).
Similarly, visual velocity feedback should remove any references to position and provide
current rate information in the form of a slope. As shown in Figure 4, this depiction of
velocity feedback draws attention to the slope of the line, while downplaying positional
information. Speciﬁcally, the x-axis is used to emphasize the velocity-salient rate of 90%
rather than the position-salient cumulative score of 9 and the higher possible scores
emphasized in the positional graphic shown in Figure 3.

Acceleration feedback. Acceleration feedback is conceptually deﬁned as the rate
of change in velocity (i.e., the rate of change in the rate of change in discrepancy
reduction) and operationally deﬁned as the change in current velocity relative to past
velocity. Thus, verbal acceleration feedback should emphasize changes in velocity (e.g.,
“You correctly responded to 80% of the items in this trial; this is 10% less than in the last
trial”). Visually, acceleration feedback should be represented as a curve in order to
emphasize changes in velocity over time. Figure 5 shows a decelerating performance
curve; note that drawing a series of tangent lines (i.e., slopes) at various points on the
curve will reveal that velocity is decreasing over time. Depictions of pure acceleration
feedback should minimize velocity- and position-salient information by smoothing

curves (as opposed to presenting a series of connected velocity slopes) and removing

21

positional references on the x-axis, respectively. The use of axes with indeterminate
bounds assists in de-emphasizing position information.

Thus, position, velocity, and acceleration feedback, properly ﬂamed, can
communicate unique aspects of performance information to learners and trainees. But
what is the practical utility of such ﬂames? The position x velocity feedback interaction
reported by Johnson and Lord (2005) provides some indication of their potential uses. As
previously mentioned, the authors found that the relationship between participants’
reports of motivational outcomes (i.e., task satisfaction, success expectancy, and goal
commitment) and performance-goal discrepancy (small or large as described by position
feedback) depended on progress rate (slow or fast as described by velocity feedback).
Speciﬁcally, when progress rate was fast, the relationship between perfonnance- goal
discrepancy and motivational outcomes was negligible. However, when progress rate
was slow, the relationship was moderately positive. This suggests that the potentially
demotivating effects of large performance-goal discrepancies may be offset by the
presence of fast discrepancy-reducing progress rates. In the following section, I will
demonstrate how this ﬁnding can be used to beneﬁt individuals during learning or
training.

Learning or Skill Acquisition Performance Proﬁles Over Time

“Practice makes perfect” is a maxim that refers to the nearly universal observation
that individuals’ task performance, as measured by speed or accuracy, tends to improve
with practice over time (N ewell & Rosenbloom, 1981). More speciﬁcally, the beneﬁts of
practice demonstrate a nonlinear relationship with performance, such that performance

improvement is rapid during early stages of task learning, but diminishes over time as

22

individuals become more skilled. This pattern has been found for such a wide variety of
perceptual-motor and cognitive skill tasks (see Newell & Rosenbloom, 1981 for a
review), that contemporary researchers describe it as “ubiquitous” and refer to it as the
Power Law of Practice, Power Law of Learning, or more simply, the “learning curve”
(e.g., Anderson, 1982; Logan, 1988; 1992; Ritter & Schooler, 2001; VanLehm, 1996).

Scholars have proposed a number of mechanisms to explain the observed
relationship between learning and performance, including Anderson’s (1982)
composition and strengthening of knowledge, Crossman’s (1959) identiﬁcation of ideal
task methods, Logan’s (1988) memory retrieval speeds, MacKay’s (1982) strength of
network connections, and Newell and Rosenbloom’s (1981) chunking of stimulus
patterns. Despite their differences in proposed mechanisms, all theories agree that the
learning curve is a negatively accelerating (i.e., decelerating) curve characterized by an
initial period of rapid performance gains that diminish over time.

Figure 6 presents a prototypical learning curve that is arbitrarily bisected into
early and late performance periods. Although acceleration is negative (i.e., decelerating)
across the two periods, position and velocity are different in early and late periods.
Speciﬁcally, in the early period, position is low relative to the standard (e. g., the upper
limit value on the x-axis) indicating a large discrepancy between current performance and
the goal, and velocity is positive and steep indicating a rapid rate of discrepancy
reduction. However, in the late period, current position is high and close to the standard
indicating a small discrepancy, and velocity remains positive, but shallow, indicating a
slow rate of discrepancy reduction. The difference between early and late period

performance suggests the potential for different feedback ﬂames to be used to maintain

23

learner motivation during different periods of the learning curve by emphasizing different
characteristics of the same performance.

For example, consider a newly-hired Transportation and Safety Administration
Airport Screening Technician undergoing training to recognize 100 types of forbidden
items via x-ray during a 10 week course. There are many ways for the trainer to provide
weekly feedback to the technician, including position-salient (i.e., number of items
recognized) or velocity-salient (i.e., rate of item recognition) performance feedback. As
item recognition is a learning task, we would expect the technician’s learning
performance to follow the Power Law of Learning; that is, we would expect rapid
recognition performance improvements during the early weeks of training as the easy
items are learned, but diminishing gains over time, and near stagnation during the ﬁnal
weeks of training as the technician struggles to recognize the more obscure or difﬁcult
items. During the opening weeks of training, which type of feedback is likely to be more
motivating to the technician: “You have learned to recognize 10 items” (i.e., position
feedback) or “You are learning to recognize items at a rate of 10 per week” (i.e., velocity
feedback)? In the ﬁnal weeks of training, which type of feedback would be most
motivating: “You have learned to recognize 98 items” (i.e., position feedback) or “You
are learning to recognize items at a rate of 1 per week” (i.e., velocity feedback)? In short,
not only might individuals respond differently to feedback ﬂamed in different ways, but
their response might also vary with time and their location on the learning curve (e. g.,
early or late). Thus, learner motivation might beneﬁt ﬂom performance feedback that is

ﬂamed in one way during one segment of a learning curve, and some other way during

24

another segment. I will now explore this potential by describing several measures of
motivation to be assessed in this study before introducing my research hypotheses.
Motivational Indicators

With the exception of Johnson and Lord (2005), past research on velocity and
acceleration feedback has only measured participants’ outcome satisfaction and mood
states. Yet, indicators of task motivation can be classiﬁed as affective (e.g., mood,
emotion), behavioral (e. g., effort, persistence), or cognitive (e.g., self-efﬁcacy, self-
evaluation). This study uses state affect, task effort, task self-efﬁcacy, goal commitment,
and self-satisfaction with performance to examine participants’ motivational responses to
various types of feedback. A broad range of dependent variables was selected in order to
enhance the study of feedback ﬂames by incorporating motivational measures that are
commonly used in research on self-regulation. While the typical self-regulation study
fails to differentiate between types of feedback, the combination of feedback ﬂames and
multiple reaction measures in this study provides a unique opportunity to exarrrine how
they interact.

State aﬂect. Despite a long history of neglect, the study of affect in the workplace
has experienced a renaissance in recent decades (Brief & Weiss, 2002). Since the 1980s,
a substantial and growmg body of cross-cultural literature suggests that affective
structure is most appropriately deﬁned by two dominant, independent dimensions of
positive and negative affect (Burke, Brief, George, Roberson & Webster, 1989; Diener,
Larson, Levine & Emmons, 1985; George, 1996; George & Brief, 1996; Meyer & Shack,
1989; Russell, 1980; 1983; Watson, Clark & Tellegen, 1988; Watson & Tellegen, 1985;

Zevon & Tellegen 1982). Individuals experiencing a positive affective state may feel

25

alert, excited, determined, strong, and active, while those experiencing a negative state
affect may feel anxiety, hostility, fear, shame, or guilt (George & Brief, 1996; Watson et
al., 1988).

There are some scholars who treat affect as information. For example, Frijda
(1988) suggests that affective experiences “signal states of the world that have to be
responded to, or that no longer need response and action” (p. 354) and Schwarz (1990)
views feelings as information capable of motivating individuals to attend to particular
environmental stimuli. Consistent with Social Cognitive Theory and Control Theory,
workers who obtain performance information ﬂom their environments are likely to react
affectively to this information and self-regulate accordingly. Speciﬁcally, George and
Brief (1996) suggest that affect can increase or decrease workplace motivation and,
subsequently, job performance. Recent related research is begimring to provide
supporting evidence as Cote (1999) reported pleasant state affect was associated with
better performance and unpleasant affect was associated with worse performance, among
samples of students, salespersons, and professional athletes. In general, feedback that is
interpreted favorably should generate positive affect while unfavorable feedback should
generate negative affect.

Task self-eﬂ‘icacy. Wood and Bandura (1989) deﬁne self-efﬁcacy as “beliefs in
one’s capabilities to mobilize the motivation, cognitive resources, and courses of action
necessary to meet given situational demands” (p. 408). Such beliefs inﬂuence all aspects
of the self-regulatory process (Bandura, 1986; 1991). For example, compared to
individuals with low self-efﬁcacy, those with high self-efﬁcacy tend to set more

challenging goals (Bandura & Cervone, 1986; Locke, Frederick, Lee & Bobko, 1984;

26

Locke & Latham, 1990; Wood & Bandura, 1989), exert greater effort and persevere in
the face of obstacles (Bandura & Cervone, 1983; 1986), and spend less time dwelling on
their deﬁciencies (Ozer & Bandura, 1989). Furthermore, there is substantial evidence of
a link between self-efﬁcacy and performance; Stajkovic and Luthans (1998) found a
weighted average correlation of .38 between self-efﬁcacy and job performance, leading
them to conclude that self-efﬁcacy may be a more useﬁrl predictor of performance than
many personality measures.

According to Wood and Bandura (1989), self-efﬁcacy can be strengthened over
time by providing successful practice experiences and realistic encouragement, both of
which are generally accompanied by performance feedback. Empirically, the provision
of positive feedback has been shown to be an optimal way to increase self-efﬁcacy (Karl,
O’Leary-Kelley & Martocchio, 1993; Martocchio & Webster, 1992). Thus, feedback that
is evaluated favorably should increase task self-efﬁcacy, while unfavorable feedback
should decrease task self-efﬁcacy.

Task effort and self-satisfaction with task performance. Bandura’s (1986; 1991)
Social Cognitive Theory proposes that motivation results ﬂom anticipated self-
satisfaction with future performance (or, alternately, self-dissatisfaction with past
performance). When a self-monitoring function provides performance feedback, the self-
evaluation function determines the magnitude of a perceived discrepancy between the
current and desired state, and the self-reaction function generates feelings of satisfaction
or dissatisfaction as appropriate. In short, the desire to feel self-satisfaction with

performance provides the incentive to mobilize greater effort. Thus, favorable feedback

27

should bolster ratings of self-satisfaction and increase task efforts, while unfavorable
feedback should have the opposite effect.

Goal commitment. Goal commitment has been deﬁned as the determination to try
for a goal (Locke et al., 1981) and it implies a persistence of effort over time and an
unwillingness to lower or abandon that goal (Campion & Lord, 1982). As with the
preceding measures, feedback that is interpreted favorably should make the goal appear
more achievable, enhance learner persistence of effort, and build commitment to that
goal. When feedback is viewed unfavorably, goal commitment is at risk as effort is
withdrawn. These responses are anticipated in the following research hypotheses.
Research Hypotheses

One way to examine the motivational effects of different forms of feedback on the
preceding motivational indicators is to explore the interactions between feedback ﬂame,
learning curves (or performance proﬁles), and time. I have already presented the three
feedback ﬂames (i.e., position, velocity, and acceleration feedback), discussed the role of
time (i.e., early and late), and introduced the learning curve (i.e., a decelerating
performance proﬁle). Before I proceed, I will introduce a second performance proﬁle
that stands in contrast to the ﬁrst.

As previously mentioned, the learning curve has been described as “ubiquitous”
due to its applicability to a wide variety of tasks (N ewell & Rosenbloom, 1981).
However, the ubiquity of the decelerating learning curve may simply reﬂect the kinds of
tasks studied by its proponents. For example, many of the simple tasks cited by Newell
and Rosenbloom (e.g., rolling cigars, reading inverted text, choice reaction tasks) afford

rapid, initial performance gains because performers can begin work immediately, acquire

28

task knowledge and skill, and realize these gains. Yet more complex tasks may require
some degree of preparation or investment of effort before any signiﬁcant performance
gains can be made. For example, in a code breaking task, the code breaker is unlikely to
produce useful text until the code has been cracked. Similarly, if the task is to prevent a
nuclear reactor meltdown, then the reactor core temperature may not be signiﬁcantly
reduced until systems are statused, the problem is diagnosed, and appropriate action is
taken. In short, more complex tasks may require initial amounts of trial and error, fact-
ﬁnding, or knowledge integration before any signiﬁcant progress can be made. Such
tasks might be characterized by an accelerating learning curve deﬁned by an initial period
of slow performance gains that grow over time (see Figure 7). Like the decelerating
curve, an accelerating curve exhibits large positional discrepancies in its early period but
small discrepancies in its later period. However, the two curves are different when
velocity is considered. Speciﬁcally, an accelerating curve possesses higher velocities in
its early period and lower velocities in its later period; this pattern is in opposition to that
of a decelerating curve.

In sum, individuals following the decelerating performance proﬁle shown in
Figure 6 will achieve rapid performance gains in the early period, but greatly diminished
returns in the late period. In contrast, those following the positively accelerating
performance proﬁle shown in Figure 7 will experience tepid initial performance gains in
the early period followed by rapid performance gains in the late period. For the purposes
of this study, combining the two contrasting learning performance proﬁles (i.e.,
decelerating and accelerating), several time periods (i.e., baseline, early, and late), and

three feedback ﬂames (i.e., position, velocity, and acceleration) provides an opportunity

29

to explore how different feedback ﬂames describing the same level of performance might
differentially beneﬁt learner motivation during different stages of a learning task.
Performance proﬁle x time interaction. There is reason to believe that
performance proﬁle will interact with time for motivational outcomes such as state affect,
task self-efﬁcacy, self-satisfaction with task performance, goal commitment, and task
effort. Hsee and Abelson (1991) found that people were more satisﬁed with a high
positive velocity outcome as compared to a low positive velocity outcome with the same
displacement, and more satisﬁed with a low negative velocity outcome as compared to a
high negative velocity outcome with the same displacement. Despite their problematic
operationalization of velocity, participants in Lawrence et al.’s (2002) study reported
more positive moods after experiencing more positive velocities. Finally, Johnson and
Lord (2005) found progress rate was positively related to task satisfaction and
expectations for success. These results suggest that velocity feedback may drive this
interaction as high positive velocities are present in the early period of the decelerating
performance proﬁle and late period of accelerating performance proﬁle condition.
Although there is no direct empirical evidence that acceleration feedback will
help velocity drive this interaction, Hsee et al.’s (1994) study of participants’ satisfaction
with accelerations curves with different displacements and velocities does provide some
support. Speciﬁcally, Hsee et al. found that participants’ ongoing reports of satisfaction
grew more positive when participants perceived a sudden increase in a constant velocity
(i.e., acceleration). In this study, participants in the decelerating performance proﬁle
condition will experience a diminishing rate of performance gains; therefore, their pattern

of motivation scores should decrease over time. In contrast, participants in the

30

accelerating performance condition will experience an increasing rate of gains and should
report greater motivation over time.

In sum, regardless of the feedback ﬂame provided, I hypothesize a performance
proﬁle x time (P x T) interaction such that participants in the decelerating performance
proﬁle condition will report increasingly less positive (and more negative) affect, task
effort, task self-efﬁcacy, satisfaction with performance, and goal commitment over time
as compared to their counterparts in the accelerating performance proﬁle condition.
Conversely, participants in the accelerating performance proﬁle condition will report or
demonstrate increasingly greater motivation over time relative to those in the decelerating
performance proﬁle condition (see Figure 8).

Feedback frame x performance proﬁle x time interaction. An understanding of
the different feedback ﬂames and their referents suggests the potential for a three-way
interaction between feedback ﬂame, performance proﬁle, and time. In Control Theory,
the referent for position feedback in the action loop is the goal, and the referent for
velocity feedback in the meta-loop is some “acceptable rate” of discrepancy reduction
(Carver & Scheier, 1988, p. 122). It is interesting to note that the source of an action loop
goal is often easily identiﬁed (i.e., self-set, dictated by the task, or supervisor-directed),
but the origin of the meta-loop referent is not. While Carver and Scheier do not speculate
as to the source of the acceptable rate, it is likely determined by the individual based on
their expectation of the rate of discrepancy reduction resulting ﬂom an initial
performance or their observations or knowledge of others’ performances and associated
rates of discrepancy reduction. To complicate matters, as performance improves, both

the individual’s actual rate of discrepancy reduction (i.e., their velocity) as well as their

31

acceptable rate of discrepancy reduction (i.e., the meta-loop standard) may change. Even
less is known about the referent for acceleration feedback, as Carver and Scheier (1998)
only brieﬂy mention the role of acceleration as the source for changes or shifts in the
intensity of experienced affect. While they do not discuss acceleration as a source of
feedback monitored in a third “uber-loop” encompassing the action and meta-loops,
presumably, individuals respond to discrepancies between the sensed rate of change in
velocity and some acceptable rate of change in velocity in order to manifest their changes
in affect. In short, the nature and origin of the referent for acceleration feedback is likely
to be at least as ambiguous as the referent for velocity feedback, and is probably
determined by the individual due to the complexity of the concept.

Given what we know about feedback ﬂames and their referents, when learners
respond to position feedback, their motivation results ﬂom the discrepancy between their
current and desired positions (Carver & Scheier, 1998). In the early period of learning,
an individual’s position is more likely to be greatly discrepant relative to the goal, while
in the late period, the learner’s position is likely to be less discrepant. Consequently,
learners receiving position feedback will likely view it less favorably and report less
positive affect (and more negative affect), exert less task effort, report less task self-
efﬁcacy, satisfaction with performance, and goal commitment in the earlier portions of
their performance proﬁles as compared to their later portions (Hsee & Abelson, 1991;
Johnson & Lord, 2005). This result should be observed whether the leamer’s
performance is consistent with an accelerating performance proﬁle or a decelerating
performance proﬁle, as the position discrepancies will always be larger early on as

compared to later (see Figure 9a).

32

However, this pattern of results is not expected to be the case when the learner
receives either velocity or acceleration feedback. As previously mentioned, learning is
rapid in the early period of the decelerating performance proﬁle but greatly diminished in
the late period. Thus, a learner performing in a manner consistent with a decelerating
performance proﬁle will receive high positive velocity feedback and more “peaked”
acceleration feedback signaling changes in high velocities. As a result, during the early
portions of a decelerating performance proﬁle, the learner is less likely to encounter large
discrepancies with respect to their acceptable rates of velocity and acceleration, and
therefore more likely to view this feedback favorably and report more positive affect (and
less negative affect), exert more task effort, report more task self-efﬁcacy, more
satisfaction with performance, and more goal commitment (Hsee et al., 1994; Johnson &
Lord, 2005; Lawrence et al., 2002; see Figure 9b and 9c). However, over time, this
pattern of reactions will reverse. As the learner approaches the latter portion of the
proﬁle, velocity is low positive and the acceleration curve is ﬂatter. These conditions are
more likely to generate large discrepancies between current and acceptable rates of
velocity and acceleration. As learners view this feedback unfavorably, they are expected
to report less positive affect (and more negative affect), exert less effort, report less self-
efﬁcacy, satisfaction, and commitment (Hsee etal., 1994; Johnson & Lord, 2005;
Lawrence et al., 2002).

By comparison, velocity and acceleration feedback are expected to produce
opposing motivational effects in the contrasting accelerating performance proﬁle
condition. During the early portion of the proﬁle, learners will receive low positive

velocity feedback or ﬂatter acceleration feedback that is more likely to create large

33

 

discrepancies. When viewed unfavorably, learners will respond with reports of less
positive affect (and more negative affect), task self-efﬁcacy, satisfaction with task
performance, and goal commitment while also demonstrating less task effort (Hsee et al.,
1994; Johnson & Lord, 2005; Lawrence et al., 2002). As they transition into the latter
portion of the proﬁle, velocity and acceleration feedback will be high positive and more
peaked, respectively. Thus feedback-referent discrepancies are likely to be lower,
learners will evaluate their feedback more favorably, and they will report more positive
affect (and less negative affect), more satisfaction, more self-efficacy, more goal
commitment, and demonstrate more effort. In short, I hypothesize a feedback ﬂame x
performance proﬁle x time interaction such that the use of position performance feedback

deviates ﬂom the pattern of results found when using velocity or acceleration feedback.

34

METHOD

Design

This study employs a 3 x 2 x 2 ﬁrlly-crossed repeated measures research design
featuring (1) a between-subjects feedback ﬂame factor with three levels (i.e., position,
velocity, and acceleration), (2) a between-subjects performance proﬁle factor with two
levels (i.e., decelerating and accelerating), and (3) a within-subjects time factor with three
levels (i.e., baseline, early, and late). In addition, several individual difference covariates
(i.e., cognitive ability and trait anxiety) were included in the analyses.
Sample

Prior to data collection, a power analysis performed in accordance with the
procedures described by Murphy and Myors (2003) suggested a sample size of 180 (i.e.,
30 per cell) would provide at least 80% power to detect a small effect size (d = .20). The
initial sample consisted of 185 undergraduate psychology students ﬂom a large
Midwestern university who were recruited to participate in the experiment in exchange
for course credit. Two participants were subsequently removed for failing to complete
the study; a third was removed because she communicated via American Sign Language
and could not distinguish between similar words in the anxiety and affect scales. Thus,
the ﬁnal sample consists of 182 participants with a mean age of 22. It is predominately
female (71%) but tests of participants’ scores revealed no systematic, signiﬁcant
differences due to sex. Caucasians comprise 59% of the sample, followed by Aﬂican

Americans (23%), and Asian Americans (8%).

35

Procedure

Preparations. At the beginning of the study, participants were asked to provide
informed consent for their participation (see Appendix C) and to permit experimenter
access to their undergraduate admissions’ standardized test scores (i.e., SAT, ACT) via
the university admissions’ ofﬁce (see Appendix D). The test scores would serve as proxy
measures of cognitive ability.

Task The task employed in this study was the Social Intuition Task (SIT)
designed by Lawrence et al. (2002) and used in their initial study of velocity feedback in
a goal-directed, behavioral context. The SIT allows experimenters to plausibly
manipulate participants’ task performance in order to assess perforrner’s reactions to their
own performance feedback. Put simply, the task requires that participants determine if a
foreign word has the same meaning as an adjacent English word. In fact, the foreign
words are bogus and the participants’ resulting performance is indeterminate. A similar
performance task, the Implicit Grammar Task (IGT; Knowlton & Squire, 1996), has been
employed by others in related research (e.g., Johnson & Lord, 2005). The fundamental
requirement of both the SIT and the IGT is that the task stimulus be sufﬁciently
ambiguous as to require that participants attend to the experimenter-provided feedback in
order to determine how they are performing.

Task stimulus. Task stimulus materials consisted of 12 tests composed of 10
foreign-English word pair items for a total of 120 items. All 60 of the foreign-English
word pairs used by Lawrence et al. (2002) were incorporated into the 12 tests. The
remaining 60 foreign-English word pairs were derived ﬂom the Thai language.

Speciﬁcally, 60 common Thai words were converted to bogus foreign words by ﬁrst

36

 

spelling them phonetically, then transposing their syllables. The original and newly-
created word pairs were then evenly distributed in an alternating fashion across each of
the 12 tests (see Appendix E).

Background. Participants were told that the purpose of the study is to learn how
peOple extract meaning ﬂom foreign words. To lend credibility to the study and the task,
participants were provided the following background information:

“Recent work on social perception has discovered that people are sensitive to a far

broader range of nonverbal and paraverbal cues than had previously been

believed. Similarly, it is now thought that the evolution of languages was far less
random than was once believed, and that words in most languages actually have
subtle connections to the semantic qualities that the words were intended to
communicate. Though the connections are subtle, people often pick them up,
even when they don’t know the language they are hearing or seeing. This process
appears to happen largely outside of consciousness, and people usually cannot
identify very clearly when it is happening” (C. S. Carver, personal

communication, March 3, 2005).

Participants then completed a brief questionnaire packet requesting both useful
demographic information (i.e., sex, race, age, language ability or exposure, word game
experience), cover items that are seemingly related to one’s social intuition (e.g., number
of siblings, social interaction patterns), and the individual difference covariates (see
Appendix F).

Instructions, familiarization trial, and baseline data. Before the ﬁrst trial,

participants were read the instructions accompanying each test and they completed a 10-

37

item familiarization trial (see Appendix G). The purpose of the familiarization trial was
threefold. First, it veriﬁed that participants understood the trial’s instructions and how to
respond to items; any ambiguity was clariﬁed by the experimenter. Second, it made
salient the action loop goal of maximizing the number of correct responses (out of a
possible 120 items) across all trials, and provided an acceptable rate of discrepancy
reduction for the meta loop (i.e., 50% to reﬂect the probability of “guessing” on the
items). Finally, it provided the basis for participants to complete baseline assessments of
state positive and negative affect, task self-efﬁcacy, self-satisfaction with task
performance, goal comrrritrnent, and task effort (see discussion of measures below).

Experimental trials. At the start of each trial, participants were asked to review
the directions, complete each of the items on the test, and return it to the experimenter for
scoring. The experimenter then made a show of “scoring” the trials, generated
performance feedback that contained the manipulation, and provided the individualized
feedback to participants. After reviewing their feedback, participants began their next
trial.

The feedback manipulation materials consist of both textual and graphical task
performance feedback (see Appendix H). Participants randomly assigned to the position
feedback ﬂame condition received verbal position-salient feedback (e. g., “Thus far, you
have correctly responded to 9 items”) and a thermometer chart depicting this information.
Those assigned to the velocity condition received velocity-salient feedback (e.g., “You
have correctly responded to 90% of the items in this trial”) and a slope chart representing
this information. Finally, participants in the acceleration condition received acceleration-

salient feedback (e. g., “You have correctly responded to 80% of the items in this trial;

38

this is 10% less than the previous trial”) and a curve that was consistent with this
information.

Participants randomly assigned to the accelerating performance proﬁle condition
received position, velocity, or acceleration feedback indicating a gradual improvement in
goal progress (i.e., maximizing correct responses) over successive trials. Speciﬁcally,
participants receiving position feedback observed that the “mercury” in their thermometer
position charts rose ﬂom 1 to 17 during their ﬁrst six trials and ﬂom 17 to 60 during their
last six trials. Those participants receiving velocity feedback observed that the slope of
the rate of discrepancy-reduction line in their velocity charts ascended ﬂom 10 to 50 in
the ﬁrst six trials and ﬂom 50 to 90 in the last six trials. For those participants receiving
acceleration feedback, their position and velocity movements created a corresponding
curve in their acceleration charts that exhibited a slowly increasing slope in the ﬁrst six
trials and a rapidly increasing slope in the last six trials.

In contrast, participants randomly assigned to the decelerating performance
proﬁle condition received feedback suggesting a gradual slowing of goal progress over
successive trials. Participants receiving position feedback observed their mercury rise
ﬂom 9 to 43 during their ﬁrst six trials and ﬂom 43 to 60 during their last six trials. For
velocity feedback, the depicted slopes grew shallow as they descended from 90 to 50 in
the ﬁrst six trials and ﬂom 50 to 10 in the last six trials. Similarly, the acceleration
curves provided to participants reﬂected a rapidly decreasing slope in the ﬁrst six trials
and a slowly decreasing slope in the last six trials. For the baseline trials in both

performance proﬁles, participants were provided with feedback identical to the feedback

39

they would receive at the end of their ﬁrst trial. This was expected to satisfy the
objectives of the baseline trial without biasing participants’ reactions.

Although participants received feedback after every trial, their reactions to
feedback were assessed on three occasions during the experiment to create the three-level
time factor (i.e., baseline, early, and late). Thus, while time is inherent in the trials of the
social intuition task, participants provided their retrospective reactions to three “blocks”
of manipulated trial feedback. State affect, task self-efﬁcacy, self-satisfaction with task
performance, goal commitment, and task effort were measured after the familiarization
trial (i.e., time 1), trial 6 (i.e., time 2) and again after trial 12 (i.e., time 3). To prevent a
maximum performance effort in the later trials, participants were not informed that trial
12 was to be their last trial until after they had completed their time 3 assessments. The
experimenter then debriefed participants (see Appendix I).

Measures

Cognitive ability. Participants were asked to self-report their undergraduate
admissions’ test scores (i.e., ACT or SAT scores) as a proxy measure of cognitive ability.
If access was granted by the participant, then missing, illegible, or out of range scores
were requested ﬂom the university’s admissions’ ofﬁce. Scores were standardized for
later use as a covariate in the analysis (see Table 1 in Appendix A for means, standard
deviations, zero-order correlations, and reliabilites of the individual difference covariates
and reaction measures).

Trait anxiety. The trait anxiety covariate was assessed using 20-items ﬂom
Spielberger’s (1983) State-Trait Anxiety Instrument (STAI; a = .93). As a trait measure,

participants were instructed to describe how they generally feel (i.e., trait), rather than

40

how they feel right now (i.e., state). Participants used a 5-point Likert scale (1 = very
slightly or not at all to 5 = extremely) to respond to items such as “I feel pleasant” and “I
am worried.”

State aﬁ'ect. State affect was measured using the Positive and Negative Affect
Schedule (PANAS; Watson et al., 1988), a 20-item assessment of the positive and
negative affect dimensions. In short, the positive affect subscale captures the extent to
which a respondent feels enthusiastic and alert (i.e., high positive affect) or upset and
lethargic (i.e., low positive affect), while the negative affect subscale reﬂects the presence
(i.e., high negative affect) or absence (i.e., low negative affect) of aversive mood states
such as anger, guilt, fear, or anxiety. As a measure of state affect, participants were
asked to indicate the extent to which they felt at the present moment, for example,
“enthusiastic” or “upset,” using a 5-point Likert scale (1 = very slightly to 5 = extremely).

Subsequent analysis of the negative affect subscale revealed a skewed and
substantially kurtotic distribution that proved resistant to attempts at normalization via
transformation. Participants across all conditions overwhelmingly reported the lack of an
aversive mood state such as anger, fear, anxiety, or guilt. Applying increasingly severe
transformations (i.e., square root, logarithmic, inverse) as outlined in Tabachnick and
Fidell (2001) failed to produce a normal distribution. Therefore, the negative affect
measure was dropped and only the positive affect subscale was retained for analysis.
This scale demonstrated acceptable internal consistency across each combination of
performance proﬁle, feedback ﬂame, and time (i.e., a = 88-96).

Task self-eﬂicacy. Task self-efﬁcacy was assessed by creating a composite of

participants’ ratings of task self-efﬁcacy magnitude and strength for various levels of task

41

performance as described in Lee and Bobko (1994). Participants were asked to report the
degree of conﬁdence (i.e., percent conﬁdence) they have in their ability to achieve
speciﬁc levels of task performance (e.g., correctly respond to 50%,of the SIT items). The
performance levels ranged ﬂom 10% through 100% and include every possible score
(i.e., 20%, 30%, 40%, etc.). Composite task self-efﬁcacy scores were calculated by
summing their non-zero conﬁdence ratings.

Self-satisfaction with performance. Participants’ satisfaction with their task
performance was measured using a 4-item scale created by Toney and Kozlowski (1999).
Using a 5-item Likert scale (1 = strongly disagree to 5 = strongly agree), participants
responded to items such as “I am happy with my performance at this point.” This scale
demonstrated adequate reliability across cells (i.e., .89-.98).

Goal commitment. The goal commitment measure was adapted ﬂom Hollenbeck,
Williams and Klein (1989). Participants responded to this 5-item measure consisting of
items such as “Quite ﬂankly, I don’t care if I achieve this goal or not” using a 5-point
Likert scale (1 = strongly disagree to 5 = strongly agree). Scale alphas across cells
ranged ﬂom .70 to .91 .

Task eﬂort. Effort, as a proxy measure of task motivation, was assessed via an
optional word game. Participants were told that such games can stimulate the language
centers of the brain and that this priming could improve their performance on future
trials. In practice, participants were provided with a word (e. g., backpacker) and asked to
use the letters found in the given word to create as many new words as possible in 2
minutes time (e.g., acre, back, creak). The number of words recorded served as an

objective measure of participants’ task motivation. A “new task effort” measure was also

42

administered to participants at the end of the study to provide an indication of their
motivation to “apply their social intuition” to a new verbal task that was unrelated to the
recently completed trials. This word game involved a 4 x 4 matrix of letters that
participants could use to form words by combining adjacent letters. Once again, the

number of words produced served as a measure of participants’ new task motivation.

43

RESULTS

Table 1 displays the means, standard deviations, zero-order correlations, and
reliabilities of the individual difference covariates and the reaction measures. Note that
cognitive ability was relatively unrelated to the reaction measures except for the new task
effort measure given at time 3. This was unexpected as it was intended to explain some
of the variance in all of the task effort measures. The other covariate, trait anxiety, was
signiﬁcantly correlated with positive affect as anticipated but also goal commitment.
Tables 2a through 2c provide means and standard deviations for the ﬁve motivational
reaction measures over time by performance proﬁle and feedback ﬂame.

Manipulation Check

At the conclusion of the study, participants were asked to rate the degree to which
they agreed with four statements such as “My performance on the task seemed to get
better over time” and “I seemed to do better earlier as compared to more recently” using a
5-point Likert scale (1 = strongly disagree to 5 = strongly agree). Independent samples t-
tests provided evidence that the feedback manipulation was perceived by participants as
intended. Speciﬁcally, participants assigned to the decelerating performance proﬁle
condition reported signiﬁcantly greater agreement with statements describing a
worsening of performance over time as compared to their counterparts assigned to the
accelerating performance proﬁle condition (t(173) = 30.92, p < .001). Conversely, those
assigned to the accelerating performance proﬁle condition reported signiﬁcantly greater

agreement with statements describing an improvement in performance relative to those

participants assigned to the decelerating performance proﬁle condition (t(178) = -3 0.42, p
< .001 ).
Interaction Hypotheses

The presence of the two hypothesized interactions were analyzed simultaneously
using repeated measures multivariate analysis of covariance (MANCOVA) with time as
the within-subjects factor; feedback ﬂame and performance proﬁle as between-subjects
factors; state positive affect, task effort, task self-efﬁcacy, task self-satisfaction with
performance, and goal commitment as the motivational reaction measures of interest; and
trait anxiety as a covariate. To conserve degrees of ﬂeedom, cognitive ability was not
used as a covariate in the overall analyses due to its low and non-signiﬁcant correlations
with each of the reaction measures (see Table 1).

The analytic strategy for testing both hypotheses involved three steps, with each
successive step contingent upon a signiﬁcant ﬁnding in the preceding step: (1) test for
signiﬁcant multivariate within-subj ects interactions, (2) identify the reaction measure or
measures contributing to the interaction via univariate tests, and (3) conduct post hoc
comparisons and prepare mean plots to verify that the pattern of mean differences are
consistent with the interactional hypotheses.

Performance proﬁle x time interaction. To review, the ﬁrst hypothesized
interaction is a two-way interaction between performance proﬁle and time (P x T) such
that participants in the decelerating performance proﬁle condition will report or
demonstrate less motivation over time while those in the accelerating performance proﬁle

condition will report or demonstrate more motivation over time (review Figure 8).

45

The multivariate test provided support for the hypothesized two-way interaction
(F(lO, 644) = 63.75, p < .001; see Table 3). Univariate tests summarized in Table 4
revealed a signiﬁcant P x T interaction for participants’ self-reports of positive affect
(F (2, 326) = 55.09, p < .001), self-efﬁcacy (F (2, 326) = 175.86, p < .001), satisfaction
with performance (F (2, 326) = 287.66, p < .001), and goal commitment (F(2, 326) =
27.70, p < .001), but not participants’ demonstrated task effort (F(2, 326) = 1.18, n.s.).

To identify the nature of these interactions, a series of univariate ANOVAs were
performed for each of the reaction measures identiﬁed. The ﬁrst four analyses included
those participants assigned to the decelerating performance proﬁle condition and grouped
the motivational scores by time. Each omnibus analysis (see Table 5) indicated at least
one signiﬁcant mean difference over time was present for positive affect (F(2, 269) =
11.18, p < .001), self-efﬁcacy (F(2, 267) = 18.51, p < .001), satisfaction with
performance (F (2, 268) = 92.59, p < .001), and goal commitment (F(2, 267) = 12.23, p <
.001). Four additional AN OVAs were then performed for participants assigned to the
accelerating performance proﬁle condition. The omnibus analyses revealed that at least
one signiﬁcant mean difference was present for self-efﬁcacy (F (2, 267) = 43.43, p <
.001) and satisfaction with performance (F(2, 270) = 116.68, p < .001). Additionally, the
analysis for positive affect proved to be marginally signiﬁcant (F(2, 269) = 2.60, p = .08).

Mean plots (see Figures 10a through 10d) revealed that the pattern of participants’
motivational scores for the decelerating performance proﬁle decreased ﬂom time 1 to
time 2 as well as ﬂom time 2 to time 3. Post hoc multiple comparisons revealed
signiﬁcant differences between the time 1 and time 2 means for positive affect,

satisfaction with performance, and goal commitment, and between time 2 and time 3 for

46

affect, self-efﬁcacy, satisfaction, and commitment. In contrast, a review of the pattern of
means for participants assigned to the accelerating performance proﬁle condition
revealed that, with one exception, participants’ motivational scores increased ﬂom time 1
to time 2 as well as ﬂom time 2 to time 3. Multiple comparisons identiﬁed signiﬁcant
mean differences for self-efﬁcacy and satisfaction with performance both ﬂom time 1 to
time 2 and time 2 to time 3.

In sum, participants assigned to both performance proﬁles responded to the range
of feedback ﬂames as expected. Those in the decelerating performance proﬁle reported
signiﬁcantly more positive affect, self-efﬁcacy, satisfaction with performance, and goal
commitment over time, while those receiving feedback consistent with an accelerating
performance proﬁle generally reported less positive motivation over time, especially self-
efﬁcacy and satisfaction. Taken as a whole, these results provide substantial support for
the hypothesized P x T interaction.

Feedback frame x performance proﬁle x time interaction. The second
hypothesized interaction is a three-way between feedback ﬂame, performance proﬁle,
and time (F x P x T) such that the pattern of results seen in the P x T interaction would
not hold for participants receiving position feedback (review Figures 9a through 9c).
Speciﬁcally, participants were expected to report or demonstrate more motivation over
time, regardless of performance proﬁle condition, because position discrepancies are
always greatest at the beginning of a curve, but smaller towards the ends. Thus,
motivation should increase over time as position discrepancies decrease even at two

differing rates as represented in the two performance proﬁles.

47

As shown in Table 3, the multivariate analysis provided support for the presence
of the three-way interaction (F(20, 1069) = 1.64, p < .05). However, the univariate tests
displayed in Table 4 revealed a signiﬁcant F x P x T interaction for participants’ reports
of self-efﬁcacy only (F (4, 326) = 6.16, p < .001). Once again, a series of ANOVAs were
performed for participants assigned to each feedback ﬂame and performance proﬁle
combination with self-efﬁcacy scores as the dependent variable. A review of Table 6
indicates the presence of at least one signiﬁcant mean difference in self-efﬁcacy scores
across time periods for each of feedback ﬂame by performance proﬁle combinations
except one. Speciﬁcally, signiﬁcant differences in self-efﬁcacy scores were identiﬁed for
position feedback provided in the accelerating proﬁle condition (F(2, 86) = 6.08, p < .01),
velocity feedback provided in both the accelerating and decelerating proﬁle conditions
(F(2, 89) = 21.29 and 19.59, p < .001, respectively), and acceleration feedback provided
in both the accelerating and decelerating proﬁle conditions (F (2, 86) = 25.77 and 3.63, p
< .001 and p < .05, respectively). The test for a signiﬁcant mean difference for position
feedback provided in the decelerating performance proﬁle condition received only
marginal support (F(2, 86) = 2.88, p = .06).

To determine if these mean differences were in the expected directions, the self-
efﬁcacy means were plotted and subjected to post hoc comparisons. As displayed in
Figures 11b and 11c, the pattern of means for participants receiving velocity and
acceleration feedback is largely consistent with the results of the P x T interaction; they
depict increasing scores over time in the accelerating proﬁle and decreasing scores over
time in the decelerating proﬁle. However, the pattern for participants receiving position

feedback was not as predicted. Rather than showing increasing scores over time for both

48

the accelerating and decelerating proﬁles, the position feedback plot (Figure 11a) reveals
a pattern of means that is similar to, but less robust than, that found in the velocity and
acceleration plots. Apparently, participants in the decelerating performance condition
reported sigriﬁcantly less task efﬁcacy at time 3 relative to time 2, rather than
sigriﬁcantly more. Therefore, despite the presence of a signiﬁcant interaction, the F x P
x T interaction was not supported as hypothesized.
Additional Analyses

As previously mentioned, participants were offered the opportunity to “apply their
social intuition” by engaging in an optional word game at the end of the study but prior to
debrieﬁng. Unlike the three preceding word games, this was not a “word generation”
task but a “word ﬁnd” task (see Appendix J). The purpose of the task was to provide an
index of participants’ motivation to use their “newly-developed” social intuition now that
the trials were completed. In turn, this permitted a test of the effects of performance
proﬁle and feedback ﬂame on a measure of participant motivation (i.e., the new task
effort score) that was psychologically removed ﬂom the trials. These effects were tested
using an ANOVA with performance proﬁle and feedback ﬂame as factors, new task
effort scores as the dependent variable, and cogritive ability as a covariate.
Unfortunately, neither feedback ﬂame, performance proﬁle, nor their interaction had a
signiﬁcant effect on participants’ new task effort scores.

This study was never intended to compare the feedback ﬂames for the purposes of
identifying the single “best” type of feedback. In fact, it is unlikely that any one feedback
ﬂame would be optimal under all conditions. However, this data set does permit a purely

exploratory examination of which feedback ﬂames produced the highest motivation

49

scores for participants assigred to each of the performance proﬁles. To examine these
effects, a series of AN OVAs were conducted for the reaction measures with feedback
ﬂame as the gouping variable (see Table 7). In the decelerating performance condition,
participants who received position feedback reported sigriﬁcantly higher positive affect
and goal commitment than did their counterparts who received either velocity or
acceleration feedback. By contrast, participants in the accelerating proﬁle who received
either velocity or acceleration feedback reported higher task self-efﬁcacy and goal
commitment scores than those who received position feedback. Furthermore, those who
received acceleration feedback reported geater positive affect and satisfaction with
performance scores than those who received position feedback. Apparently, participants
experiencing declining performance reported geater motivation when they received
position feedback, while those enjoying improving performance reported geater
motivation when they received acceleration or velocity (i.e., non-position) feedback.
While these results are far ﬂom deﬁnitive in identifying an optimal feedback ﬂame, they
do raise some interesting ideas for future research which will be discussed in geater

detail in the following section.

50

DISCUSSION

The purpose of this study was to (1) conceptually and operationally deﬁne
position, velocity, and acceleration feedback for use as performance information during
learning or skill acquisition, and (2) examine individuals’ affective, behavioral, and
cogritive reactions to performance feedback presented in position-, velocity-, and
acceleration-salient ﬂames in a goal-directed, self-regulatory context. The initial
proponents of velocity and acceleration information as feedback had been inconsistent in
their deﬁnitions and applications of these ﬂames, leaving some doubt as to the distinction
between them as well as their demonstrated effects on individuals’ performance. This
study sought to remove these doubts by drawing ﬂom the original metaphor ﬂom the
physical sciences (i.e., the laws of motion) to explicitly deﬁne each feedback ﬂame and
draw attention to their characteristic components for improved application. These
clariﬁed ﬂames could then be studied to examine their effects on individuals’ motivation
as they pursue a task goal.

The ﬁrst hypothesis, a P x T interaction, proposed that, regardless of feedback
ﬂame, participants assigred to a decelerating performance proﬁle condition would report
or demonstrate less motivation over time, while those assigred to an accelerating
performance proﬁle condition would report or demonstrate geater motivation over time.
This interaction was supported for four of the ﬁve motivational measures used in the
study: state positive affect, task self-efﬁcacy, self-satisfaction with task performance,

and goal commitment.

51

 

Simple effects analyses for positive affect revealed that participants in the
decelerating performance proﬁle reported sigriﬁcantly less positive affect at time 2
relative to time 1, and at time 3 relative to time 2. Apparently, receiving position,
velocity, or acceleration information that described a worsening of performance served to
reduce their enthusiasm while inspiring lethargy. Although participants in the
accelerating performance proﬁle reported more positive affect at time 2 relative to time 1,
as well as at time 3 relative to time 2, the mean differences were not sigriﬁcant. Still,
this trend is consistent with Lawrence et al. (2002) who reported that their participants
expressed more positive moods after experiencing more positive velocities, even if their
velocity manipulation did have acceleration components. The lack of sigriﬁcant mean
differences in the accelerating performance proﬁle may be due to the choice of task,
which was both ambiguous and repetitive. Speciﬁcally, the task may have engendered
little enthusiasm among participants as evidenced by a lack of skew in the positive affect
measure indicating neither enthusiasm nor lethargy, as well as the severe skew in the
defunct negative affect measure which indicates the near absence of anxiety or guilt.

Post hoc analyses of the self-efﬁcacy measure also revealed that participants in
the decelerating performance proﬁle reported less task efﬁcacy at time 2 relative to time
1, and sigriﬁcantly less efﬁcacy at time 3 relative to time 2. The lack of a sigriﬁcant
mean difference in the early period may reﬂect participants’ skepticism of the highly
favorable feedback in the ﬁrst few trials. Speciﬁcally, participants in the decelerating
performance proﬁle condition received a 90% on a test of “forei gr” words after both the
familiarization trial and trial 1. Given the ambiguity of the task, some may have rejected

the high scores as an indication of their actual level of social intuition, and reported less

52

task efﬁcacy in response. This view is supported by experimenter observations which
noted some disbelief among participants when they received their feedback. On the other
hand, participants assigned to the accelerating performance proﬁle reported sigriﬁcantly
more task efﬁcacy at time 2 relative to time 1, and at time 3 relative to time 2. This result
is consistent with the theory and literature that concludes self-efﬁcacy is strengthened by
positive performance feedback (Martocchio & Webster, 1992; Wood & Bandura, 1999).
The presence of a sigriﬁcant mean difference in the early period of this condition may
reﬂect the fact that participants were more likely to believe their initial poor performance
(i.e., they received a 10% on the familiarization trial and trial 1) as the experimenter
observed that some participants responded to their feedback with self-deprecating humor
or comments that they had “expected” to do that poorly.

Planned comparisons of participants’ self-satisfaction with performance scores
indicated that participants in the decelerating performance proﬁle reported sigriﬁcantly
less satisfaction at time 2 relative to time 1, and sigriﬁcantly less satisfaction at time 3
relative to time 2. Conversely, those in the accelerating proﬁle reported sigriﬁcantly
more satisfaction at time 2 relative to time 1 as well as at time 3 relative to time 2. These
results are consistent with Hsee and Abelson (1991) and Hsee et al. (1994) who
concluded that people were more satisﬁed with more positive velocity outcomes and
more positive accelerations, respectively, as well as with Johnson and Lord (2005) who
found that participants’ velocities were positively related to their reports of task
satisfaction.

Finally, the results of the post hoc analyses of the goal commitment measure were

similar to that of the positive affect measure. Commitment in the decelerating

53

 

performance proﬁle decreased sigriﬁcantly ﬂom time 2 relative to time 1 and ﬂom time
3 relative to time 2. However, in the accelerating proﬁle, there were no sigriﬁcant mean
differences as scores remaining relatively unchanged over time. Once again, the nature
of the task may help to explain the observed results as an uninspiring task may not
generate increases in goal commitment, but may be more susceptible to decreases in
commitment in the face of unfavorable feedback.

It is interesting to note that the observed P x T interaction did not hold for the task
effort measure. It is possible that the word game did not measure participants’ task
motivation, but instead reﬂected a desire to do “something” as a diversion ﬂom the study.
Recall that the word game was an optional activity. However, those who chose not to
participate in the game were instructed to sit quietly for the duration of the game. In
practice, few could stand to do nothing for two minutes. Perhaps the effort measure did
not measure task motivation, but the motivation to escape the task regardless of condition
or time.

Overall, the results of the P x T interaction analysis suggests that the set of
position, velocity, and acceleration feedback ﬂames, as conceptually and operationally
deﬁned in this paper, operate in a predictable manner consistent with theory and past
research. Not only did the observed results for the range of motivational measures
support past research which applied more ambiguous conceptions of velocity and
acceleration, but the use of contrasting performance proﬁles provided additional evidence
that the feedback ﬂames affected participants’ motivation whether their task performance

was improving or worsening.

54

 

The hypothesized F x P x T interaction proposed that the observed P x T
interaction pattern would hold for the velocity and acceleration feedback ﬂames, but not
the position ﬂame. Speciﬁcally, an understanding of the referents for each ﬂame
suggests that, for position feedback, discrepancies will always be geater in the early
period of either performance proﬁle while, for velocity or acceleration feedback, they can
be either more or less in the early period of an accelerating or decelerating performance
proﬁle, respectively. Despite the presence of a signiﬁcant three-way interaction for task
efﬁcacy, the simple effects analysis did not support this pattern of results. Apparently,
participants responded to position feedback in much the same way they responded to
velocity and acceleration feedback. Speciﬁcally, participants in the accelerating proﬁle
reported increasing task self-efﬁcacy over time while those in the decelerating proﬁle
reported decreasing efﬁcacy over time.

Carver and Scheier (1990; 1998) proposed that, in the action loop, position
feedback is compared to the referent standard to determine if a discrepancy is present,
and the system takes actions to reduce sensed discrepancies. Why, then, did participants
receiving position feedback not sense and respond appropriately to geater positional
discrepancies early in both performance proﬁle conditions as shown in Figure 9a? It is
possible that they did not interpret position feedback in an action loop consistent with
Control Theory. Despite attempts to provide only position-salient information in the
position feedback condition, participants may have converted their position feedback into
a more popular or useful “percent correct” score (i.e., velocity feedback). For example,
during data collection, at least eight participants were observed recording their position

feedback scores on their desktops or study packets. They used this record to subtract

55

 

their preceding position ﬂom their current position to determine the number of items they
answered correctly in the last trial. Since each trial consisted of 10 items, this number
could then be multiplied by 10 to obtain the percentage of items they responded to
correctly (i.e., the operational equivalent of velocity feedback). In the meta loop, velocity
feedback is compared to some “acceptable rate” of discrepancy reduction, and the system
produces affect in response to sensed rate discrepancies (Carver & Scheier, 1990; 1998).
Therefore, if participants routinely used simple math to convert their position feedback
into velocity feedback, then the observed results are not unexpected. That is, the
observed pattern of efﬁcacy means for position feedback is consistent with that of
participants responding to velocity feedback (see Figure 9b).
Implications

Despite the complex distinctions between the position, velocity, and acceleration
feedback ﬂames, participants were able to comprehend their respective feedback ﬂames
and respond appropriately in a goal-directed, self-regulatory context. When they
received feedback intended to be favorable, they responded with increased motivation;
when they received feedback intended to be judged unfavorably, they reported or
demonstrated decreased motivation. Furthermore, their responses were generally
congruent with their location, with respect to time, within their assigred performance
proﬁles. That is, motivation improved with favorable feedback in the early portion of a
decelerating proﬁle as well as in the late portion of an accelerating proﬁle. Taken as a
whole, this suggests that the proposed conceptualizations and operationalizations of the
three feedback ﬂames are acceptable. Not only are they consistent with the results of

earlier research on satisfaction with outcomes, but they also support the results of the two

56

 

studies of velocity feedback as applied in a goal-directed context. Therefore, researchers
who wish to study self-regulated learning or feedback ﬂames can use the guidance
offered here to formulate and deliver their feedback.

Practitioners may be wondering which type of feedback to use for different tasks.
This study found that while each ﬂame was interpretable by participants, it was clear that
some participants would have preferred to receive their feedback in one ﬂame rather than
another. Speciﬁcally, several participants assigned to the position feedback condition
went to geat lengths to track their position feedback for conversion into velocity
feedback. This observation suggests that, given the task, some participants found
velocity feedback to be easier to understand or more useful for self-regulation. While
this study necessitated the application of each feedback ﬂame, practitioners should
consider the nature of the task or goal when selecting the type of feedback to provide.
For example, if an individual’s goal is phrased in terms of achieving a numerical standard
(e. g., learn 25 ways to sigral a rescue plane), then position feedback may be easier to use
for self-regulation. If the task does not suggest a feedback ﬂame, perhaps individuals can
be given the option of receiving their feedback ﬂamed in terms of position, velocity, or
acceleration. For maximum ﬂexibility, the optimal solution may be to provide multiple
ﬂames as a means of communicating all aspects of an individual’s goal progess.
Limitations & Future Research

This study employed the Social Intuition Task developed by Lawrence et al.,
(2002) for use in studying velocity feedback in a goal-directed context. Like Knowlton
and Squire’s (1996) Implicit Grammar Task, the SIT is desigred to be ambiguous so that

participants will accept and respond to the manipulated feedback provided to them by the

57

experimenter. However, it is important to recogrize that participants are neither
performing nor learning in this experiment, even if they truly believe they are improving
their social intuition. Therefore, we cannot be certain that individuals engaged in a real
world task involving desirable goals and actual performance would respond to these
feedback ﬂames in the same way. Unfortunately, the limitations of the ambiguous task
paradign must be accepted if we are to study participants’ reactions to the two
contrasting performance proﬁles. Future research should study feedback ﬂames in the
context of a realistic task involving actual performance and desirable goals. It is likely
that the thrill of actually acquiring geater knowledge or skill would produce a geater
motivational response than that observed in an ambiguous task in which participants were
merely told that their performance was improving. Conversely, the real world experience
of slowing knowledge or skill acquisition would likely produce a geater demotivating
response as compared to that observed in this study.

The new task motivation measure also proved to be a limitation of the study as it
may have been a measure of cognitive ability rather than task effort or motivation. The
task effort word game measures were selected because they represented behaviorally-
based, objective measures of task motivation. In retrospect, it would not have been
difﬁcult to add a subjective assessment of task effort. Future studies using the ambiguous
task paradign should continue to incorporate a broad range of affective, behavioral, and
cogritive reaction measures, including a self-report measure of task effort and
motivational states (e.g., state goal orientation; Elliot & Church, 1997; VandeWalle, Cron

& Slocum, 1999).

58

Another limitation of this study concerns the operationalization of the feedback
ﬂames. Past research did not carefully review the laws of motion metaphor before
deﬁning and applying position, velocity, and acceleration feedback. As a result, it was
possible to make the mistake of, for example, providing velocity feedback with position
or acceleration components. This study represents an attempt to use the metaphor to
provide guidelines for applying the feedback ﬂames as independently as possible, for the
sake of exploring their unique effects on task motivation. However, this proved to be
easier said than done. Conceptually, position, velocity, and acceleration are related; any
physics text provides a series of equations that reveal how any one can be derived ﬂom
any of the others. In retrospect, this study’s attempt to provide “pure” position, velocity,
and acceleration feedback to participants was not ﬁrlly successﬁrl. Speciﬁcally, the
acceleration feedback contained velocity-salient information. While the feedback curves
removed positional references ﬂom their axes, the accompanying verbiage referred to a
velocity-salient current score (e.g., 10% in Figure 5) even as it emphasized an
acceleration-salient increase in performance (e. g., 10% more than last trial). This may
explain why the results of analyses involving velocity and acceleration looked similar.
Future research on feedback ﬂames should pursue both an isolation and an integation
track. The purpose of the isolation track is to conﬁrm the results of this study using pure
position, velocity, and acceleration feedback. As a start, this can be done by removing
the velocity-salient information ﬂom the acceleration feedback and by using a prime
number of items in a trial to make it more difﬁcult for participants to convert position to
velocity information. The integation track addresses the idea that it may not be possible

or even advisable to completely separate position-, velocity-, and acceleration-salient

59

information. In practice, velocity and acceleration information are often provided
concurrently (e. g., the position and rate of movement of a car’s speedometer needle) and
participants have demonstrated a willingress to go to geat lengths to calculate velocity
ﬂom position information. Such research might determine that particular combinations
of feedback ﬂames (i.e., position with velocity or velocity with acceleration) are more
effective than any one ﬂame.

One question that this study was not intended to address is: which feedback
ﬂame works best? The additional analyses introduced in the results section seem to
indicate that individuals may realize geater motivation gains if they receive velocity or
acceleration feedback when their performance follows an accelerating performance
proﬁle. This may be due to the fact that both ﬂames provide rate or trend information
that may lead individuals to believe that their rosy performance is destined to continue.
However, when individuals’ performance follows a decelerating proﬁle, unfavorable
rates and trends suggest that performance will only get worse. Perhaps individuals are
less likely to see this future when position feedback is supplied. In any case, these
speculations raise a number of empirical questions that can be investigated in a controlled
setting.

Conclusions

In the 15 years since Carver and Scheier (1990) ﬁrst introduced the concept of
velocity feedback, little progess has been made in the attempt to understand its
relationship to position and acceleration feedback, or how it functions in the self-
regulatory cycles associated with learning and skill-acquisition. However, the tide is

turning. Within the last three years, Lawrence et al. (2002) published the ﬁrst study of

60

velc

join

reﬁ
fee
reft

511;

{CS

velocity feedback in a self-regulatory context and Johnson and Lord (2005) studied the
joint effects of position and velocity feedback on individuals performing an ambiguous
task. This study has contributed to this ongoing effort to understand feedback ﬂames by
reﬁning the conceptual and operational deﬁnitions of position, velocity, and acceleration
feedback and examining individuals’ application of this feedback in the context of their
referents and Control Theory. While the hypothesized F x P x T interaction was not
supported, the proposed feedback ﬂames are instructive to those who wish to continue
this research as students and trainees stand to beneﬁt ﬂom their application once

researchers identify how and when they operate effectively.

61

APPENDICES

62

APPENDIX A

Tables

63

Table 1. Means, standard deviations, zero-order correlations, and reliabilities of the
individual difference and reaction measures.

 

 

Measures M SD 1 2 3 4 5

1. Standardized Cognitive Ability 0.00 1.00 --

2. Trait Anxiety 43.41 13.25 .02 (.93)

3. State Positive Affect (Time 1) 27.10 8.84 -.08 433: (.92)‘

4. State Positive Affect (Time 2) 27.04 9.11 -.09 -.29 .80 (.92)

5. State Positive Affect (Time 3) 25.68 10.74 -.11 -.28” .61“ .87" (.95)
6. Task Effort (Time 1) 11.86 4.83 .14 -.09 .10 .14 .16‘
7. Task Effort (Time 2) 11.68 4.34 .12 -.02 .05 .13 .13
8. Task Effort (Time 3) 12.28 4.66 .10 .02 -.05 .05 .05
9. Task Self-Efﬁcacy (Time 1) 469.64 238.07 -.06 -.09 .23“ .09 -.06
10. Task Self-Efﬁcacy (Time 2) 510.13 209.13 -.08 -.02 .18' .09 .05
11. Task Self-Efﬁcacy (Time 3) 523.52 236.84 .01 -.13 .10 .27" .38“
12. Performance Satisfaction (Time 1) 12.13 6.15 .02 -.16' .27“ -.00 -.21”
13. Performance Satisfaction (Time 2) 12.87 4.04 -.02 -.12 .05 .19‘ .21”
14. Performance Satisfaction (Time 3) 12.46 5.25 -.01 -.06 -.09 .18‘ .38“
15. Goal Commitment (Time 1) 17.82 3.83 -.03 -24“ .33“ .29“ .27"
16. Goal Commitment (Time 2) 17.19 4.07 -.04 -.26“ .32“ .41“ .42“
17. Goal Commitment (Time 3) 16.60 4.68 -.10 -.23” .20“ .41” .54”
18. New Task Effort 9.66 4.52 .27” .00 .03 .06 13

 

Note: Reliabilities on the diagonal. 'p < .05; "p < .01

64

Table I (cont’d).

 

 

Measures 6 7 8 9 10 ll 12
l. Standardized Cognitive Ability

2. Trait Anxiety

3. State Positive Affect (Time 1)

4. State Positive Affect (Time 2)

5. State Positive Affect (Time 3)

6. Task Effort (Time 1) --

7. Task Effort (Time 2) .63” --

8. Task Effort (Time 3) .56” .58" --

9. Task Self-Efﬁcacy (Time 1) .01 .00 .02 ~-

10. Task Self-Efﬁcacy (Time 2) .01 .03 .03 .79' --

11. Task Self-Efﬁcacy (Time 3) .16' .11 .10 .18' .44" --

12. Performance Satisfaction (Time 1) -.16' -.10 -.14 .52” .34” -.27” (.98)
13. Performance Satisfaction (Time 2) .08 .05 .05 -.14 .07 .28" -.02
14. Performance Satisfaction (Time 3) .14 .10 .07 -.42” -.19‘ .46“ -52”
15. Goal Commitment (Time 1) -.04 -.02 -.08 .15‘ .17‘ .13 .15‘
16. Goal Commitment (Time 2) -.02 -.01 -.08 .08 .14 .25“ .06
17. Goal Commitment (Time 3) .01 -.00 -.02 -.05 .08 .36" -.15‘
18. New Task Effort .38“ .37“ .36" -.05 .03 .12 -.06

 

Note: Reliabilities on the diagonal. 'p < .05; "p < .01

65

Table I (cont’d).

 

Measures 1 3 14 l 5 l 6 l 7 l 8
Standardized Cognitive Ability

Trait Anxiety

State Positive Affect (Time 1)

State Positive Affect (Time 2)

State Positive Affect (Time 3)

Task Effort (Time 1)

Task Effort (Time 2)

Task Effort (Time 3)

. Task Self-Efﬁcacy (Time 1)

10. Task Self-Efﬁcacy (Time 2)

ll. Task Self-Efﬁcacy (Time 3)

12. Performance Satisfaction (Time 1)

13. Performance Satisfaction (Time 2) (.94).

14. Performance Satisfaction (Time 3) .61 (.98)

‘O?°>l.o‘$":ﬂl~”!°:"

15. Goal Commitment (Time 1) -.08 -.05. (.80).

16. Goal Commitment (Time 2) .07 .15” .76” (.82.).

17. Goal Commitment (Time 3) .13 .33 .61 .75 (.86)

18. New Task Effort .18“ .17“ -.05 .03 .03 --

 

Note: Reliabilities on the diagonal. 'p < .05; "p < .01

66

Table 2a. Descriptive statistics for reaction measures by time and performance proﬁle
for participants receiving position feedback.

 

Position Feedback
Time 1 Time 2 Time 3 Marginal Means
Reaction Measure M SD M SD M SD M SD

 

Decelerating Performance Proﬁle

Positive Affect 30.44 8.10 30.11 8.52 25.48 10.34 28.21 9.13
Task Effort 9.56 5.38 10.70 4.76 10.96 5.75 10.45 5.23

Task Self-Efﬁcacy 592.22 237.36 564.33 210.37 466.81 182.86 550.24 216.12
Satisfaction 17.07 4.15 12.56 4.11 8.59 3.83 12.72 5.12

Goal Commitment 19.56 3.72 18.15 4.08 16.70 5.21 17.90 4.50

Accelerating Performance Proﬁle

Positive Affect 24.61 8.73 26.82 8.35 26.11 10.51 25.17 9.24
Task Eﬂ‘ort 13.79 3.63 13.57 3.26 14.75 4.45 13.69 3.98

Task Self-Efﬁcacy 319.54 185.93 339.36 187.15 475.61 198.22 378.78 195.80
Satisfaction 7.25 4.35 12.82 4.04 14.29 3.40 1 1.30 4.97

Goal Commitment 17.25 4.49 16.50 4.66 16.46 4.32 16.83 4.33

67

Table 2b. Descriptive statistics for reaction measures by time and performance proﬁle
for participants receiving velocity feedback.

Velocity Feedback

 

Time 1 Time 2 Time 3 Marginal Means
Reaction Measure M SD M SD M SD M SD

Decelcrating Performance Proﬁle

Positive Affect 28.97 7.85 24.63 7.02 21.23 6.87 24.82 7.77
Task Eﬂon 10.97 5.15 11.47 4.24 11.60 5.29 11.39 4.80

Task Self-Efﬁcacy 621.83 172.72 630.77 166.13 400.83 153.13 551.17 192.80
Satisfaction 17.23 3.00 11.40 2.77 8.00 3.30 12.00 4.94

Goal Commitment 17.80 3.09 16.73 3.46 14.50 3.24 16.30 3.47

Agelerating Performance Proﬁle

Positive Affect 25.79 8.57 28.68 9.94 30.57 11.84 27.48 10.47
Task Effort 11.68 4.13 11.50 3.99 12.39 3.98 11.94 4.00

Task Self-Efﬁcacy 310.96 169.47 438.50 157.75 683.14 264.58 468.32 246.20
Satisfaction 7.86 4.33 13.61 4.39 16.57 3.13 12.43 5.22

Goal Commitment 17.61 4.72 18.18 4.91 18.96 5.01 18.20 4.78

68

Table 2c. Descriptive statistics for reaction measures by time and performance proﬁle for
participants receiving acceleration feedback.

 

Acceleration Feedback
Time 1 Time 2 Time 3 Marginal Means
Reaction Measure M SD M SD M SD M SD

 

Decelcrating Performance Proﬁle

Positive Affect 26.93 9.87 24.10 9.66 21.31 9.66 23.82 9.86
Task Effort 11.62 4.81 10.76 5.12 11.59 4.14 11.40 4.62

Task Self-Efﬁcacy 577.10 240.19 568.38 236.33 431.93 257.32 528.60 251.32
Satisfaction 16.76 4.19 12.34 4.60 10.14 4.92 13.00 5.34

Goal Commitment 16.97 3.62 15.90 3.80 14.90 4.68 15.86 4.06

Accelerating Performance Proﬁle

Positive Affect 27.89 9.32 31.04 8.92 32.14 11.28 30.18 9.67
Task Eﬂ‘ort 13.29 4.69 1 1.61 4.08 12.54 3.73 12.78 4.34

Task Self-Efﬁcacy 385.75 194.13 492.71 167.01 703.04 180.74 528.45 222.60
Satisfaction 7.50 3.66 15.21 3.59 17.89 3.00 13.48 5.57

Goal Commitment 17.79 2.71 18.07 3.21 19.04 3.96 18.14 3.61

 

69

Table 3. Results of the repeated measures multivariate analysis of covariance.

 

 

Between Subjects df F r12 p
Intercept 5, 159 196.23'" .86 .000
Performance Proﬁle (P) 5, 159 6.12‘” .16 .000
Feedback Frame (F) 5, 318 1.13‘ .34 .034
P x F 10, 318 1.98 .06 .059
Trait Anxiety (A) 5, 159 5.85 .16 .155
Within Subjects

Time (T) 10, 644 1.54 .02 .120
T x P 10, 644 63.75'” .50 .000
T x F 20, 1069 1.74‘ .03 .023
TxA 10,644 1.40 .02 .191
T x P x F 20, 1069 1.64‘ .03 .032

 

Note: F-test of Wilks’ Lambda; 'p < .05; mp < .001

70

Table 4. Results of the repeated measures univariate analyses of covariance.

 

 

Positive Affect Task Effort Task Self-Efﬁcacy

Between Subjects df F r72 (if F r12 df F r12
Intercept 1 273.18'" .63 1 156.39'" .49 1 134.9?" .45
Performance Proﬁle (P) 1 4.20‘ .03 1 9.40" .05 1 8.00" .47
Feedback Frame (F) 2 0.06 .00 2 0.35 .00 2 2.18 .03
P x F 2 3.17‘ .04 2 2.93 .04 2 2.75 .03
Trait Anxiety (A) 1 19.65‘” .11 1 1.01 .01 1 1.41 .01
Error 163 (198.92) 163 (43.62) 163 (93330.26)
Within Subjects

Time (T) 2 0.07 .00 2 0.96 .01 2 4.50' .03
T x P 2 55.09‘“ .25 2 1.18 .01 2 175.86‘“ .52
T x F 4 1.11 .01 4 1.47 .02 4 2.04 .02
T x A 2 0.62 .00 2 1.38 .01 2 2.79 .02
T x P x F 4 1.92 .02 4 0.20 .00 4 6.16‘“ .07
Error 326 (17.93) 326 (8.89) 326 (13419.32)

 

Note: Mean squared error in parentheses. 'p < .05; p < .01

71

; p<.001

Table 4 (cont’d).

 

 

 

Satisfaction with
Performance Goal Commitment

Between Subjects df F r12 df F r12
Intercept 1 378.20'" .70 1 458.14'" .74
Performance Proﬁle (P) l 0.02 .00 1 3.78 .23
Feedback Frame (F) 2 2.42 .03 2 0.32 .00
P x F 2 1.20 .01 2 3.09‘ .04
Trait Anxiety (A) 1 6.39" .04 1 14.10’” .08
Error 163 (22.51) 163 (38.613)
Within SubLects
Time (T) 2 0.30 .00 2 0.87 .01
T x P 2 287.66‘” .64 2 27.70‘“ .15
T x F 4 3.00‘ .04 4 1.86 .02
T x A 2 0.62 .00 2 0.48 .00
T x P x F 4 0.82 .01 4 1.32 .02
Error 326 (10.76) 326 (4.40)
Note: Mean squared error in parentheses. 'p < .05; p < .01"° mp < .001

72

Table 5. Results of the Simple effects univariate analyses for the performance proﬁle x

 

time interaction.

Dependent Variable 3’] F r12
Decelcrating Performance Proﬁle

State Positive Affect 2, 269 11.18‘: .08

Task Self-Efﬁcacy 2, 267 18.51' .12

Satisfaction with Performance 2, 268 92.59‘” .41

Goal Commitment 2, 267 12.23‘” .08

Accelerating Performance Proﬁle

State Positive Affect 2, 269 260* .02
Task Self-Efﬁcacy 2, 267 43.43’” .25
Satisfaction with Performance 2, 270 116.68'” .46
Goal Commitment 2, 269 0.190 .00

 

T'p < .001; *p< .10

73

 

Table 6. Results of the feedback ﬂame x performance proﬁle x time interaction’s simple
effects univariate analyses for the task self-efﬁcacy measure.

 

 

Performance Proﬁle df F n2
Roam—99M
Accelerating 2,86 6.08“ .12
Decelcrating 2, 86 288* .06
Velocitv F eedbac_k
Accelerating 2,89 2129‘” .32
Decelcrating 2, 89 19.59‘” .31

Acceleration Feedback
Accelerating 2, 86 25.77'“ .38

Decelcrating 2, 86 3.63‘ .08
T}: < .05; "p < .01; "'p < .001; 1p < .10

74

Table 7. Feedback ﬂames that produced Sigliﬁcantly higher motivational scores by
performance proﬁle.

 

Reaction Measure
Task Satisfaction with Goal
Performance Proﬁle Positive Affect Self-Efﬁcacy Performance Commitment
Decelcrating P > (V and A) -- -- P > (V and A)
Acceleration (V and A) > P A > P A > P (V and A) > P

 

P = position feedback; V = velocity feedback; A = acceleration feedback

75

APPENDD( B

Figures

76

Figure 1. A nominal meta loop with an embedded action loop.

 

Goal:
All “acceptable rate” of
discrepancy reduction

 

 

 

 

 

 

 

V

 

y

 

Input:
Rate of discrepancy
reduction

 

 

 

 

 

V

 

 

Output:
Positive or negative
affect

 

 

 

 

Environment:
Discrepancy reduction in the
action loop

 

 

A

 

 

 

77

 

 

Figure 2. Lawrence et al.’s (2002) velocity feedback conditions.

 

    

 

 

— High Positive Velocity

.5 9° "1 Medium High Positive Velocity
2 80 .1 Intermediate Positive Velocity
m """""" Medium Low Positive Velocity
g 70 —- - - Low Posrtlve Velocntv
11.1
g 60 —
§ 50 _i O... .0 . ’
E 0.... O... ’
S.) 40— c ....................... ’ ’
3 3o _ ...................... ’ ’
.‘P. I
t a
31’ 10 — ’

I l I I l I

1 2 3 4 5 6

Trial Block

Adapted ﬂom “Velocity toward goal attainment in immediate experience as a
determinant of affect,” by J. W. Lawrence, C. S. Carver, and M. F. Scheier, 2002,
Journal of Applied Social Psychology, 32, p. 794.

78

Figure 3. An example of position-salient feedback.

100;:
930—2

80 _:
Thus far, you have 70::
correctly responded so;
to a total of g items. —;

Trial 1

79

 

Figure 4. An example of velocity-salient feedback.

 

You have correctly
responded to 90% of
the items in this trial.

 

 

 

Trial 1

80

O

dwU'INCO
COCO

Figure 5. An example of acceleration-salient feedback.

A

You have correctly

responded to 10% of
the items in this trial;
this is 10% less than \

the previous trial. ’
Trial 12

 

81

Figure 6. A decelerating learning curve characterized by rapid early performance gains
that slowly diminish over time.

Performance
(# Correct Responses)

b----------

 

 

V

Early Late

82

Figure 7. An accelerating learning curve characterized by slow early performance gains
that rapidly increase over time.

\

Performance
(# Correct Responses)

 

 

 

Early Late

83

Figure 8. Hypothesized performance proﬁle x time interaction.

 

 

m Accelerating
4E» Proﬁle

.8.

3

O

E

C

o

'33

9

'43 Decelerating
2 Proﬁle

Baseline Early Late

84

Figure 9a. Hypothesized feedback ﬂame x performance proﬁle x time interaction for
position feedback.

For Position Feedback:

Accelerating
Proﬁle

Decelerating
Proﬁle

Motivational Outcomes

 

 

v7

Baseline Early Late

85

Figure 9b. Hypothesized feedback ﬂame x performance proﬁle x time interaction for

velocity feedback.

Motivational Outcomes

 

 

For Velocity Feedback:
Accelerating
Proﬁle
Decelcrating
Proﬁle
Baseline Early Late

86

Figure 9c. Hypothesized feedback ﬂame x performance proﬁle x time interaction for
acceleration feedback.

For Acceleration Feedback:

 

 

a, Accelerating
‘3 Proﬁle

.8

3

O

a

C

O

'5

9

*3 Decelcrating
2 Proﬁle

Baseline Early Late

87

Figure 100. Performance proﬁle x time interaction for state positive affect.

 

 

 

 

 

 

 

 

+Accelerating
Performance Proﬁle

+ Decelerating
Performance Proﬁle

 

 

 

 

 

 

50
2 4o-——--———e -
< . . .
2 30 2840 * 28‘15 28¢ 86
a
m 20
3 22.51
a
10
0
1 2 3
Time

 

’p< .05; “p < .01

88

 

 

Figure 10b. Performance proﬁle x time interaction for task self-efﬁcacy.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1000 l

900

800
g 700 61221 n.“
E 600 60239 595 '14 A! +Accelerating
In 500 ”* Performance Proﬁle
h ———4
'5 / a...\. +Deceletating
2 400 ,/...r 424.17 434.83 “—1 Pem'mance Pmﬁ'e
g 300

340.42
200
100 — .__
0
1 2 3
Time

 

 

 

”p < .01; ”‘p < .001

89

Figure 10c. Performance proﬁle x time interaction for self-satisfaction with

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

performance.

l 20
g 16 82
" ' 16.05
E 15 *tt tilt
g 13 6
“- +Accelemting
g 10 "I. 12.06 I.“ Performance Proﬁle
: / \- —--— Decelcrating
é? 8.86 Performance Proﬁle
'9 7.48
to 5 .
fr
2
Q
m 0 .

1 2 3
Time
“,2 < .001

90

Figure 10d. Performance proﬁle x time interaction for goal commitment.

 

 

 

 

 

 

 

 

 

 

 

 

 

25
20 - C
18.01
E- ' 17560 L296
E 15 ' 16.78 H .. , +Accelerating '
E 15.24 Performance Proﬁle
5 + Decelerating
T. 10 "‘—‘ Performance Proﬁle
0
(D
5 _.
0
1 2 3
Time

 

 

‘p < .05; ”p < .01

91

 

Figure 11a. Feedback ﬂame x performance proﬁle x time interaction for task-self
efﬁcacy with position feedback.

 

 

For Position Feedback

1000

800

 

700
600
500
400
300
200
1 00

+Accelerating
Performance Proﬁle

—-— Decelcrating
Performance Proﬁle

 

 

Task Self-Efﬁcacy

 

Time

 

 

 

 

‘p < .05; "p < .01

92

Figure 11b. Feedback ﬂame x performance proﬁle x time interaction for task-self
efﬁcacy with velocity feedback.

 

 

For Velocity Feedback

 

+Accelerating
Performance Proﬁle

+ Decelerating
Performance Proﬁle

 

 

Task Self-Efficacy

 

Time

 

 

 

 

’p < .05; ”‘p < .001

93

Figure 11c. Feedback ﬂame x performance proﬁle x time interaction for task-self
efﬁcacy with acceleration feedback.

 

For Acceleration Feedback

1000

 

800
700
600
500 --
4m .53
300
200
1 00

l

 

568.38 ear: 1 _. +Accelelating ' :'
Performance Proﬁle i

+ Decelerating
Performance Proﬁlel

 

 

 

 

Task Self-Efficacy

 

 

Time

 

L

.p < .05; mp < .001

 

 

 

94

APPENDIX C

Informed Consent Form

95

INFORMED CONSENT FORM
for the Social Intuition Study

Overview. This study examines how people use their “social intuition” to determine the meaning of
unfamiliar words. Should you choose to participate in this study, you will be asked to provide some
information about yourself, including demographic information (e. g., age, gender, race), personal
preferences (e.g., hobbies, social interaction), and your SAT or ACT test scores as a measure of cognitive
ability. Next, you will be asked to review several lists of foreign words to decide what they mean. Your
responses will then be used to determine how people’s unique characteristics affect the use of their social
intuition.

In order to verify forgotten or illegible SAT or ACT test scores, we will request your authorization to
release your test scores via the MSU Admissions Cmce. You can still participate in this study even if you
choose not to mt this authorization. Please read and initial m of the following statements to indicate
your desired level of participation in this study:

I would like to participate in this study and I am also willing to authorize release of my
Initials SAT/ACT test scores via the MSU Admissions Oﬁce using the attached form.

 

OR

I would like to participate in this study but I gm NOT willing to authorize release of my
Initials SAT/ACT test scores via the MSU Admissions Ofﬁce.

 

Time & Compensation. This study will take approximately one hour of your time to complete. In
exchange for your participation, you will receive one hour of research study extra credit.

Risks. There are no foreseeable risks or discomforts associated with your participation in this study.

Confidentiality. Your privacy will be protected to the maximum extent allowable by law. The responses
you provide during this study will be held in strict conﬁdence by the investigator. You will not be
identiﬁable in any report of this study’s results. When available, you may request a copy of this study’s
results.

Participation. Your participation in this study is completely voluntary. Without penalty, you may refuse
to participate in all or part of this study and may refuse to answer some or all questions.

Contact Information. If you have any questions about this study, please contact the investigator, Dan
Watola, at 340C Psychology Building, East Lansing, MI 48824, (517)432-7752, watolada@msu.edu; or
the Department of Psychology Chair, Dr. Neal Schmitt, at (517) 355-8305, schmitt@msu.edu. If you have
any questions or concerns regarding your rights as a study participant, or are dissatisﬁed at any time with
any aspect of this study, you may contact—anonymously, if you wish—Peter Vasilenko, Ph.D., Chair of
the University Committee on Research Involving Human Subjects (UCRIHS) at (517) 355-2180,
ucrihs@msu.edu, 202 Olds Hall, East Lansing, MI 48824.

Your signature below indicates your voluntary ageement to participate in this study. A copy of this for
will be provided to you upon your request.

 

Participant’s Signature Date

 

Participant’s Printed Name

96

APPENDIX D

Authorization to Release Test Scores Form

97

Authorization to Release ACT and/or SAT Scores

As part of the “Social Intuition” research project conducted by the MSU Department of
Psychology, I authorize the MSU Ofﬁce of Admissions to release my ACT and/or SAT
scores to Dr. Steve Kozlowski. AS soon as my test scores are linked to the other
information I am providing as a participant in this study, any information identifying me
personally will be permanently removed ﬂom the data ﬁle. Furthermore, this information
will not be released to any other research teams or individuals and the results of this study
will only report average test scores that do not permit personal identiﬁcation.

 

 

Sigred Date

 

 

Printed Name PID

98

APPENDIX E

Social Intuition Task Trials 1-12

99

TRIAL 1

Directions: Read the foreign word and decide if it expresses the same meaning
as the adjacent English word. If you decide the two words have similar
meanings, circle “YES”. If you decide their meanings are not similar, circle “NO”.
Your goal is to maximize the number of correct responses across all trials. When
you are ﬁnished with this trial, return this page to your folder and wait for further
instructions.

 

FOREIGN WORD ENGLISH WORD RESPONSE
1. Akce ‘Rain YES NO
2. Neetdu Internet YES NO
3. Senaz Future YES NO
4. Doob Point YES NO
5. Fefe Tree YES NO
6. Tackrote Taxi YES NO
7. Osha Enemy YES NO
8. Wylow Wine YES NO
9. Pahu Food YES NO
10. Bortki Keyboard YES NO

100

TRIAL 2

Directions: Read the foreign word and decide if it expresses the same meaning
as the adjacent English word. If you decide the two words have similar
meanings, circle “YES”. If you decide their meanings are not similar, circle “NO”.
Your goal is to maximize the number of correct responses across all trials. When
you are ﬁnished with this trial, return this page to your folder and wait for further
instructions.

 

FOREIGN WORD ENGLISH WORD RESPONSE
1 . Rajas Harmful YES NO
2. Tanakan Bank YES NO
3. Dhakk Cart YES NO
4. Cownack Journalist YES NO
5. Chudoba Round YES NO
6. Seefanya Toothpaste YES NO
7. Takur Help YES NO
8. Balasin Painting YES NO
9. Naha Dwelling YES NO

10. Pawnzee Bridge YES NO

101

 

TRIAL 3

Directions: Read the foreign word and decide if it expresses the same meaning
as the adjacent English word. If you decide the two words have similar
meanings, circle “YES”. If you decide their meanings are not similar, circle “NO”.
Your goal is to maximize the number of correct resggnses across all trials. When
you are ﬁnished with this trial, return this page to your folder and wait for further
instructions.

 

FOREIGN WORD ENGLISH WORD RESPONSE
1 . Mbiom ha Lucky YES NO
2. Pimcroy Printer YES NO
3. Murda Sit YES NO
4. Maijot Letter YES NO
5. Rokna Title YES NO
6. Gileena Clock YES NO
7. Pastha Fork YES NO
8. Amsue Toilet YES NO
9. Kattora Forgive YES NO
10. Mook Nose YES NO

102

TRIAL 4

Directions: Read the foreign word and decide if it expresses the same meaning
as the adjacent English word. If you decide the two words have similar
meanings, circle “YES”. If you decide their meanings are not similar, circle “NO”.
Your goal is to maximize the number of correct responses across all trials. When
you are ﬁnished with this trial, return this page to your folder and wait for further
instructions.

 

FOREIGN WORD ENGLISH WORD RESPONSE
1 . Ehu Obedient YES NO
2. Bet Ball YES NO
3. Tea Beach YES NO
4. Taykah Fighting YES NO
5. Chunga Ice YES NO
6. Pakroam Campsite YES NO
7. Ghalana Dog YES NO
8. Dygra Rabbit YES NO
9. Madgura Year YES NO
1 0. Angdi Bed YES NO

103

TRIAL 5

Directions: Read the foreign word and decide if it expresses the same meaning
as the adjacent English word. If you decide the two words have similar
meanings, circle “YES”. If you decide their meanings are not similar, circle “NO”.
Your goal is to maximize the number of correct responses across all trials. When
you are ﬁnished with this trial, return this page to your folder and wait for further
instructions.

 

FOREIGN WORD ENGLISH WORD RESPONSE
1 . Kachna Rotten YES NO
2. Soquke Daughter YES NO
3. Starat Open YES NO
4. Hommootsa Library YES NO
5. Sunut Hero YES NO
6. Fanmore Dentist YES NO
7. Eroste Barrel YES NO
8. Weepandee Compact Disk YES NO
9. Oihu Crumb YES NO
10. Ahoo Head YES NO

104

TRIAL 6

Directions: Read the foreign word and decide if it expresses the same meaning
as the adjacent English word. If you decide the two words have similar
meanings, circle “YES”. If you decide their meanings are not similar, circle “NO”.
Your goal is to maximize the number of correct responses across all trials. When
you are ﬁnished with this trial, return this page to your folder and wait for further
instructions.

 

FOREIGN WORD ENGLISH WORD RESPONSE
1 . Pupupu Affection YES NO
2. Dreedon Music YES NO
3. Itibari Growth YES NO
4. Pate Fan YES NO
5. Peleka Calf YES NO
6. Binang Airplane YES NO
7. Rozsah Water YES NO
8. Homepa Blanket YES NO
9. Zidor Tomato YES NO
10. Roakbe Jungle . YES NO

105

TRIAL 7

Directions: Read the foreign word and decide if it expresses the same meaning
as the adjacent English word. If you decide the two words have similar
meanings, circle “YES”. If you decide their meanings are not similar, circle “NO”.
Your goal is to maximize the number of correct responses across all trials. When
you are ﬁnished with this trial, retum this page to your folder and wait for further
instructions.

 

FOREIGN WORD ENGLISH WORD RESPONSE
1. Umatu Job YES NO
2. Na Chest YES NO
3. Chitr East YES NO
4. Sapra Phone YES NO
5. Latta Shadow YES NO
6. Chideck Boy YES NO
7. Plavec Under YES NO
8. Bowgre Suitcase YES NO
9. Andi Sweet YES NO
10. Morsaydu Motorcycle YES NO

106

TRIAL 8

Directions: Read the foreign word and decide if it expresses the same meaning
as the adjacent English word. If you decide the two words have similar
meanings, circle “YES”. If you decide their meanings are not similar, circle “NO”.
Your goal is to maximize the number of correct responses across all trials. When
you are ﬁnished with this trial, return this page to your folder and wait for further
instructions.

 

FOREIGN WORD ENGLISH WORD RESPONSE
1 . Dohain Chair YES NO
2. Nitten Tennis YES NO
3. Haka Dusk YES NO
4. Teetan Week YES NO
5. Shauri Kick YES NO
6. Ahgot Air Conditioning YES NO
7. Gharsa Pray YES NO
8. Roketah Infant YES NO
9. Chinitti Hard YES NO
10. Satonsue Zoo YES NO

107

TRIAL 9

Directions: Read the foreign word and decide if it expresses the same meaning
as the adjacent English word. If you decide the two words have similar
meanings, circle “YES”. If you decide their meanings are not similar, circle “NO”.
Your goal is to maximize the number of cgrrect responses across all trials. When
you are ﬁnished with this trial, return this page to your folder and wait for further
instnlctions.

 

FOREIGN WORD ENGLISH WORD RESPONSE
1 . Lagwa Memory YES NO
2. Gn'gone Scissors YES NO
3. Safha Car YES NO
4. Dib Father YES NO
5. Caj Over YES NO
6. Cowahoo Knee YES NO
7. Rak Get YES NO
8. Firote Train YES NO
9. Jaunze Finish YES NO
10. Egow Couch YES NO

108

TRIAL 1 0

Directions: Read the foreign word and decide if it expresses the same meaning
as the adjacent English word. If you decide the two words have similar
meanings, circle “YES”. If you decide their meanings are not similar, circle “NO”.
Your goal is to maximize the number of correct responses across all trials. When
you are ﬁnished with this trial, return this page to your folder and wait for further
instructions.

 

FOREIGN WORD ENGLISH WORD RESPONSE
1 . Pitzu Improve YES NO
2. Natubla Tuna Fish YES NO
3. Makua Direct YES NO
4. Neesawbry Mail YES NO
5. Kamba Land YES NO
6. Yatun Sick YES NO
7. Donest Report YES NO
8. Dumgra Button YES NO
9. Hertu Art YES NO
10. Tayuwit Radio YES NO

109

TRIAL 1 1

Directions: Read the foreign word and decide if it expresses the same meaning
as the adjacent English word. If you decide the two words have similar
meanings, circle “YES”. If you decide their meanings are not similar, circle “NO”.
Your goal is to maximize the number of correct responses acms all trials. When
you are ﬁnished with this trial, return this page to your folder and wait for further
instructions.

 

FOREIGN WORD ENGLISH WORD RESPONSE
1 . Mutlag Insult YES NO
2. Binamsee Airport YES NO
3. Ganda Even YES NO
4. Atku Bottle YES NO
5. Balwa Change YES NO
6. Diwon Eyeglasses YES NO
7. Rahm Fire YES NO
8. Royah Boat YES NO
9. Elizkoi Bird YES NO

10. Rayto Television YES NO

110

TRIAL 1 2

Directions: Read the foreign word and decide if it expresses the same meaning
as the adjacent English word. If you decide the two words have similar
meanings, circle “YES”. If you decide their meanings are not similar, circle “NO”.

Your goal is to maximize the number of correct responses across all trials. When

you are ﬁnished with this trial, return this page to your folder and wait for further

 

instructions.
FOREIGN WORD ENGLISH WORD RESPONSE

1 . Kalati Education YES NO
2. Dugree Bone YES NO
3. Ranta Human YES NO
4. Toptra Sun YES NO
5. Ribost King YES NO
6. Atroodam Police YES NO
7. Pesha Amount YES NO
8. Labinsin Artist YES NO
9. Khet Conﬂict YES NO

10. Andoo Month YES NO

111

APPENDIX F

Demogaphic and Individual Difference Measures

112

PWNQMPP’P!‘

Demographic Information

What is your sex?

What is your race?

What is your age?

If you took the SAT, write down your total score (math + verbal) in the Space below.
If you took the ACT, write down your total score in the space below.

How many brothers and sisters do you have?

Approximately how many people do you have a conversation with in an average day?
Do you often play word games (e.g., crossword puzzles, word ﬁnds, word of the day)?
Other than English, how many languages do you Speak ﬂuently?

Individual Differences

Trait Anxiety

P?°>‘.°‘S":‘>S’°.NZ“

I feel calm.

I feel secure.

I am tense.

I am regetful.

I feel at ease.

I feel upset.

I am presently worrying over possible misfortunes.
I feel rested.

I feel anxious.

. Ifeel comfortable.

. Ifeel self-conﬁdent.
. Ifeel nervous.

. 1am jittery.

I feel “high strung.”

. I am relaxed.

I feel content.

. I am worried.
. I feel overexcited and rattled.

I feel joyful.
I feel pleasant.

113

APPENDIX G

F arniliarization Trial

114

FAMILIARIZATION TRIAL

Directions: Read the foreign word and decide if it expresses the same meaning
as the adjacent English word. If you decide the two words have similar
meanings, circle “YES”. If you decide their meanings are not similar, circle “NO”.
Your goal is to maximize the ntﬂber pf correct responses across all trials. When
you are ﬁnished with this trial, return this page to your folder and wait for further
instructions.

 

FOREIGN WORD ENGLISH WORD RESPONSE
1 . Palola Degree YES NO
2. Sendotu Peace YES NO
3. Kana New YES NO
4. Adelu Last YES NO
5. Pundra Faster YES NO
6. Luberri Cut YES NO
7. K0 Spirit YES NO
8. Mbogo Red YES NO
9. Phirkta Work YES NO

10. Larazi Leap YES NO

115

APPENDD( H

Feedback Manipulation

116

Decelcrating Performance Proﬁle with Position Feedback

6789101112

Trial

60—

0|

MUD-F
600000

.3

"F

K\

Decelcrating Performance Proﬁle with Position Feedback

.x
on O O
O

l

.h
OOOOOOOOOO

Eulululnlulnlululllululnlnlululllulnll
sl—
“l-

—L
In
00

"ION

de

6789101112

Decelcrating Performance Proﬁle with Velocity Feedback

 

E90

{—70

{—50

 

 

 

 

 

Trials 1-6
- - - Trials 7-12

Accelerating Performance Proﬁle with Velocity Feedback

 

 

 

 

 

 

 

Tn'als 1-6
- - - Trials 7-12

118

Decelerating Performance Proﬁle with Acceleration Feedback

 

 

V

Trial

Accelerating Performance Proﬁle with Acceleration Feedback

 

 

 

119

APPENDIX I

Debrieﬁng Form

120

THANK YOU for participating in this study!

Although the stated purpose of this study is to examine social intuition, the actual
purpose is to examine your reactions to the type of feedback provided to you throughout
the study. Current research suggests that people may respond more or less favorably to
feedback based upon when it is provided, the way in which it is provided, and their
unique individual differences. Therefore, in addition to your reactions, we will also
examine your demographic information and personal preferences to understand how they
may have affected your reactions to the feedback you had received. Ultimately, the
results of this study could be used to determine how to provide the right feedback at the
right time to enhance the motivation and performance of students or trainees. The social
intuition “cover story” was used to ensure that your knowledge of the actual purpose of
this study would not affect your reactions to the feedback. Therefore, it is critical that
you do not discuss the details of this study with anyone who might later participate in this
study.

If you have any questions or concerns regarding your rights as a participant in this study,
you may contact MSU’s Ofﬁce of Research Ethics and Standards at (517) 355-2180. If
you are interested in receiving any additional information about this study, please contact
Dan Watola at (517) 432-7752 or watolada@msu.edu, or read the journal article cited
below.

THANK YOU again for your participation!

Hsee, C. K., & Abelson, R. P. (1991). Velocity relation: Satisfaction as a function of
the ﬁrst derivative of outcome over time. Journal of Personality and Social
Psychology, 60, 341-347.

121

APPENDIX J

Motivational Measures

122

Motivational Measures

State Positive Affect Subscale (Times 1-3)
1. Enthusiastic
2. Interested
3. Determined
4. Excited
5. Inspired
6. Alert

7. Active

8. Strong

9. Proud

10. Attentive

Self-Satisfaction with Task Performance (Times 1-3)

1. I am satisﬁed with my overall performance on this task.
2. I am pleased with how I am doing.

3. My current performance satisﬁes me.

4. At this point, I am happy with my performance.

Goal Commitment (Times 1-3)

1. I am committed to pursuing this goal.

2. I am willing to put forth a great deal of effort to achieve this goal.
3. Quite frankly, I don’t care if I achieve this goal or not.

4. It wouldn’t take much to make me abandon this goal.

5. I think this goal is a good goal to shoot for.

Task Self-Efﬁcacy (Times 1-3)

 

I am % conﬁdent that I can answer % of the items correctly in future trials.
1. % 10%
2. % 20%
3. % 30%
4. % 40%
5. % 50%
6. % 60%
7. % 70%
8. % 80%
9. % 90%
10. % 100%

123

Task Effort (Time 1)
Write down all of the words you can create using the letters found in the word:

BATHROOM

Task Effort (Time 2)
Write down all of the words you can create using the letters found in the word:

BACKPACKER

Task Effort (Time 3)
Write down all of the words you can create using the letters found in the word:

FOOTFALLS

New Task Effort (Time 3)
Write down all the words you can form by linking letters (left, right, below, above, and
diagonally).

 

J
T
E
HOG

 

 

 

I
R
R

HEM

 

2>OW

 

 

 

 

 

 

Manipulation Check (Time 3)

1. My performance on the task seemed to get better over time.
2. My performance on the task seemed to get worse over time.
3. I seemed to do better earlier as compared to more recently.
4. I seemed to do worse earlier as compared to more recently.

124

REFERENCES

Ammons, R. B. (1956). Effects of knowledge of performance: A survey and tentative
theoretical formulation. Journal of General Psychology, 54, 279-299.

Anderson, J. R. (1982). Acquisition of cognitive skill. Psychological Review, 89, 369-
406.

Balcazar, R, Hopkins, B. L., & Suarez, Y. (1985). A critical, objective review of
performance feedback. Journal of Organizational Behavior Management, 7, 65-
89.

Bandura, A. (1986). A social cognitive theory. Englewood Cliffs, NJ: Prentice-Hall.

Bandura, A. (1991). Social cognitive theory of self-regulation. Organizational Behavior
and Human Decision Processes, 50, 248-287.

Bandura, A., & Cervone, D. (1983). Self-evaluative and self-efﬁcacy mechanisms
governing the motivational effects of goal systems. Journal of Personality and
Social Psycholog, 45, 1017-1028.

Bandura, A., & Cervone, D. (1986). Differential engagement of self-reactive inﬂuences
in cognitive motivation. Organizational Behavior and Human Decision
Processes, 38, 92-113.

Brief, A. P., & Weiss, H. M. (2002). Organizational behavior: Affect in the workplace.
Annual Review of Psychology, 53, 279-307.

Burke, M. J ., Brief, A. P., George, J. M., Roberson, L., & Webster, J. (1989). Measuring
affect at work: Conﬁrmatory analyses of competing mood structure with

conceptual linkage to cortical regulatory systems. Journal of Personality and
Social Psychology, 5 7, 1091 -1 102.

Campion, M. A., & Lord, R. G. (1982). A control systems conceptualization of the goal-
setting and changing process. Organizational Behavior and Human Performance,
30, 265-287.

Carver, C. S., & Scheier, M. F. (1982). Control theory: A useful conceptual framework
for personality-social, clinical, and health psychology. Psychological Bulletin,
92, 111-135.

Carver, C. S., & Scheier, M. F. (1990). Origins and functions of positive and negative
affect: A control-process View. Psychological Bulletin, 97, 19-35.

125

Carver, C. S., & Scheier, M. F. (1998). On the self-regulation of behavior. Cambridge,
UK: Cambridge University Press.

Cdte, S. (1999). Affect and performance in organizational settings. Current Directions
in Psychological Science, 8, 65-68.

Crossman, E. R. F. W. (1959). A theory of the acquisition of speed-skill. Ergonomics,
2, 153-166.

Diener, E., Larson, R. J ., Levine, S., & Emmons, R. A. (1985). Intensity and frequency:
Dimensions underlying positive and negative affect. Journal of Personality and
Social Psychology, 48, 1253-1265.

Elliot, E. S., & Church, M. A. (1997). A hierarchical model of approach and avoidance
achievement motivation. Journal of Personality and Social Psychology, 72, 24-
33.

Erez, M. (1977). Feedback: A necessary condition for the goal setting-performance
relationship. Journal of Applied Psychology, 62, 624-627.

Frijda, N. H. (1988). The laws of emotion. American Psychologist, 43, 349-358.

George, J. M. (1996) Trait and state affect. In K. R. Murphy (Ed.), Individual
diﬂ‘erences and behavior in organizations (pp. 145-171). San Francisco: J ossey-
Bass.

George, J. M., & Brief, A. P. (1996). Motivational agendas in the workplace: The
effects of feelings on focus of attention and work motivation. In B. M. Staw & L.
L. Cummings (Eds), Research in organizational behavior (Vol. 18, pp. 75-109).
Greenwich, CT: JAI Press.

Higgins, E. T. (1987). Self-discrepancy: A theory relating self and affect.
Psychological Review, 94, 319—340.

Hollenbeck, J. R., Williams, C. R., & Klein, H. J. (1989). An empirical examination of
the antecedents of commitment to difficult goals. Journal of Applied Psychology,
74, 18-23.

Hsee, C. K., & Abelson, R. P. ( 1991). Velocity relation: Satisfaction as a function of
the ﬁrst derivative of outcome over time. Journal of Personality and Social
Psychology, 60, 341-347.

Hsee, c. K., Abelson, R. P., & Salovey, P. (1991). The relative weighting of position
and velocity in satisfaction. Psychological Science, 2, 263-266.

126

Hsee, C. K., Salovey, P., & Abelson, R. P. (1994). The quasi-acceleration relation:
Satisfaction as a ﬁmction of the change in velocity of outcome over time. Journal
of Experimental Social Psychology, 30, 96-111.

Johnson, R. E., & Lord, R. G. (April, 2005). Moving beyond discrepancies: The
importance of progress rate. In Aﬂective experiences at work: The role of self-
regulation. In S. Sonnentag (Chair), Symposium conducted at the meeting of the
Society for Industrial and Organizational Psychology Conference, Los Angeles,
CA.

Ilgen, D. R., Fisher, C. D., & Taylor, M. S. (1979). Consequences of individual
feedback on behavior in organization. Journal of Applied Psychology, 64, 349-
371.

Kahneman, E., & Tversky, A. (1979). Prospect theory: An analysis of decision-making
under risk. Econometrica, 47, 263-291.

Kahneman, E., & Tversky, A. (1984). Choices, values and frames. American
Psychologist, 39, 341-350.

Kanfer, R. (1990). Motivation theory and industrial and organizational psychology. In
M. D. Dunnette & L. M. Hough (V 01. Eds.), Handbook of industrial and
organizational psychology: Vol. 1. (2nd ed., pp. 75-170). Palo Alto, CA:
Consulting Psychologists Press.

Karl, K. A., O’Leary-Kelly, A. M., & Martocchio, J. J. (1993). The impact of feedback
and self-efﬁcacy on performance in training. Journal of Organizational
Behavior, 14, 379-394.

Kluger, A. N., & DeNisi, A. (1996). The effects of feedback interventions on
performance: A historical review, a meta-analysis, and a preliminary feedback
intervention theory. Journal of Applied Psychology, 119, 254-284.

Knowlton, B. J ., & Squire, L. R. (1996). Artiﬁcial grammar learning depends on
implicit acquisition of both abstract and exemplar-speciﬁc information. Journal
of Experimental Psychology: Learning, Memory, and Cognition, 22, 169-181.

Kozlowski, S. W. J ., Toney, R. J ., Mullins, M. E., Weissbein, D. A., Brown, K. G., &
Bell, B. S. (2001). Developing adaptability: A theory for the design of
integrated-embedded training systems. In E. Salas (Ed.), Advances in human
performance and cognitive engineering research (Vol. I, pp. 59-123).
Amsterdam: JAI/Elsevier Science.

Latham, G. P., & Locke, E. A. (1991). Self-regulation through goal setting.
Organizational Behavior and Human Decision Processes, 50, 212-247.

127

Lawrence, J. W., Carver, C. S., & Scheier, M. F. (2002). Velocity toward goal
attainment in immediate experience as a determinant of affect. Journal of Applied
Social Psychology, 32, 788-802.

Lee, C., & Bobko, P. (1994). Self-efﬁcacy beliefs: Comparison of ﬁve measures.
Journal of Applied Psychology, 79, 364-369.

Locke, E. A. (1982). Relation of goal level to performance with a short work period and
multiple goal levels. Journal of Applied Psychology, 67, 512-514.

Locke, E. A., Frederick, E., Lee, C., & Bobko, P. (1984). Effect of self-efﬁcacy, goals,
and task strategies on task performance. Journal of Applied Psychology, 69, 241 -
251.

Locke, E. A., & Latham, G. P. (1990). A theory of goal setting and task performance.
Englewood Cliffs, NJ: Prentice Hall.

Locke, E. A., Shaw, K. N., Saari, L. M., & Latham, G. P. (1981). Goal setting and task
performance; 1969-1980. Psychological Bulletin, 90, 125-152.

Logan, G. D. (1988). Toward an instance theory of automatization. Psychological
Review, 95, 492-527.

Logan, G. D. (1992). Shapes of reaction-time distributions and shapes of learning
curves: A test of the instance theory of automaticity. Journal of Experimental
Psychology: Learning, Memory, and Cognition, 18, 883-914.

Lord, R. G., & Hanges, P. J. ( 1987). A control system model of organizational
motivation: Theoretical development and applied implications. Behavioral
Science, 32, 161-178.

MacKay, D. G. (1982). The problem of ﬂexibility, ﬂuency, and speed-accuracy trade-
off in skilled behavior. Psychological Review, 89, 483-506.

Martocchio, J. J., & Webster, J. (1992). Effects of feedback and cognitive playfulness
on performance in microcomputer software training. Personnel Psychology, 45,
553-578.

Mayer, G. J ., & Shack, J. R. (1989). Structural convergence of mood and personality:
Evidence for old and new directions. Journal of Personality and Social
Psychology, 5 7, 691-706.

Mento, A. J ., Steel, R. P., Karren, R. J. (1987). A meta-analytic study of the effects of

goal setting on task performance: 1966-1984. Organizational Behavior and
Human Decision Processes, 39, 52-83.

128

Miller, G. A, Galanter, E., & Pribram, K. H. (1960). Plans and the structure of
behavior. New York: Holt, Rinehart, & Winston.

Mitchell, T. R., & Daniels, D. M. (2003). Motivation. In W. C. Borman, D. R. Ilgen, &
R. J. Klirnoski (V 01. Eds.), Handbook of psychology: Vol. 12. Industrial and
organizational psychology. (pp. 225-276). Hoboken, NJ: Wiley.

Murphy, K. R., & Myors, B. (2004). Statistical power analysis: A simple and general
model for traditional and modern hypothesis tests (2“ld ed.). Mahwah, NJ:
Lawrence Erlbaum Associates.

Newell, A., & Rosenbloom, P. S. (1981). Mechanisms of skill acquisition and the law of
practice. In J. R. Anderson (Ed.), Cognitive skills and their acquisition (pp. 1-55).
Hillsdale, NJ: Lawrence Erlbaum Associates.

Ozer, E. M., & Bandura, A. (1989). Mechanisms governing empowerment effects: A
self-efﬁcacy analysis. Journal of Personality and Social Psychology, 58, 472-
486.

Pritchard, R. D., Jones, S. D., Roth, P. L., Stuebing, K. K, & Ekeberg, S. E. (1988).
Effects of group feedback, goal setting, and incentives on organizational
productivity. Journal of Applied Psychology, 73, 337-358.

Ritter, F. E., & Schooler, L. J. (2001). The learning curve. In N. J. Smelser & P. B.
Baltes (Eds.), International encyclopedia of the social and behavioral sciences
(pp. 8602-8605). Amsterdam: Pergamon.

Russell, J. A. (1980). A circumplex model of affect. Journal of Personality and Social
Psychology, 39, 1 161-1178.

Russell, J. A. (1983). Pancultural aspects of human conceptual organization of
emotions. Journal of Personality and Social Psychology, 45, 1281-1288.

Schwarz, N. (1990). Feelings as information. In E. T. Higgins & R. M. Sorrentino
(Eds.), Handbook of motivation and cognition: Foundations of social behavior
(Vol. 2, pp. 527-561). New York: Guilford Press.

Simon, H. A. (1967). Motivational and emotional controls of cognition. Psychological
Review, 74, 29-39.

Spielberger, C. D. (1983). Manual for the State-Trait Anxiety Inventory (S T A1).
Palo Alto, CA: Consulting Psychologists Press.

Stajkovic, A. D., & Luthans, F. (1998). Self-efﬁcacy and work-related performance: A
meta-analysis. Psychological Bulletin, 124, 240-261.

129

 

Tabachnick, B. G., & Fidell, L. s. (2001). Using multivariate statistics (4th Ed.).
Boston: Allyn & Bacon.

Toney, R. J ., & Kozlowski, S. W. J. (April, 1999). Effects of evaluative feedback and
task difficulty on learning and training performance. In S. Gully (Chair),
Learning to fail or failing to learn ?: The role of errors, failure, and feedback in
learning environments. Symposium conducted at the meeting of the Society for
Industrial and Organizational Psychology Conference, Atlanta, GA.

Tubbs, M. E. (1986). Goal-setting: A meta-analytic examination of the empirical
evidence. Journal of Applied Psychology, 71, 474-483.

VandeWalle, D., & Cron, W. L., & Slocum, J. W. (1999). The role of goal orientation
following performance feedback. Journal of Applied Psychology, 82, 629-640.

VanLehn, K. (1996). Cognitive skill acquisition. Annual Review of Psychology,
4 7, 5 13-539.

Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief
measures of positive and negative affect: The PAN AS scales. Journal of
Personality and Social Psychology, 54, 1063-1070.

Watson, D., & Tellegen, A. (1985). Toward a consensual structure of mood.
Psychological Bulletin, 98, 219-235.

Weiner, N. (1948). Cybernetics: Control and communication in the animal and the
machine. Cambridge, MA: MIT Press.

Wood, R. E., & Bandura, A. (1989). Social cognitive theory of organizational
management. Academy of Management Review, 14, 361-3 84.

Wood, R. E., Mento, A. J ., & Locke, E. A. (1987). Task complexity as a moderator of
goal effects: A meta-analysis. Journal of Applied Psychology, 72, 416-425.

Wright, P. M. (1990). Operationalization of goal difficulty as a moderator of the goal
difﬁculty performance relationship. Journal of Applied Psychology, 75, 227-234.

Zevon, M. A., & Tellegen, A. (1982). The structure of mood change: An

idiographic/nomothetic analysis. Journal of Personality and Social Psychology,
43 , 1 11-122.

130