2011. Vol.2, No.3, 187-201
Copyright © 2011 SciRes. DOI:10.4236/psych.2011.23030
Self-Regulation of Goals and Performance: Effects of Discrepancy
Feedback, Regulatory Focus, and Self-Efficacy*
Jessica M. Nicklin1#, Kevin J. Williams2
1Department of Psychology, University of Hartford, West Hartford, USA;
2State University of New York at Albany, Albany, USA.
Received January 1st, 2011; revised February 24th, 2011; accepted April 2nd, 2011.
We adopted a social cognitive approach of motivation (Bandura, 1986, 1989, 2003) to examine the influence of
normative feedback and self-set goals on positive discrepancy creation and goal revision in the face of a novel
task. The moderating effects of self-efficacy and regulatory focus were also examined. A laboratory study in-
cluding 297 undergraduate students demonstrated that feedback, whether based on normative standards of per-
formance or goal-performance discrepancies was a strong predictor of positive discrepancy creation and goal re-
vision. Self-efficacy was also an independent predictor of goal revision, but regulatory focus was not. These
findings have important practical implications for a variety of performance contexts (e.g., work, school, sports).
Individuals will modify their goals based largely on feedback received (goal-performance discrepancies and
normative standards); however, self-efficacy independently influences goal revision beyond the effects of feed-
back. Other implications for research and practice are discussed.
Keywords: Motivation, Performance, Self-Regulation, Goal-Performance Discrepancies, Feedback, Regulatory
Focus, Self-Efficacy
Goals are a common means of motivating people and are
used as standards for evaluating performance (e.g., Locke &
Latham, 1990, 2002). Goals that are specific and explicit pro-
vide individuals with direction that help enhance and sustain
motivation in achievement settings (e.g., Locke & Latham,
2002). Through self-regulation, individuals are able to guide
goal-directed activities over time (Karoly, 1993).
Bandura’s social cognitive theory (SCT: 1986, 1989, 2003)
suggests a two-process model of self-regulation that identified
discrepancy production and discrepancy reduction processes.
Discrepancy production occurs when individuals set desirable,
challenging goals at a level higher than previous levels of per-
formance in an attempt to motivate themselves toward higher
levels of performance. Discrepancy reduction processes occur
when individuals seek to reduce undesirable goal-performance
discrepancies (GPD) in order to receive a positive self-evaluation.
Although there is some empirical support for social cognitive
approaches to motivation (e.g., Donovan & Williams, 2003;
Donovan & Hafsteinsson, 2006; Williams, Donovan, & Dodge,
2000), insufficient attention has been paid to how people use
normative standards and self-set goals to regulate performance
over time on novel tasks. Therefore, the goal of the present
study is to examine how discrepancy feedback, which indicates
how performance deviates from normative standards and self-
set goals, influences discrepancy production and discrepancy
reduction processes. The roles of regulatory focus and self-
efficacy will also be examined, which have not been studied
together in the context of dynamic self-regulation.
Theoretical Rationale and Hypotheses
A Social Cognitive Approach to Motivation
Social cognitive theory (SCT: Bandura, 1986, 1989, 1991,
2003) suggests that individuals form cognitive representations
of action to be performed (intentions), motivate themselves and
guide their actions in the anticipation of future events (fore-
thought), and evaluate their motivation, behavior, values, and
meaning of life with purpose (self-reactiveness). Thus, SCT
views humans as agents of their own behavior, anticipatorily
moving through the world with intention and purpose.
One way to motivate and regulate behavior is through the use
of goals. Many contemporary theories of motivation view per-
sonal goals as a primary regulator of behavior (e.g., Bandura,
2003; Donovan & Williams, 2002; Locke & Latham, 1990).
Although SCT recognizes the value of goals for regulating
performance, Bandura argues that the actual goals themselves
do not motivate individuals. Rather, the positive self-evaluation
and anticipated self-satisfaction gained from reaching goals
provides incentive for action (Bandura, 1989; 1991). For exam-
ple, a goal of a 3.5 grade point average (GPA) does not moti-
vate a person; rather, the anticipated self-satisfaction derived
from reaching the goal provides the incentive for action and
enhanced effort.
*This article is based on Jessica M. Nicklin’s doctoral dissertation, com-
pleted under the supervision of Kevin J. Williams at the University at Al-
any, State University of New York. A previous version of this article was
presented at the 25th Annual Conference of the Society for Industrial and
Organizational Psychology, Atlanta, Georgia, April 2010.
#Correspondence concerning this article should be addressed to Jessica M.
icklin, Department of Psychology, University of Hartford, West Ha
CT 06117. Email:
In order to gauge progress towards goals, self-regulation re-
lies on self-observation and evaluation, which provides the
information and feedback needed for setting realistic goals and
for evaluating one’s progress toward them (Bandura, 1991).
When performance does not match goals, a negative discrep-
ancy is observed, which creates self-dissatisfaction that serves
as an incentive motivator for enhanced effort (Bandura, 1989).
Bandura labeled this a two process model of discrepancy pro-
duction and discrepancy reduction.
Discrepancy production refers to the process by which indi-
viduals set desirable, challenging goals at a level higher than
previous levels of performance, thereby creating a discrepancy
between their current performance level and performance goal.
This discrepancy provides feedback about competence/ability.
The self-dissatisfaction felt due to the discrepancy will auto-
matically trigger the individual to reduce the incongruity (Ban-
dura, 1986, 1989). Thus, individuals also engage in discrepancy
reduction processes by monitoring goal-performance discrep-
ancies (GPD) and working toward reducing the discrepancy in
order to achieve a positive self-evaluation. For example, if a
person sets a goal GPA of 3.5 for the semester (discrepancy
production), and receives a 2.5 on a midterm (feedback), the
person can use this feedback to evaluate performance relative to
the goal. Because there is a negative discrepancy between the
goal (3.5) and current performance (2.5) a sense of dissatisfac-
tion is felt. The individual will either increase effort or reduce
the goal so that performance matches the standard. Effective
self-regulation is characterized by alternating cycles of dis-
crepancy production and discrepancy reduction, expectantly
leading to higher levels of performance through goal attainment.
Comparative feedback is essential in the ongoing regulation of
motivation. Two forms of comparative feedback will be dis-
cussed: normative standards and GPD.
Self-Regulation and Performance Feedback
Normative standards. When people perform a task for the
first time, they rely on normative standards or assigned goals to
monitor and judge performance. As they gain experience, the
regulation of performance is likely to shift from external to
internal standards. In an early test of SCT, Bandura and Cer-
vone (1986) asked participants to pursue a challenging standard
in a strenuous activity and provided them with feedback that
their effort fell markedly, moderately, or minimally short of the
standard. Results showed that participants were dissatisfied
with large discrepancies, and as the discrepancy narrowed, they
become more satisfied with their attainments. When they were
informed that their performance fell just short of or surpassed
their goals, some even raised their goals.
Despite evidence of positive discrepancy creation (PDC: set-
ting future goals higher than previous levels of performance),
little attention has been paid to the relative amount of PDC in
response to success and failure. Based on prospect theory
(Tversky & Kahneman, 1992), Heath, Larrick, and Wu (1999)
found that motivation to improve is stronger when performance
is below a standard than when it is above. Thus, losses of a
particular value are more painful than similar size gains are
pleasurable. This suggests that negative discrepancies may
produce greater motivation than positive discrepancies of com-
parable amounts.
Based on SCT, people with both positive and negative dis-
crepancies should raise goals above current levels of perform-
ance, but for different reasons. Individuals with negative per-
formance discrepancies will be dissatisfied with current levels
of performance and will be motivated to reduce the discrepancy.
Thus, they should have a larger amount of PDC than individu-
als with a positive discrepancy, who have already exceeded a
standard. These individuals will be satisfied with performance
and will have a smaller PDC. Based on this reasoning, we pre-
dict the following:
Hypothesis 1: The tendency to create a positive discrepancy
between one’s goal and recent performance will be stronger
when performance is below a normative standard (negative
discrepancy feedback), than when performance is above a
normative standard (positive discrepancy feedback).
Goal performance discrepancies as feedback. As indi-
viduals pursue a task over time, the source of feedback be-
comes the person’s current performance relative to their goals
and past performance. In the dual-cyclic model, goal perform-
ance discrepancies (GPD: performance—goal) become the
primary source of feedback for future goal revision and per-
formance. For instance, Williams et al., (2000) and Donovan
and Williams (2003) asked varsity track and field athletes to
report their goals and progress during the track and field season.
In both studies, participants created positive discrepancies by
setting competition goals higher than their best previous per-
formance. Participants were most likely to lower their goals
when performance fell substantially short of the goal, but were
likely to raise their goals when GPD were small. The PDC ef-
fect has also been demonstrated in educational contexts (e.g.,
Phillips, Hollenbeck, & Ilgen, 1996); however, to our knowl-
edge little research has examined the dual-cyclic model when
the task is novel. Based on SCT, it is expected that in the face
of novel tasks the same relationship between GPD and goal
revision will apply.
Hypothesis 2: GPD feedback will be related to goal revision,
such that positive discrepancies (performance > goal) will re-
sult in upward goal revision and negative discrepancies (per-
formance < goal) will result in downward goal revision. The
degree of goal revision will be related to the size of the GPD.
Task persistence may also be related to the size and direction
of GPD. If a negative GPD is large or the goal is perceived as
unattainable, individuals may give up on the goal and direct
attention to another task. However, if a negative GPD is small
and perceived as attainable, then according to SCT, individuals
should persist at the task in order to raise current performance
to the goal. Conversely, individuals with very large positive
GPD may become bored with the task and direct effort toward
another goal. Therefore, we predict:
Hypothesis 3: A curvilinear relationship will exist between
task withdrawal and GPD, such that people will be mostly like-
ly to withdrawal from the task when large negative and large
positive GPD exist.
Individual Va ri a bl es and Motivati on
Although some research lends support for Bandura’s model,
there is little known about specific individual variables that
influence goal revision processes; although some initial re-
search has been conducted to examine this issue. For instance,
Donovan and Hafesteinsson (2006) found that self-efficacy and
goal orientation moderated the GPD-goal revision relationship.
Phillips et al. (1996) found that performance expectancy and
need for achievement moderated the relationship between GPD
and goal revision. The present study seeks to explore the influ-
ence of two individual variables on self-regulated motivation on
a novel task: regulatory focus and self-efficacy.
Regulatory focus. Higgins’ (1997) regulatory focus theory
suggests that there are different ways of approaching desired
end states. A promotion focus is a desire for ideals, advance-
ment and accomplishments. A prevention focus is a concern for
oughts, protection, and security. For example, a student with a
promotion focus may be more motivated to strive for advance-
ment and reach the ideal GPA; whereas a student with a pre-
vention focus may be more motivated to get the grade s/he
“ought” to get and avoid looking bad.
Given that regulatory focus concerns variation in
self-regulation, it is possible that regulatory focus plays a role
in self-regulated motivation as delineated by Bandura. Empiri-
cal research supports this supposition. For instance, Van-Dijk
and Kluger (2004) found that under a promotion focus, positive
feedback increased motivation more than negative feedback,
whereas under a prevention focus, negative feedback increased
motivation more than positive feedback. Spiegel and Higgins
(2001) found that promotion-focused participants performed
better on the second round of an anagram task after receiving
success feedback on the first round of the task, whereas preven-
tion participants performed better on the second round of the
task after receiving failure feedback on the first round of the
task. Idson and Higgins (2000) found that with a promotion
focus, a more positive performance improvement was present
following success than failure feedback, whereas for the pre-
vention focus a more positive performance improvement was
present following failure than success feedback.
Idson and Higgins (2000) concluded that these patterns are
consistent with the belief of high promotion individuals that
there is “everything to gain.” This belief is maintained after
success and reduced after failure. In contrast, the belief of high
prevention individuals that there is “everything to lose” is
maintained after failure. Given evidence that regulatory focus
influences how one responds to performance feedback it is
possible that regulatory focus also influences how one responds
to GPD that vary in size and direction, which are also a form of
feedback. Therefore, we predict:
Hypothesis 4a: When positive GPD exist, those with a pro-
motion focus will be more likely to revise their goals upward
than those with a prevention focus.
Hypothesis 4b: When negative GPD exist, those with a pre-
vention focus will be less likely to lower their goals than those
with a promotion focus.
Self-efficacy. Self-efficacy is “self-belief in one’s capabili-
ties to exercise control over events and accomplish desired
goals” (Wood & Bandura, 1989, p. 364). Bandura (1989) em-
phasized the importance of self-efficacy for perseverance and
performance. Thus, in the face of negative performance dis-
crepancies, it is likely that those high in self-efficacy will con-
tinue to generate PDC because they believe they are capable of
succeeding, whereas those low in self-efficacy will be likely to
reduce or abandon their goals because they do not believe they
are capable of succeeding. Conversely, when in the face of
positive GPD, it is likely that those high in self-efficacy will
create even larger discrepancies because they believe they are
capable of succeeding; whereas, those low in self-efficacy will
be likely to reduce or abandon their goals because they do not
believe they are capable of succeeding.
Previous research supports this prediction. Bandura and
Cervone (1986) found that participants’ perceived self-efficacy
increased as performance discrepancy narrowed; the stronger
the perceived self-efficacy for goal attainment, the higher the
level of self-set goals and the greater the self-dissatisfaction
with substandard attainments. Donovan and Hafesteinsson
(2006) found that those high in self-efficacy revised their goals
upward to a greater extent in response to a positive GPD than
individuals low in self-efficacy. The present study seeks to
expand these findings by examining the influence of
self-efficacy on goal revision in the face of a novel task .
Hypothesis 5: Self-efficacy will be positively related to goal
revision, such that higher self-efficacy will be associated with
greater levels of upward goal revision.
Present Study
The present study examined the effects of performance dis-
crepancies, regulatory focus, and self-efficacy on PDC and goal
revision. We used a novel task, the children’s game OPERA-
TION. This permitted us to provide normative standards of
feedback and to assess performance and goal revision over
multiple trials. Participants were unlikely experienced at the
game; thereby making it a novel task for many. We also chose
this task because it was moderately difficult and provided op-
portunities for success and failure. Pilot studies were conducted
prior to the main study. Pilot Study 1 assessed the reliability of
measures of regulatory focus. Pilot Study 2 examined the ap-
propriateness of the study task.
Pilot Studies
Pilot Study 1
Participants and procedures. One-hundred undergraduate
students from a psychology class at a Northeastern university in
the United States volunteered to participate in this study. Par-
ticipants were asked to answer a survey. The study duration
was ten minutes.
Measures. Regulatory focus was measured using the Regu-
latory Focus Questionnaire (RFQ) developed by Higgins,
Friedman, Harlow, Idson, Ayduk, and Taylor (2001) and the
Promotion/Prevention Scale (PP) developed by Lockwood,
Jordan, and Kunda (2002).
The RFQ (Higgins et al., 2001) consists of two subscales,
which are intended to measure individuals’ subjective histories
of promotion and prevention success on a five-point scale (1 =
never/certainly false to 5 = very often/certainly true). The pro-
motion subscale consists of six items. A sample item is: “How
often have you accomplished things that got you psyched to
work even harder?” The prevention subscale contains five
items. A sample item is: “How often did you obey rules and
regulations that were established by your parents?”
The PP (Lockwood et al., 2002) measure consists of two
subscales, which are intended to measure promotion and pre-
vention on a 9-point rating scale (1 = not at all true of me to 9 =
very true of me). A sample promotion item is: I frequently
imagine how I will achieve my hopes and aspirations.” A sam-
ple prevention item is: “In general, I am focused on preventing
negative events in my life.”
Results. The RFQ (
prevention = .69;
promotion = .79) and the
PP (
prevention = .79;
promotion = .86) both reached acceptable
levels of reliability. Exploratory factor analysis with oblimin
rotation on the RFQ measure did not show support for two
distinct factors. The scree plot showed four clear bends, and
four eigenvalues that exceeded 1.00. When two factors were
forced, the items loaded as expected. Similarly, the PP measure
showed 4 - 5 bends in the scree plot and five eigenvalues that
exceeded 1.00. When two factors were forced, most items
loaded as expected; however, cross-loadings were present.
The PPpromotion was positively correlated with the RFQpromotion (r
= .41, p < .01), and negatively related to the PPprevention (r = .41,
p < .01) and RFQprevention, r = .21, p < .05. However, the PPpre-
vention and the RFQprevention were not correlated (r = .07, ns) indi-
cating a lack of convergent validity for the prevention subscales.
Pilot Study 2
Participants and procedures. The purpose of pilot study 2
was to re-examine the reliability of the regulatory focus meas-
ures and to determine the suitability of the experimental task.
Sixty undergraduate students (53.3% female) from a psychol-
ogy research pool at a Northeastern university in the United
States volunteered to participate in this study for research credit.
Students that participated in Pilot Study 1 were not eligible to
participate in Pilot Study 2.
Participants first filled out the RFQ and PP to assess chronic
regulatory focus. Then they played the game OPERATION.
The game board is an illustration of a surgery patient, and the
goal of the game is to remove as many pieces (body parts) from
the game board as possible with metal tweezers without hitting
the sides. If the sides of the game board are touched a buzzing
sound occurs. Participants were asked to remove as many piec-
es from the game board as possible in 60-seconds. For every
successful piece removed, participants were awarded points
based on difficulty of the piece. When a mistake was made (the
tweezer hit the side of the board) participants would lose points
from their score. Point values were determined by a perform-
ance norming study involving a separate sample prior to this
Upon completion of the performance trial, a negative per-
formance discrepancy was created by informing participants
that they missed the “average” score by 10, 30, 50, or 100
points. Participants were asked if they would like to play the
game OPERATION again or put together a 50 piece puzzle. If
they chose OPERATION they reported a goal.
Measures. As in Pilot Study 1, regulatory focus was meas-
ured using the RFQ and the PP. Negative normative feedback
was provided to participants in order to create perceptions of
substandard performance. As a manipulation check, participants
were asked: “How would you rate your performance on this
task on an 11-point Likert format scale (1 = extremely far
below average; 6 = average; 11 = extremely far above average).
Goal performance discrepancy creation was assessed by asking
participants to set a new goal for a second performance trial and
then subtracting their initial performance score from the goal.
Persistence was assessed by giving participants a choice be-
tween OPERATION or a puzzle.
Results. Responses to the manipulation check item demon-
strated that the difference between the feedback conditions was
significant, F (3, 56) = 6.86, p < .01,
2 = .27. The means on an
11-pt scale (with 6 being equal to average) were as follows: 10
points below average (M = 5.13, SD = .99), 30 points below
average (M = 3.80, SD = 1.97), 50 points below average (M =
3.40, SD = 1.40), and 100 points below average, M = 2.67, SD
= 1.59. Helmert contrasts demonstrated that only the small (10
points) condition was significantly (p < .01) different than the
other groups.
Univariate analysis of variance demonstrated that normative
feedback was significantly related to discrepancy production, F
(3, 23) = 6.77, p < .01,
2 = .47. Individuals in the 10 point
condition set their performance goals 18.57 points above their
initial performance, SD = 14.35. Those in the 30 and 50 point
conditions set their performance goals 20.00 (SD = 10.49) and
38.33 (SD = 18.07) points above their initial performance, re-
spectively. Participants in the 100 point condition raised their
performance goals on average 61.56 (SD = 30.56) points above
their initial performance. All Helmert contrasts were significant
(p < .05). Task choice (puzzle or operation) was not related to
any of the variables.
The internal reliabilities of the RFQ subscales were
= .62 and
prevention = .82. The PP subscales demonstrated internal
reliabilities of
promotion = .79 and
preventi on = .75. RFQprevention was
positively correlated with RFQpromotion (r = .29, p < .05) and PPpro-
motion (r = .30, p < .05), which demonstrates a lack of discriminant
validity. The PPpromotion and PPprevention were not correlated, r = –.01,
ns. Furthermore, by examining the scree plot and eigenvalues
greater than 1.00, exploratory factor analysis with oblimin rotation
demonstrated five factors for the PP measure and three factors for
the RFQ. When two factors were forced for each of the scales, the
items loaded where expected, but cross loadings were present.
There were no significant correlations between regulatory focus
and performance or goal change.
Summary of Pilot Studies
Normative performance feedback influenced goal change.
The main study expands these findings by examining goal revi-
sion across repeated performance trials in the face of positive
and negative performance discrepancies. Further, as demon-
strated in both studies, the use of chronic regulatory focus
measures lacked clear validity evidence, although the sample
sizes were small. Therefore, regulatory focus will be manipu-
lated as a state variable in the main study opposed to a trait
variable in the pilot studies.
Participants were 297 undergraduate students (50.7% male)
from a large public university in the northeast United States.
Recruitment was conducted through the university’s research
pool. Students earned credits toward the research requirement.
The majority of the sample was Caucasian, 66.7%. The re-
mainder of the sample was diverse: 9.1% African American,
7.7% Hispanic/Latino, 12.5% Asian, and 3.7% other. The mean
age was 18.71 (SD = 1.39; range 17 - 27) and a majority of the
sample were Freshmen, 62%. Those that participated in the
pilot studies were not eligible to participate in this study.
Study task. Participants were asked to play the game OP-
ERATION. Sixty-seven percent of the sample reported that
they had played the game as a child, but only 9.1% reported
they had played it in the past five years. This information sug-
gests that the vast majority of participants did not have recent
experience with the game. The variability in scores described
below further indicates that participants, in general were not
experts at the game.
Participants had 60 seconds to remove as many pieces from
the game board as possible. Each piece was given a
pre-determined point value based on difficulty. If the partici-
pants successfully removed the piece without making a mistake,
they would receive the point value for that piece. If they hit the
side of the game board, they lost half the point value of that
piece from their score.
Manipulated variables. Prior to the task, participants were
placed in either a promotion or prevention focus frame via task
instructions modeled after Shah and Higgins (1997). The pro-
motion focus frame indicated that their chances of succeeding
at the task would be greatest if they maximize the number of
points earned by getting as many pieces out as possible. The
prevention focus indicated that their chances of failing at the
task would be lessened if they minimize the number of points
lost by making the fewest mistakes possible. The instructions
were framed such that the goal was the same: to perform well
on the task. However, the focus was manipulated.
The second manipulation was feedback direction. Partici-
pants were given an initial opportunity to engage in the OP-
ERATION task in the absence of normative standards or salient
goals. All participants were told their initial score either ex-
ceeded the average score by 50 points or missed the average
score by 50 points. Fifty was chosen based on Pilot Study 2
Experimental procedures. The two manipulated variables
produced four conditions. Participants were randomly assigned
to one of the conditions using a random numbers table. Upon
arrival, participants were asked to sign an informed consent and
the experimenter checked to ensure that they had not already
participated in the pilot studies.
Participants were first asked to complete the Lockwood et al.
(2002) regulatory focus measure to assess baseline levels of
regulatory focus. Participants were then given an assessment of
manual dexterity using the Grooved Pegboard Test (Lafayette
Instrument Company, 1997).
Participants were next read instructions by the experimenter:
We are interested in examining your strategy and performance
on tasks that require precision and careful attention to detail.”
Initial directions for the task were manipulated for the promo-
tion or prevention focus as described above. Participants filled
out a measure of self-efficacy and engaged in an initial trial of
OPERATION. The experimenter tallied the score and told par-
ticipants that their score had either exceeded or missed the av-
erage by 50 points.
For trials 2 and 3 participants were given a self-efficacy
questionnaire for each upcoming trial, and then asked to set a
goal for the trial. Participants were reminded of the task direc-
tions and were asked to engage in the task for 60 seconds. After
each performance trial, participants were given their feedback
and told how they did relative to their goal set (eliciting GPD).
Finally, after setting the goal for trial 4 participants were
asked if they would like to play OPERATION again or put
together a puzzle. After engaging in the final task of their
choice participants filled out a demographic questionnaire and
were debriefed. The experiment took approximately 40 min-
Task performance. Task performance was operationalized
as the total score earned during each 60-second performance
trials (total points gained minus total points lost). The experi-
menter tallied the scores and provided the feedback to partici-
Self-set performance goals. Performance goals were meas-
ured prior to trial 2, trial 3, and trial 4. Participants verbally
expressed the goals and wrote them down.
Goal change. Similar to Donovan and Williams (2003) and
Williams et al. (2000), goal revision was operationalized as the
point difference between the participant’s revised performance
goal and the previous performance goal. Positive goal revision
indicated that the goal was made more difficult, whereas nega-
tive scores indicated that the individuals lowered their goals.
This was calculated for trial 3 (goal set trial 3 - goal set trial 2)
and trial 4 (goal set trial 4 - goal set trial 3).
Positive discrepancy creation. Similar to Donovan and Ha-
festeinsson (2006), PDC was operationalized as the point dif-
ference between the participant’s current performance and the
goal set for the next trial. Positive scores indicated that indi-
viduals raised their performance goal above their current per-
formance; negative scores indicated that individuals lowered
their score performance scores below their current performance.
PDC was calculated for trials 2 (goal trial 2 - performance trial
1), 3 (goal trial 3 - performance trial 2), and 4 (goal trial 4 -
performance trial 3).
Task persistence. Persistence was operationalized as whether
the participants chose to play OPERATION for a fourth trial or
withdraw from the task to do a puzzle.
Goal performance discrepancies. Similar to Donovan and
Williams (2003) and Williams et al. (2000), GPD was opera-
tionalized as the point difference between the individual’s score
for a trial and the goal for that trial. Positive discrepancy scores
indicate performance above one’s goal level, whereas negative
discrepancy scores indicate performance below one’s goal level.
It was calculated for trial 2 (performance trial 2 - goal trial 2)
and trial 3 (performance trial 3 - goal trial 3).
Self-efficacy. Self-efficacy for task performance was meas-
ured before each trial using a six-item measure on a 5-point
scale (1 = strongly disagree; 3 = neutral; 5= strongly agree)
adapted from previous self-regulation research (Bandura, 1991;
Donovan & Haffenstein, 2006; Donovan & Williams, 2002). A
sample item was: I feel confident in my ability to perform well
at this task.
Manipulation checks. The extent to which participants per-
ceived there to be a performance discrepancy after the first trial
was assessed by asking participants to report where they fell
relative to average performance on an 11-point scale
(1—extremely far below average; 6—average; 11—extremely
far above average). This measure was also used as a proxy
measure of attention to the task and feedback comprehension. If
a participant did not respond in accordance with the intended
manipulation, the participant was dropped from the study (n =
A measure of regulatory focus was included to determine if
differences existed between participants in the promotion and
prevention focus frames and if they understood the directions.
Participants were asked: “1.) During your trials, to what extent
was your primary focus to minimize the number of mistakes
made while playing OPERATION?” and “2.) During your trials,
to what extent was your primary focus to maximize the number
of pieces removed while playing OPERATION? (1 = to a very
little extent to 10 = to a very large extent)”
Control Variables. Chronic regulatory focus was measured
using the 18-item measure developed by Lockwood et al. (2002).
In the pilot studies the Lockwood scale demonstrated better
reliability than the Higgins et al. scale.
Motor control/manual dexterity was assessed prior to the
study task using the Grooved Pegboard Test (Lafayette Instru-
ment Company, 1997). Participants were asked to insert metal
pegs into key-shaped holes on the pegboard in a specific se-
quence. Scores represent time, in seconds, with the dominant
hand and then with the non-dominant hand.
After the fourth trial, participants were asked to report if they
have ever played the game OPERATION as a child and if they
played the game in the past five years. Participants were also
asked to report how much they enjoyed playing the game on a
5-point scale (1 = not at all enjoyable; 3 = neutral; 5 = ex-
tremely enjoyable), how much effort they dedicated to the ex-
periment (1 = no effort at all; 3 = neutral; 5 = much effort), and
how anxious they were while engaging in the task (1 = not at
all anxious; 3 = neutral; 5 = extremely anxious). Finally, par-
ticipants reported general demographic information such as
gender, race, and academic standing.
Correlation coefficients, descriptive statistics, and reliability
coefficients can be found in Table 1.
Manipulation Checks
Initial feedback. One-way ANOVA demonstrated that par-
ticipants in the negative feedback group (M = 3.90, SD = 1.04)
perceived their initial score to be significantly (F(1, 295) =
1150.07, p < .01, η2 = .80) lower than those in the positive
feedback group, M = 7.89, SD = .99. Thus, the feedback ma-
nipulation worked as intended.
Regulatory focus. One-way ANOVA demonstrated that
those in the promotion frame condition (M = 7.98, SD = 1.66),
indicated that they “focused on maximizing the number of
pieces removed while playing OPERATION” significantly
more (F(1,295) = 4.44, p < .05) than those in the prevention
frame condition (M = 7.56, SD = 1.74); however, the strength
of the effect was small, η2 =.02. A second one-way ANOVA
showed that there was not a significant difference between the
promotion (M = 7.15, SD = 1.98) and prevention frame (M =
6.93, SD = 2.04) for the manipulation check item “to what ex-
tent was your primary focus to minimize the number of errors
made while playing OPERATION?” These findings should be
kept in mind when interpreting the results.
Tests of H ypotheses
Hypothesis 1. As can be seen in Table 1, normative feed-
back direction was significantly (p <.01) related to PDC, r =
.29. Individuals who missed the average created a larger PDC
for trial 2 (M = 47.35, SD = 29.23) by setting a higher score
relative to their performance than the participants who ex-
ceeded the average, M = 28.03, SD = 34.76. Likewise, per-
formance feedback (GPD) was strongly correlated with PDC
for trials 3 (r = .78, p <.01) and trials 4 (r = .75, p < .01),
indicating that as GPD became larger and positive, individuals
engaged in less PDC. This supports Hypothesis 1.
Hypothesis 2. Supporting hypothesis 2, a significant positive
correlation (r =.79, p <.01) between GPD for trial 2 and goal
revision for trial 3 was found. Similarly, GPD for trial 3 was
positively correlated (r =.73, p <.001) with goal revision for
trial 4.
Hypothesis 3. Correlation analysis (r = .03, ns) and logistic
regression demonstrated that trial 3 GPD was not a significant
predictor of task choice on trial 4, Wald =.31, SE =.002, ns.
The distribution of GPDs was similar for those individuals who
chose the puzzle (n = 120, M = 1.50, SD = 75.61) and who
chose to play OPERATION again, n = 177, M = 3.50, SD =
Hypothesis 4. Table 1 demonstrates that regulatory focus
frame was not related to goal revision for trial 3 (r = .05, ns) or
for trial 4, r =.03, ns. Likewise, regulatory focus frame was not
related to PDC for any of the trials. As shown in Tables 2 and 3,
multiple regression analyses demonstrated that regulatory focus
frame was not a significant predictor of goal revision for trials 3
or 4, respectively. Likewise, no interactions were found be-
tween regulatory focus and GPD on goal revision. Thus, Hy-
pothesis 4 was not supported. It should be noted that the analy-
ses were reexamined with chronic regulatory focus included as
a control variable and the results were unchanged.
Hypothesis 5. Table 1 demonstrates that self-efficacy for tri-
al 3 was significantly correlated with goal revision for trial 3 (r
=.39, p <.01) and self-efficacy for trial 4 was significantly cor-
related with goal revision for trial 4, r =.25, p <.01. Thus, as
one’s self-efficacy increased so did upward goal revision. This
supports Hypothesis 5.
As shown in Tables 4 and 5, multiple regression analysis
demonstrated that self-efficacy was a significant predictor of
trial 3 goal revision (β =.12, p <.01) independent of trial 2 GPD
(β =.76, p <.01). Likewise, self-efficacy was a significant pre-
dictor of trial 4 goal revision (β =.17, p <.01) independent of
trial 3 GPD, β =.30, p <.01. This provides further support for
hypothesis 5.
As demonstrated in Table 4 an interaction between trial 3
self-efficacy and GPD for trial 2 was found, β =.52, p <.01.
When GPD was large and positive, those with higher levels of
self-efficacy engaged in more upward goal revision than those
with lower levels of self-efficacy. When GPD was negative,
self-efficacy did not impact goal revision (see Figure 1). This
interaction effect was not found for trial 4.
Relationship between Goals and Performance over
Repeated measures ANOVA demonstrated a significant in-
crease in both performance (F (3, 528) = 56.95, p <.001, par-
tialη2 =.24) and goal level (F (2, 590) = 91.68, p <.001, partial
η2 =.24) over trials. The average level of self-set goals increased
Table 1.
Summary of intercorrelations, means, standard deviations, and scale reliabilities for study variables.
Measure M SD 1 2 3 4 5 6 7 8 9
1. Chronic RFpro 7.45 .90 (.80)
2. Chronic RFpre 5.46 1.26 .18** (.74)
3. Dexterity 1.07 .10 .05 .03 —
4. Feedback .51 .50 .16** .06 .11 —
5. RF manip .50 .50 .02 .04 .05 .00 —
6. SE1 3.50 .68 .13* .11 .14* .03 .05 (.90)
7. ScoreT1 68.67 77.85 .05 .04 .30** .02 .00 .06 —
8. GoalT2 106.26 73.87 .09 .03 .28** .11 .05 .17**.90** —
9. PDC2 37.59 33.51 .07 .02 .09 .29** .10 .25**.33** .11 —
10. SE2 3.50 .66 .22** .06 .18** .39** .03 .50**.30** .33**.05
11. ScoreT2 108.54 80.00 .07 .06 .21** .01 .05 .08 .43** .47**.02
12. GPDT2 2.27 79.54 .01 .10 .05 .09 .01 .08 .40** .46** .07
13. GoalT3 133.11 72.97 .09 .01 .29** .09 .08 .16**.66** .76**.15**
14. PDC3 24.58 49.90 .03 .11 .08 .11 .03 .11 .26** .36**.18**
15. Goal Ch 3 26.85 50.91 .01 .04 .01 .03 .05 .02 .37** .36**.07
16. SE3 3.63 .74 .13* .07 .08 .24** .03 .40**.03 .01 .09
17. ScoreT3 131.63 77.59 .02 .00 .26** .12* .03 .11 .43** .45**.00
18. GPDT3 1.48 75.77 .11 .01 .01 .03 .05 .05 .20** .27** .15**
19. GoalT4 148.80 74.24 .01 .02 .30** .11 .03 .21* .56** .65**.14*
20. PDCT4 17.13 52.10 .01 .03 .03 .14 .00 .14* .16** .26**.20**
21. Goal Ch 4 16.08 49.98 .17** .05 .02 .05 .07 .08 .13* .15** .03
22. SE4 3.74 .67 .14* .08 .02 .13* .02 .43**.00 .03 .06
23. Task choice .60 .49 .02 .11 .09 .02 .05 .06 .02 .01 .03
24. Experience1 1.33 .47 .10 .15** .07 .03 .04 .07 .16** .12* .10
25. Experience2 1.91 .31 .06 .04 .07 .01 .01 .05 .06 .04 .05
26. Effort 3.94 .67 .17** .03 .14* .10 .05 .14* .07 .03 .09
27. Anxiety 2.51 .99 .17** .13* .05 .04 .03 .03 .03 .03 .01
28. Enjoyment 4.17 .75 .03 .01 .02 .07 .13* .10 .00 .02 .04
29. Gender 1.49 .50 .14 .12* .25** .12* .06 .08 .03 .01 .08
30. Age 18.71 1.39 .06 .13* .05 .04 .00 .09 .07 .05 .05
31. Ethnicity 1.77 1.24 0.02 0.01 0.13 0.09 0.02 0.08 0.12* 0.06 0.15**
Measure 10 11 12 13 14 15 16 17 18 19 20
1. Chronic RFpro
2. Chronic RFpre
3. Dexterity
4. Feedback
5. RF manip
6. SE1
7. ScoreT1
8. GoalT2
9. PDC2
10. SE2 (.90)
11. ScoreT2 .08 —
12. GPDT2 .23** .57** —
13. GoalT3 .21** .79** .09 —
14. PDC3 .19** .45** .78** .19** —
15. Goal Ch 3 .18** .46** .79** .33** .25** —
16. SE3 .49** .37** .36** .28** .17** .39** (.93)
17. ScoreT3 .02 .50** .08 .50** .07 .06 .02 —
18. GPDT3 .18** .25** .00 .46** .26** .26** .25** .55** —
19. GoalT4 .22** .63** .03 .77** .11* .16** .20** .77** .05 —
20. PDCT4 .27** .15** .08 .35** .27** .14* .25** .40** .75** .28** —
21. Goal Ch 4 .00 .22** .08 .31** .11 .23** .12* .41** .73** .37** .09
22. SE4 .45** .13* .11 .11 .06 .11 .55** .28** .18** .27** .03
23. Task choice .09 .01 .01 .00 .02 .01 .01 .03 .03 .01 .03
24. Experience1 .09 .09 .02 .11 .01 .03 .09 .09 .01 .10 .01
25. Experience2 .02 .03 .00 .02 .02 .02 .01 .02 .05 .02 .02
26. Effort .13* .07 .05 .08 .00 .08 .18** .09 .18** .02 .12*
27. Anxiety .12* .03 .06 .05 .11 .02 .03 .13* .18* .01 .21**
28. Enjoyment .18** .08 .06 .04 .07 .02 .15** .08 .04 .09 .02
29. Gender .09 .04 .05 .00 .06 .02 .05 .02 .02 .07 .12*
30. Age .03 .07 .02 .02 .09 .06 .04 .01 .01 .02 .03
31. Ethnicity .02 .06 .01 .05 .03 .01 .06 .09 .04 .07 .03
Measure 21 22 23 24 25 26 27 28 29 30 31
1. Chronic RFpro
2. Chronic RFpre
3. Dexterity
4. Feedback
5. RF manip
6. SE1
7. ScoreT1
8. GoalT2
9. PDC2
10. SE2
11. ScoreT2
12. GPDT2
13. GoalT3
14. PDC3
15. Goal Ch 3
16. SE3
17. ScoreT3
18. GPDT3
19. GoalT4
20. PDCT4
21. Goal Ch 4 —
22. SE4 .25** (.91)
23. Task choice .03 .05 —
24. Experience1 .00 .09 .06 —
25. Experience2 .08 .01 .06 .17** —
26. Effort .14* .20** .00 .00 .04 —
27. Anxiety .05 .06 .00 .03 .03 .28** —
28. Enjoyment .08 .29** .02 .02 .03 .28** .20** —
29. Gender .09 .16** .00 .04 .04 .04* .04 .03 —
30. Age .04 .02 .06 .12* .03 .04* .03 .06 .07 —
31. Ethnicity .04 .01 .04 .40** .09 .01 .04 .00 .05 .12* —
N = 297. * p < .05 ** p < .01. Scale reliabilities are on the diagonals in (): 1. Chronic RFpro = 9 items from Lockwood et al. (2002) on a 9 point scale (1 = not at all true of
me to 9 = very true of me). 2. Chronic RFpre = 9 items from Lockwood et al. (2002) on a 9 point scale (1 = not at all true of me to 9 = very true of me). 3. Dexterity =
finger dexterity using the Grooved Pegboard Test with dominant hand. 4. Feedback = feedback manipulation (0 = negative, 1 = positive). 5. RFmanip = regulatory focus
manipulation (0 = prevention, 1= promotion). 6. SE1 = Self-efficacy trial 1 (composite of six items 1—strongly disagree to 5—strongly agree). 7. Score T1 = total score
trial 1. 8. GoalT2 = goal set for trial 2. 9. PDC2 = positive discrepancy creation for trial 2 (goal trial 2 - performance trial 1). 10. SE2 = self-efficacy trial 2. 11. Score T2 =
total score trial 2. 12. GPDT2: goal performance discrepancy trial 2 (performance trial 2 - goal trial 2). 13. GoalT3 = goal set trial 3. 14. PDC3 = positive discrepancy
creation for trial 3 (goal trial 3 - performance trial 2). 15. Goal Ch 3 = goal change/revision for trial 3 (goal time 3 - goal time 2). 16. SE3 = self efficacy trial 3. 17. Sco-
reT3 = total score trial 3. 18. GPDT3 = goal performance discrepancy trial 3 (performance trial 3 - goal trial 3). 19. GoalT4 = goal set for trial 4. 20. PDCT4 = positive
discrepancy creation for trial 4 (goal trial 4 - performance trial 3). 21. Goal Ch 4 = goal change/revision for trial 4 (goal time 4 - goal time 3). 22. SE4 = self-efficacy trial 4.
23. Task choice: choice on last trial (0 = puzzle, 1 = operation). 24. Experience 1: Did you play operation as a child? (0 = no, 1 = yes). 25. Experience 2: have you played
operation in the past 5 years? (0 = no, 1 = yes). 26. Effort: 1 = no effort at all; 5 = much effort. 27. Anxious: 1 = not at all anxious; 3 = neutral; 5 = extremely anxious. 28.
Enjoyment: 1 = not at all enjoyable; 3 = neutral; 5 = extremely enjoyable. 29. Gender (0 = female, 1 = male). 30. Age. 31. Ethnicity = (0 = not Caucasian, 1 = Caucasian).
in a linear fashion from 106.26 (trial 2 goal) to 148.79 (trial 4
goal). Similarly, average performance increased in a linear
fashion from 68.67 (trial 1) to 146.10 (trial 4). All changes
across trials were significant (p < .01), demonstrating effective
self-regulation of performance. Participants set increasingly
difficult goals and were able to match performance to their
Table 2.
Hypothesis 4: regression analyses for regulatory focus on goal revision trial 3.
Dependent Variable: Goal Revision Trial 3 (goal T3 - goal T2)
Model Variable β SE t-value R2 ΔR2
1 Dexterity 0.00 0.03 0.05 .01 .01
Experience 0.02 9.57 0.26
Effort 0.08 4.80 1.21
Enjoyment 0.00 4.19 0.05
Anxiety 0.01 3.16 0.09
2 Dexterity 0.04 0.00 1.12 .64 .64**
Experience 0.02 5.77 0.59
Effort 0.03 2.91 0.84
Enjoyment 0.05 2.56 1.35
Anxiety 0.06 1.92 1.67
RF 0.05 3.62 1.42
GPD2 0.80 0.02 22.55**
3 Dexterity 0.04 0.00 1.12 .64 .00
Experience 0.02 5.78 0.59
Effort 0.03 2.91 0.84
Enjoyment 0.05 2.58 1.34
Anxiety 0.06 3.64 1.68
RF 0.05 3.67 1.41
GPD2 0.80 0.03 15.40**
GPD2 X RF 0.00 0.05 0.05
N = 256 * p < .05, ** p < .01.
Table 3.
Hypothesis 4: regression analyses for regulatory focus on goal revision trial 4.
Dependent Variable: Goal Revision Trial 4 (goal T4 – goal T3)
Model Variable β SE t R2 ΔR2
1 Dexterity 0.06 0.01 1.08 .05 .05
Experience 0.09 9.34 1.51
Effort 0.19 4.61 3.02**
Enjoyment 0.14 4.03 2.28*
Anxiety 0.03 3.04 0.42
2 Dexterity 0.04 0.00 0.94 .54 .50**
Experience 0.04 6.51 1.02
Effort 0.07 3.28 1.48
Enjoyment 0.06 2.85 1.32
Anxiety 0.08 2.14 1.98*
RF 0.04 4.04 1.05
GPD3 0.72 0.03 17.50**
3 Dexterity 0.04 0.00 0.95 .54 .00
Experience 0.04 6.53 0.99
Effort 0.07 3.27 1.48
Enjoyment 0.06 2.87 1.28
Anxiety 0.08 2.14 1.97*
RF 0.04 4.05 1.04
GPD3 0.71 0.04 11.36**
GPD3 X RF 0.02 0.05 0.36
N = 256 * p < .05, ** p < .01.
Table 4.
Hypothesis 5: regression analyses for self-efficacy on goal revision trial 3.
Dependent Variable: Goal Revision Trial 3 (goal T3 - goal T2)
Model Variable β SE t R2 ΔR2
1 Dexterity 0.00 0.01 0.00 .01 .01
Experience 0.02 9.72 0.36
Effort 0.08 4.81 1.21
Enjoyment 0.00 4.19 0.04
Anxiety 0.01 3.18 0.14
2 Dexterity 0.03 0.00 0.77 .65 .65**
Experience 0.02 5.78 0.53
Effort 0.01 2.89 0.32
Enjoyment 0.06 2.51 1.49
Anxiety 0.06 1.90 1.74
SE3 0.12 0.44 3.22**
GPD2 0.76 0.02 20.16**
3 Dexterity 0.02 0.00 0.64 .66 .01**
Experience 0.02 5.71 0.66
Effort 0.01 2.85 0.17
Enjoyment 0.04 2.49 1.13
Anxiety 0.06 1.87 1.57
SE3 0.16 0.46 4.09**
GPD2 0.23 0.11 1.34
SE3 X GPD2 0.52 0.01 3.08**
N = 256 * p < .05, ** p < .01.
Table 5.
Hypothesis 5: Regression analyses for self-efficacy on goal revision trial 4.
Dependent Variable: Goal Revision Trial 4 (goal T4 - goal T3)
Model Variable β SE t R2 ΔR2
1 Dexterity 0.01 0.01 0.09 .01 .01
Experience 0.01 9.51 0.23
Effort 0.08 4.77 1.22
Enjoyment 0.01 4.19 0.17
Anxiety 0.01 3.15 0.16
2 Dexterity 0.01 0.01 0.02 .10 .09**
Experience 0.03 9.12 0.56
Effort 0.01 4.69 0.07
Enjoyment 0.00 4.15 0.07
Anxiety 0.05 3.05 0.86
SE4 0.17 0.76 2.77**
GPD3 0.30 0.04 5.01**
3 Dexterity 0.00 0.01 0.03 .10 .01
Experience 0.03 9.13 0.57
Effort 0.00 4.75 0.01
Enjoyment 0.01 4.16 0.09
Anxiety 0.05 3.08 0.78
SE4 0.18 0.79 2.82**
GPD3 0.48 0.22 1.43
SE4 X GPD3 0.18 0.01 0.55
N = 256 * p < .05, ** p < .01.
eg (-2SD)
eg (-1SD)MeanPos (+1SD)Pos (+2SD)
Goal Revisio
Low SE (-2SD)
Low SE (-1SD)
Med SE (Mean)
High SE (+1SD)
High SE (+2SD)
Figure 1.
Trial 2 GPD X trial 3 self-efficacy on goal revision. Interactive effect of trial 2 GPD (goal set – minus performance) and self-efficacy for trial 3 on
trail 3 goal revision. GPD: M = 2.27, SD = 79.54; SE: M = 21.77, SD = 4.43. As per Jaccard (1990, p. 1990) effects were evaluated at low, medium,
and high by calculating one SD above and below the mean for each variable. For further examination 2 SD above and below each mean was also
Trial 1Trial 2Trial 3Trial 4
P erformance
Figure 2.
Performance and goal change over time. All performance changes and goal changes across trials were significant at p < .001. Values on the axis
represent mean performance score.
goals (see Figure 2). Finally, traditional goal setting effects
were tested by regressing performance on goal level, control-
ling for previous performance, self-efficacy, and finger dexter-
ity for the last 3 trials. For trial 2 (β = .44), trial 3 (β =.23) and
trial 4 (β = .39) goal difficulty was a significant (p < .01) pre-
dictor of performance.
Chronic Reg ulatory Foc u s
Finally, we conducted a factor analysis on the combined
samples from the two pilot studies and the main study, N = 454
for the Lockwood et al. (2002) promotion/prevention measure.
Given that a sample size of over 400 is very good for factor
analysis (Comrey, 1992) this may be a better assessment of the
Lockwood et al. measure. The combined sample resulted in
acceptable levels of reliability: αpre = .73 and αpro = .81. Ex-
ploratory factor analysis with oblimin rotation of the two sub-
scales resulted in three Eigenvalues that exceeded 1.00 and
three bends in the scree plot.
Summary of Findings
This study sought to examine the effects of discrepancy
feedback information from two sources: normative standards
and self-set goals on goal revision processes on a novel task.
The results demonstrate that people are able to manage dis-
crepancy creation and reduction processes to boost performance
over trials on a novel task. When presented with normative
feedback, the vast majority of participants set performance
goals higher than their initial performance, and those given
negative feedback created higher discrepancies than those given
positive feedback. Discrepancy creation was also observed
across later trials, with the size of PDC inversely related to goal
success. These findings provide strong evidence that people
have a tendency to challenge themselves by setting goals above
current levels of performance, and that this tendency is greater
when they are not performing up to their own or normative
Further, personal goal discrepancy feedback was a strong
predictor of goal revision, even more so than normative dis-
crepancy feedback. Individuals surpassing their goals consis-
tently raised their goals relative to the size of the positive dis-
crepancy, and individuals missing their goals lowered their
goals relative to the size of the negative discrepancy. In general,
however, participants kept goals higher than previous levels of
performance. These findings complement previous research
(Donovan & Hafesteinsson, 2006; Donovan & Williams, 2003)
and support the main premises of SCT: individuals engage in a
dual-cyclic process of discrepancy production and reduction
and respond to performance feedback by revising goals ac-
Results also suggest that self-efficacy is a consistent inde-
pendent predictor of goal revision, and at times interacts with
GPD to affect goal revision. Consistent with SCT, goal aspira-
tions are in part a function of people’s belief in their capability
to perform a task. It should be noted, however that the effects of
self-efficacy were not consistent across trials. Although self-
efficacy independently predicted goal change for both trials 3
and 4, an interaction was observed for trial 3 goal revision only.
This may be due to the fact participants did not develop a
strong sense of self-efficacy regarding the task, either because it
was novel or they had not played it in a long time. Future re-
search should continue to examine the interactive effect of
self-efficacy and feedback.
Contrary to hypothesis 3, task persistence was not related to
the size of GPD. Task choice on trial 4 was not correlated with
GPD or any variables of interest. Participants may have chosen
to withdrawal from the task for reasons not related to the task
itself or because fatigue had set in. Although free choice meas-
ures of motivation and persistence are common, few research-
ers have bothered to ask participants why they chose the par-
ticular task they did. By simply asking participants why they
chose a particular task, researchers may gain a better under-
standing of reasons for task persistence and withdrawal.
Perhaps granting individuals the option to set and revise their
own goals, as was examined in the present study, reduces the
likelihood of withdrawing from the task (i.e., the task can al-
ways be made easier or harder as per the discretion of the indi-
vidual). If participants are given assigned goals and large dis-
crepancies exist, then it is possible they would be more likely to
withdrawal from the task, because they would not have control
over revising the goals upward or downward. Future research
should examine how task persistence and withdrawal are influ-
enced by the presence of assigned goals versus self-set goals,
and the factors influencing task persistence.
Finally, when measured as both a chronic trait variable and
an induced situational variable, regulatory focus did not predict
PDC or goal revision. This is surprising given that several stu-
dies have shown regulatory focus to interact with feedback
information to produce an effect on future performance (e.g.,
Idson & Higgins, 1997; Spiegel & Higgins, 2001). Although
the manipulations were modeled after previous research, the
results of the manipulation checks demonstrated that partici-
pants may not have understood the desired focus. Higgins et al.
(1994) demonstrated that promotion regulation is more associ-
ated with approach motivation, whereas prevention regulation
is more associated with avoidance motivation, and that motiva-
tion is stronger when focus matches the end-state (Higgins,
2000). It is possible that the task directions did not provide a
strong enough distinction between approach and avoidance
motivation or perhaps since the prevention group did not begin
with any points, they did not feel like there was something to be
lost. Future studies should ensure that the prevention group has
something to lose, and the promotion group has something to
Future research should also examine if regulatory focus is
only a relevant predictor of goal setting in specific contexts.
Previous research has utilized tasks indicative of intelligence,
such as essay writing (Freitas et al., 2002), evaluations of life
events (Higgins et al., 1994), and anagram tasks (Forster, Grant,
Chen Idson, & Higgins, 2001; Idson & Higgins, 2000; Spiegel
& Higgins, 2001). Thus, another explanation for the lack of
support for regulatory focus as a predictor of goal revision and
performance in the present study is that the individuals’
self-concepts may not have been threatened. Performance on
OPERATION may not be relevant to people; whereas, success
on an anagram task provides information about intellectual
ability. Perhaps, regulatory focus only matters when people
engage in tasks that provide self- relevant feedback.
Finally, the findings from the three samples demonstrate lit-
tle support for the two distinct dimensions of promotion and
prevention. The present research failed to show consistently
high reliabilities and validity evidence for Higgins’ measure or
Lockwood’s measure of regulatory focus. Likewise, the ma-
nipulation of prevention focus proved to be difficult in the main
study. Researchers should consider newly developed measures
of regulatory focus (e.g., Regulatory Focus Scale: Fellner, Hol-
ler, Kirchler, & Schabmann, 2007) or may want to revisit the
regulatory focus construct all together. Perhaps the distinction
between regulatory focus and goal orientation should be re-
examined. Alternatively, promotion focus may simply be a
component of achievement motivation.
Limitations and Directions for Future Research
As with any study, there are limitations that should be noted.
First, the sample under examination in the present study was
comprised of college level students, engaging in a novel task
for course credit. Future research should examine how the
GPD-goal revision relationship operates in different contexts
and under varying circumstances. For instance, Schmidt and
DeShon (2007) examined performance in the face of multiple
tasks. Future research may want to extend these findings by
examining how individuals self-regulate in the presence of
multiple goals. This would be valuable, because in more natu-
ralistic settings people frequently have more than one task or
goal to pursue.
Although support was found for SCT in terms of goal revi-
sion as a function of feedback information, no support was
found for regulatory focus. This may have been mostly due to
methodological issues surrounding the measurement and ma-
nipulation of regulatory focus. Despite the attempts to pilot
these measures, future research should re-examine if regulatory
focus impacts goal revision using clearer measures.
Implications for Theory and Prac tice
This research lends converging support that individuals en-
gage in a dual-cyclic process of discrepancy production and
discrepancy reduction, and that self-efficacy may be an impor-
tant moderator of this relationship. This study extends current
self-regulation research by demonstrating that in the face of a
novel task, discrepancy feedback originating first from norma-
tive standards and then from self-set goals, is a strong predictor
of goal revision and performance. Above and beyond all other
variables examined, feedback information was the most pow-
erful determinant of future goals and performance. This was
consistent across the four trials.
Although this study did not find that regulatory focus moder-
ates this relationship, these findings can act as a guide for re-
search and theory development. Researchers should continue to
examine cognitive and dispositional variables that impact goal
revision processes in order to develop a comprehensive model
of motivation. One potential avenue for future research is to
re-examine the influence of regulatory focus on goal revision
with more refined and reliable measurement. Given that the
present study demonstrated difficulties measuring and manipu-
lating regulatory focus, future research should examine the
impact, if any; regulatory focus has on self-regulation. Perhaps,
regulatory focus does not have a substantial impact on goal
revision processes, or does not add any predictive power over
goal orientation (Dweck, 1986; Elliot & Friedman, 2007). Re-
searchers should continue to examine the effects of regulatory
focus and goal orientation on goal revision and performance in
achievement contexts.
Finally, a thorough understanding of factors that influence
self-directed motivation is important, especially given the
unique challenges of today’s workplace stemming from global-
ization, telecommuting, self-employment, dual-income house-
holds, and unemployment. These workplace changes require
greater self-monitoring and regulation on the part of workers.
Thus, it is important for both individuals and organizations to
be aware of the powerful effects goals have on performance,
and how the dual-cyclic process of self-regulation facilitates
effective performance management. Helping individuals self-
regulate, by encouraging goal-setting and performance moni-
toring, may be an effective way for organizations to get the
most out of their employees, and for individuals to have a sat-
isfying and rewarding work experience. Likewise, managers
should make attempts to increase the self-efficacy of employees.
Facilitation of “you can do” perceptions, may enhance “I will
do” outcomes among employees. Similarly, these findings can
also apply to other contexts such as academics (e.g., grades),
health (e.g., diet and exercise), finances (e.g., paying bills), and
family (e.g., organizations and chores), among others.
The present research demonstrates that goal revision and
subsequent performance largely rely on performance feedback
from normative standards and self-set goals. Individuals use
information about their current performance levels compared to
desired performance levels to modify goals. Individuals are
further likely to increase goals beyond their current levels of
performance, especially when performance is not meeting ex-
pectations. Self-efficacy is a consistent independent predictor of
goal revision as well. This research provides support for a
dual-cyclic model of discrepancy production and reduction and
highlights the importance of self-efficacy. These findings have
a practical utility in broad range of contexts (e.g., work, school,
sports, etc.).
We thank Sylvia Roch, Monica Rodriguez, and the anony-
mous reviewers at Psychology for their helpful comments. We
also thank Dr. Joseph Voelker, Dean of the College of Arts and
Sciences at the University of Hartford, for funding publication
Bandura, A. (1986). Social foundations of thought and action: A social
cognitive theory. Englewood Cliffs, NJ: Prentice Hall.
Bandura, A. (1989). Self-regulation of motivation and action through
internal standards and goal systems. In L. A. Pervin (Ed.), Goal con-
cepts in personality and social psychology (pp. 19-85). Hillsdale, NJ:
Lawrence Erlbaum Associates.
Bandura, A. (1991). Social cognitive theory of self-regulation. Organ-
izational behavior and Human Decision Processes, 50, 248-287.
Bandura, A. (2003). Social cognitive theory: An agentic perspective.
Annual review of Psychology, 52, 1-26.
Bandura, A., & Cervone, D. (1986). Differential engagement of
self-reactive influences in cognitive motivation. Organizational Be-
havior and Human Decision Processes, 38, 92-113.
Bandura, A., & Locke, E. A. (2003). Negative self-efficacy and goal
effects revisited. Journal of Applied Psychology, 88, 87-99.
Comrey, A. L. (1992). A first course in factor analysis. (2nd ed.).
Hillsdale, NJ: LEA, Publishers.
Donovan, J. J., & Hafsteinsson, L. G. (2006). The impact of
goal-performance discrepancies self-efficacy, and goal orientation on
upward goal revision. Jounral of Applied Social Psychology, 36,
2006. doi:10.1111/j.0021-9029.2006.00054.x
Donovan, J. J., & Williams, K. J. (2003). Missing the mark: Effects of
time and causal attributions on goal revision in response to goal-
performance discrepancies. Journal of Applied Psychology, 88,
379-390. doi:10.1037/0021-9010.88.3.379
Dweck, C. S. (1986). Motivational processes affecting learning. American
Psychologist, 41, 1040-1048.
Elliot, A. J., & Freidman, R. (2007). Approach-avoidance: A central
characteristic of personal goals. In B. R. Little, K. Salmera-Aro and
S. D. Phillips (Eds.), Personal Project Pursuit: Goals, Action, and
Human Flourishing (pp. 97-118). Mahwah, NJ: Lawrence Erlbaum
Fellner, B., Holler, M., Kirchler, E., & Scabmann, A. (2007). Regula-
tory Focus Scale (RFS): Development of a scale to record dispotional
regulatory focus. Swiss Journal of Psychology, 66, 109-116.
Forster, J., Grant, H., Chen Idson, L., & Higgins, T. E. (2001). Suc-
cess/failure feedback, expectancies, and approach/avoidance motiva-
tion: How regulatory focus moderates classic relations. Journal of
Experimental Social Psychology, 37, 253-260.
Heath, C., Larrick, R. P., & Wu, G. (1999). Goals as reference points.
Cognitive Psychology, 38, 79-109. doi:10.1006/cogp.1998.0708
Higgins, T. E. (1997). Beyond pleasure and pain. American Psycholo-
gist, 52, 1280-1300. doi:10.1037/0003-066X.52.12.1280
Higgins, T. E. (2000). Making a good decision: Value from fit. Ameri-
can Psychologist, 55, 1215-1230.
Higgins, T. E., Friedman, R. S., Harlow, R. E., Chen Idson, L., Aykuk,
O. N., & Taylor, A. (2001). Achievement orientations from subjec-
tive histories of success: Promotion pride versus prevention pride.
European Journal of Social Psychology, 31, 3-23.
Higgins, T. E., Roney, C., Crowe, E., & Hymes, C. (1994). Ideal versus
ought predilections for approach and avoidance: Distinct
self-regulatory systems. Journal of Personality and Social Psychol-
ogy, 66, 276-286. doi:10.1037/0022-3514.66.2.276
Idson, L. C., & Higgins, E. T. (2000). Howe current feedback and
chronic effectiveness influence motivation: Everything to gain versus
everything to lose. European Journal of Social Psychology, 30,
Karoly, P. (1993). Mechanisms of self-regulation: A systems view.
Annual Review of Psychology, 44, 23-52.
Locke, E. A., & Latham, G. P. (1984). Goal setting: A motivational
technique that works!. Englewood Cliffs, NJ: Prentice Hall.
Locke, E. A., & Latham, G. P. (1990). A theory of goal setting and task
performance. Englewood Cliffs, NJ: Prentice Hall.
Locke, E. A., & Latham, G. P. (2002). Building a practically useful
theory of goal setting and task motivation: A 35-year odyssey.
American Psychologist, 57, 705-717.
Lockwood, P., Jordan, C. H., & Kunda, Z. (2002). Motivation by posi-
tive or negative role models: regulatory focus determines who will
best inspire us. Journal of Personality and Social Psychology, 83,
854-864. doi:10.1037/0022-3514.83.4.854
Phillips, J. M., Hollenbeck, J. R., & Ilgen, D. R. (1996). Prevalence and
prediction of positive discrepancy creation: Examining a discrepancy
between two self-regulation theories. Journal of Applied Psychology,
81, 498-511. doi:10.1037/0021-9010.81.5.498
Shah, J., & Higins T. E. (1997). Expectancy X value effects: Regu-
latory focus as a determinant of magnitude and direction. Journal of
Personality and Social Psychology, 73, 447-458.
Spiegel, S., & Higgins, T. E. (2001). Regulatory focus and means sub-
stitution in strategic task performance. Unpublished manuscript,
Columbia University.
Tabachnick, B. G., & Fidell, L. S. (2001). Using multivariate statistics
(4th ed.). Needman Heights, MA: Allyn & Bacon.
Tversky, A., & Kahneman, D. (1992). Advances in prospect theory:
Cumulative representations of uncertainty. Journal of Risk and Un-
certainty, 5, 297-323. doi:10.1007/BF00122574
Van-Dijk, D., & Kluger, A. N. (2004). Feedback sign effect on motiva-
tion: Is it moderated by regulatory focus?. Applied Psychology: An
International Review, 53, 113-135.
Williams, K. J., Donovan, J. J., & Dodge, T. L. (2000). Self-regulation
of performance. Goal establishment and goal revision processes in
athletes. Human Performance, 13, 159-180.
Wood, R., & Bandura, A. (1989). Social cognitive theory of organiza-
tional management. Academy of Management Review, 14, 361-384.