See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/233130558

The Effect of Commitment to a Learning Goal, Self-Efficacy, and

the Interaction Between Learning Goal Difficulty and
Commitment on Performance in a Business Simulation

Article  in  Human Performance · July 2011

DOI: 10.1080/08959285.2011.580807

CITATIONS READS

25 520

2 authors:

Gerard Seijts Gary P. Latham

The University of Western Ontario University of Toronto
68 PUBLICATIONS   2,239 CITATIONS    242 PUBLICATIONS   28,484 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Leader Character and subjective well-being View project

the relevance of priming goals in the subconscious on organizational behaviour View project

All content following this page was uploaded by Gary P. Latham on 28 May 2014.

The user has requested enhancement of the downloaded file.

This article was downloaded by: [University of Toronto Libraries]
On: 06 October 2011, At: 14:33
Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Human Performance
Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/hhup20

The Effect of Commitment to a Learning

Goal, Self-Efficacy, and the Interaction
Between Learning Goal Difficulty and
Commitment on Performance in a
Business Simulation
a b
Gerard H. Seijts & Gary P. Latham
a
University of Western Ontario
b
University of Toronto

Available online: 08 Jul 2011

To cite this article: Gerard H. Seijts & Gary P. Latham (2011): The Effect of Commitment to a
Learning Goal, Self-Efficacy, and the Interaction Between Learning Goal Difficulty and Commitment
on Performance in a Business Simulation, Human Performance, 24:3, 189-204

To link to this article: http://dx.doi.org/10.1080/08959285.2011.580807

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation
that the contents will be complete or accurate or up to date. The accuracy of any
instructions, formulae, and drug doses should be independently verified with primary
sources. The publisher shall not be liable for any loss, actions, claims, proceedings,
demand, or costs or damages whatsoever or howsoever caused arising directly or
indirectly in connection with or arising out of the use of this material.
 Human Performance, 24:189–204, 2011
Copyright © Taylor & Francis Group, LLC
ISSN: 0895-9285 print/1532-7043 online
DOI: 10.1080/08959285.2011.580807

The Effect of Commitment to a Learning Goal,

Self-Efficacy, and the Interaction Between Learning Goal
Difficulty and Commitment on Performance in
a Business Simulation
Downloaded by [University of Toronto Libraries] at 14:33 06 October 2011

Gerard H. Seijts
University of Western Ontario

Gary P. Latham
University of Toronto

The effect of commitment to a learning goal, self-efficacy, and the interaction between learning goal
difficulty and goal commitment with performance was investigated using a highly complex business
simulation. Participants (n = 128) needed to acquire knowledge in order to perform the task effec-
tively. The correlation between commitment to the learning goal and performance was positive and
significant (r = .47, p < .001). Commitment was also a moderator of the learning goal–task per-
formance effect. The relationship between self-efficacy and performance was partially mediated by
commitment to the learning goal. Performance was a partial mediator of the relationship between goal
commitment and self-efficacy. Seventy-five percent of the participants self-set a performance goal.
The correlation between self-set performance goals and performance was positive and significant
(r = .31, p < .001).

Crucial to building a theory of goal setting (Latham & Locke, 2007; Locke & Latham, 1990,
2002) is the search for moderator variables or boundary conditions (Locke, 2007). This is because
they alter the strength of the causal relationship between a specific high goal and task performance
by weakening or amplifying the effect. Five moderator variables included in goal-setting theory
are ability, feedback, task complexity, situational constraints, and goal commitment. The latter
variable, defined as an individual’s determination to attain a goal, is arguably the most important.
This is because commitment is the sine qua non of goal setting. Locke and Latham (1990) argued,
“It is virtually axiomatic that a goal that a person is not really trying for is not really a goal and
therefore cannot have much effect on subsequent action” (p. 124).
The mediators in goal-setting theory are choice, effort, persistence, and strategy. Although
cognitive factors play a role in all four mediators, the first three are more motivational in nature
than is the fourth. With regard to a specific high performance goal, where the emphasis is on

Correspondence should be sent to Gerard H. Seijts, University of Western Ontario, Richard Ivey School of Business,
1151 Richmond Street North, London, Ontario, Canada N6A 3K7. E-mail: gseijts@ivey.ca
 190 SEIJTS AND LATHAM

a desired outcome, the choice to exert effort and persist until the goal is attained is central to
the beneficial effect of goal setting on subsequent performance. Strategy is important to the
extent that the performance goal cues the recall of procedures that have proven effective for
goal attainment in the past.
To date, no research has shown whether goal commitment moderates the relationship between
setting a learning goal and performance. This issue is important because the emphasis of a learn-
ing goal is on discovering or mastering, as opposed to merely recalling, appropriate strategies,
processes, or procedures necessary for goal attainment. A focus on learning processes or pro-
cedures for performing effectively is not qualitatively the same as exerting effort and merely
persisting until a desired level of performance has been reached. Thus it is possible that with
Downloaded by [University of Toronto Libraries] at 14:33 06 October 2011

regard to goal commitment, a learning goal may operate differently than a performance goal,
where the latter is dependent primarily on sheer effort and persistence.

ABILITY AND TASK COMPLEXITY

More than 1,000 studies have demonstrated that a specific difficult performance goal
(e.g., increase sales by 15%) increases a person’s performance (Mitchell & Daniels, 2003). In
fact, the relationship is linear (Latham & Locke, 2007). However, this is only the case when a
person has the ability to perform the task effectively. For example, Kanfer and Ackerman (1989)
found that urging Air Force trainees to do their best resulted in higher performance on a simu-
lated air traffic control task than assigning them a specific high performance goal to attain. This
is because the trainees had no prior experience in, or training on, air traffic control. Hence they
had no knowledge or skills to draw upon when performing the task. Moreover, they were not
informed that they should take the time to learn ways to master this complex task before focusing
on performing well.
Subsequent experiments, using complex tasks, have replicated the findings obtained by Kanfer
and Ackerman. Earley, Connolly, and Ekegren (1989) explained that a specific high performance
goal on a task that people lack the ability to perform may make them so anxious to succeed that
they scramble to discover task-relevant strategies in an unsystematic manner therefore impeding
their performance. Mone and Shalley (1995) concluded that, in the absence of ability, focusing
on a specific high performance goal interferes with learning the task-relevant strategies that will
enable people to accelerate their effectiveness. For example, a specific high performance goal
may misdirect the systematic search for task-relevant strategies (e.g., when individuals shift from
one strategy to another on an ad hoc or trial-and-error basis). Therefore, people fail to learn in a
timely fashion the most effective way(s) to perform a task that is complex for them.
Since the publication of that research, there have been several studies that have contrasted the
effect of performance versus learning goals on tasks where people lack the ability to perform
them. A performance goal, as noted earlier, refers to the desired aim or end of an action in terms
of level of performance to be attained on a task (Latham & Locke, 2007; Locke & Latham, 1990);
in contrast, a learning goal refers to discovering and mastering the requisite strategies, processes,
or procedures for performing effectively as opposed to relying on the knowledge and skill one
already possesses (Seijts & Latham, 2005).
At least eight experiments have shown that the setting of a specific learning goal (e.g., discover
seven ways to increase your teaching evaluations) predisposes individuals to take the additional
 LEARNING GOAL 191

time necessary to seek and process the information required for performing effectively on tasks
they lack the ability to perform. Winters and Latham (1996) manipulated task complexity. On a
task that people already had the ability to perform, a specific high-performance goal led to signif-
icantly higher performance than either a specific high-learning or a do-your-best goal. On a task
requiring knowledge acquisition, the authors replicated Kanfer and Ackerman’s (1989) finding.
People who were told to do their best performed significantly better than those with a specific
high performance goal. But those with a specific high-learning goal had significantly higher per-
formance than those with either a do-your-best or a specific high performance goal. Seijts and
Latham (2001) obtained similar findings using a complex class-scheduling task. Drach-Zahavy
and Erez (2002) found that individuals with a learning goal had higher levels of performance than
Downloaded by [University of Toronto Libraries] at 14:33 06 October 2011

those with a performance goal on a highly complex stock market prediction task. Seijts, Latham,
Tasa, and Latham (2004) reported that people with a learning goal attained more market share
on an interactive, computer-based simulation of the cellular telephone industry than those with a
specific high performance goal. Kozlowski and Bell (2006) found that a learning goal improved
both affective and cognitive self-regulatory processes compared to a performance goal. Noel and
Latham (2006) found that participants who had a learning goal kept their simulated businesses
running longer than those who had a performance goal. Latham, Seijts, and Crim (2008) found
that the higher the learning goal, the higher the person’s performance, particularly for those lower
in cognitive ability. Cianci, Klein, and Seijts (2010) reported that individuals assigned a learning
goal experienced less tension and performed better following negative feedback than individuals
assigned a performance goal. These studies consistently show that a learning goal rather than a
performance goal should be set when people have yet to acquire the ability to master a task.

GOAL COMMITMENT

As previously stated, a core premise of goal-setting theory is that goal commitment is the sine
qua non of the goal–performance relationship (Latham & Locke, 2007; Locke & Latham, 1990,
2002). Thus, consistent with the research on performance goals, we hypothesized that commit-
ment to a learning goal has a direct positive effect on performance of a task that requires the
acquisition of knowledge. Goal-setting theory also states that commitment moderates the effect
of goals on performance when goal level varies among individuals. Yet, no study to date has
empirically shown that goal commitment is a moderator of the effect of a learning goal on per-
formance. It is important to do so because it may be that performance goals activate more lower
level cognitive functions (e.g., attention, memory, and comprehension) in comparison to learning
goals (e.g., analysis and evaluation, planning, and self-monitoring). Learning goals may activate
a wider range of cognitive functions than performance goals (e.g., figuring out how to perform a
task through trial and error, and changing strategies as needed; Anderson, 1985). The extent to
which higher level cognitive functions are activated by learning goals is based on the extent to
which a task is truly complex for an individual.
Several of the tasks used in earlier experiments on learning goals were perceived by the par-
ticipants as only moderately complex. For example, the scheduling task used by Latham et al.
(2008) to test for goal difficulty effects was perceived by the participants to be only moderately
complex, with an average score of 3.62 on a 5-point scale ranging from 1 (low) to 5 (high). Many
participants in that experiment mastered the task-relevant strategies. Hence, no evidence of a
 192 SEIJTS AND LATHAM

moderating effect for goal commitment was found. A task that is perceived as high on complex-
ity by individuals allows for a more rigorous test of the learning goal commitment–performance
relationship than one that is perceived as moderately complex.
Goal-setting theory states that goal difficulty level should be more highly related to task per-
formance among individuals with high commitment than among those with low commitment to
a goal. This hypothesis has yet to be empirically supported with regard to a learning goal.
In a review of the literature, Locke and Latham (1990) concluded that considerable effort on
the part of an experimenter is typically required to get variance in goal commitment. This is
because the performance goals that are assigned in laboratory and field experiments are typically
viewed by participants as legitimate and hence are readily accepted by them (Locke, Latham, &
Downloaded by [University of Toronto Libraries] at 14:33 06 October 2011

Erez, 1988). Nevertheless, there is empirical support for the necessity of commitment to a high-
performance goal in order for it to increase performance (e.g., Erez & Zidon, 1984). Using a
highly complex task should introduce sufficient variance in goal commitment to detect whether
this is also the case with a learning goal. Thus the first and second hypotheses tested in this
experiment were as follows:
H1: There is a significant and positive relationship between commitment to a learning goal
and performance on a task that requires the acquisition of knowledge.
H2: There is a significant interaction between the difficulty level of a learning goal and goal
commitment on performance of a task that requires the acquisition of knowledge. That
is, performance is significantly higher for those individuals who are highly committed to
a high-learning goal than it is for those with low commitment.

SELF-EFFICACY AND LEARNING GOALS

Very few experiments on learning goals have measured self-efficacy. Yet self-efficacy, along with
goal setting, has been shown to be a fundamental component of self-management (e.g., Frayne &
Latham, 1987; Latham & Budworth, 2006; Latham & Frayne, 1989; Yanar, Budworth, & Latham,
2009). Social cognitive theory (Bandura, 1986, 1997) states that whether individuals commit to
a goal depends, in part, on their level of self-efficacy regarding task performance. Self-efficacy is
defined as an individual’s belief that a specific task can be mastered. Bandura (1997) explained
that “effective functioning requires both skills and the efficacy beliefs to use them well” (p. 37).
Experiences with a particular task provide individuals with cues whether they have the capabili-
ties to succeed. Self-efficacy, therefore, is partly rooted in existing abilities. Voluminous studies
have shown that self-efficacy has a direct effect on performance (Bandura, 1986, 1997). Prior
studies involving specific high performance goals have also shown that self-efficacy plays a
role in keeping individuals committed to a course of action, in particular when obstacles to
performance mastery are encountered (e.g., Bandura & Cervone, 1986). This is because high
self-efficacy mobilizes and sustains the effort needed to perform a task effectively. What is not
known is whether the effect of self-efficacy on performance is partially mediated by commitment
when a learning rather than a performance goal is set. Thus the third hypothesis tested in the
experiment was as follows:
H3: Commitment to a learning goal partially mediates the relationship between self-efficacy
and performance of a task that requires the acquisition of knowledge.
 LEARNING GOAL 193

Studies have reported reciprocal relationships between performance and self-regulatory vari-
ables including self-efficacy, information processing, and commitment (e.g., Durham, Knight, &
Locke, 1997; Seijts et al., 2004; Wood & Bandura, 1989). For example, those individuals with
high commitment to the goal perform at higher levels than those with low commitment, and this
in turn contributes to an increase in self-efficacy. Our experimental design allowed us to explore
such reciprocal relationships as we measured variables multiple times. Thus the fourth hypothesis
tested in the experiment was as follows:

H4: Performance partially mediates the relationship between commitment to a learning goal
and self-efficacy on a task that requires the acquisition of knowledge.
Downloaded by [University of Toronto Libraries] at 14:33 06 October 2011

METHOD

Sample and Design

Undergraduate business students (N = 128) were randomly assigned to one of two goal condi-
tions: a specific low (n = 64) versus high (n = 64) learning goal. Their mean age was 18.57 years
(SD = 1.56). Eighty-one male students and 47 female students participated in the experiment. All
of them had taken courses in accounting, finance, entrepreneurship, marketing, and operations.
The experiment was conducted in a computer laboratory, with 6 to 8 individuals in each session.
Course credit was given for participation.

Experimental Task

The Cellular Industry Business Game (CIBG), created by Audia, Locke, and Smith (2000), was
used to test the hypotheses. The CIBG is an interactive, computer-based simulation that is based
on events that actually occurred in the U.S. cellular telephone industry.
The CIBG was used for two reasons. First, the task is highly complex as defined by Wood’s
(1986) criteria for component, dynamic, and coordinative complexity. The participants must pro-
cess numerous information cues to do well on the task. Furthermore, the task uses a complex set
of formulas to link strategic actions to performance outcomes; these formulas vary over time to
reflect the ongoing changes in the industry. Performance on the task cannot be increased solely
through effort and persistence. Effective performance requires information seeking and integra-
tion to discover appropriate strategies. The participants in a previous experiment on learning
goals (Seijts et al., 2004) perceived the task to be highly complex, with an average score of 4.12
on a 5-point scale ranging from 1 (low) and 5 (high).
Second, using a realistic, relevant task enhances the extent to which results generalize to
field settings (Latham & Lee, 1986; Locke, 1986). As was the case with the actual deci-
sion makers in the cellular telephone industry, the participants in the present experiment had
not acquired effective performance routines; they needed to learn how to increase market
share.
Participants played the role of Mr. Douglas, the founder, CEO, and sole decision maker of
a cellular telephone company for 13 decision periods, each corresponding to a year of activity.
They were told that there would be 15 rounds of decision making. The experiment ended after
 194 SEIJTS AND LATHAM

the 13th round to prevent an “endgame effect.” All 13 decision periods were completed in one
session; there was no time limit on completing the session.
Each participant started with a 7% market share. During each round, participants could make
decisions concerning 10 areas of activity: pricing, advertising, research and development, sales
force, cost containment, radio wave capacity, additional products, finance, geographic scope, and
alliances with other companies. Each area of activity allowed numerous options. For example,
in the finance area, participants could raise funds by issuing bonds, issuing public shares, or
borrowing from a bank; pay down debt on any one of the three fund-raising methods; or issue
dividend payments.
Following each decision period, all participants obtained feedback regarding market share,
Downloaded by [University of Toronto Libraries] at 14:33 06 October 2011

number of cell phone subscribers, and operating profit. An optional screen provided longitu-
dinal results on 12 other business-related indicators (e.g., advertising expenses, total debt, and
radio wave capacity). Market share was the direct result of the decisions that each individual
made.
During the first eight decision periods competition is restricted on a regional basis. The five
effective strategic options to increase market share in this phase were (a) buying licenses to oper-
ate in all available markets, (b) acquiring additional radio wave capacity allowing the company
to carry additional calls and avoid network jams, (c) raising capital to finance new investments,
(d) increasing the sales force and advertising expenditures, and (e) concentrating on business
rather than private users. Participants had the choice of using each of these strategic options in
any given decision period.
Following Year 8, however, the industry experienced a radical environmental change in the
form of deregulation. Consequently, the participants in this experiment were given several mes-
sages warning that deregulation was likely. The five strategic options that were successful before
the deregulation ceased to be effective. For example, aggressive advertising strategies after Year 9
led to a decline in operating profits. The two strategic options that were most effective after dereg-
ulation were (a) the creation of strategic alliances to ensure wide geographic coverage and (b) cost
containment.

Procedure

A preparatory case that described the company and cellular telephone industry was distributed to
participants 2 days prior to the experiment. Reading the preparatory case was a prerequisite for
participation in the experiment. In the computer laboratory, participants were given instructions
on how and when to complete the questionnaire booklet. Once the participants were familiar with
the task, they were asked to open the booklet containing the goal manipulation and measures of
the experimental variables.
Participants were instructed to focus on discovering or learning strategies that would help them
to increase market share during the completion of the simulation. Defining learning goals in terms
of an emphasis on mastering the underlying complexities of a task for effective performance is
consistent with prior studies on learning goals (e.g., Latham et al., 2008; Noel & Latham, 2006;
Seijts et al., 2004). The specific goal instructions were as follows:

The most important indicator of performance is total market share. Thinking about strategies to help
you increase market share results in higher performance. Past users of the task have shown that a
 LEARNING GOAL 195

goal of identifying and implementing two different strategies by the end of the task is difficult, yet
attainable. Research has shown that setting a difficult, yet attainable goal maximizes performance.
Thus, your goal as the new CEO is to identify and implement two or more strategies to achieve
market share by the end of the task.

Participants in the high-learning goal condition were assigned a goal of six or more strategies.
These assigned learning goals were based on Seijts et al. (2004).

Measures
Downloaded by [University of Toronto Libraries] at 14:33 06 October 2011

Performance

The market share that each participant had attained at the end of the 13th and final decision-
making round was assessed. Participants who bankrupted their organization were assigned a
market share of zero.

Goal Commitment

Commitment to the assigned learning goal was measured after decision-making Rounds 2,
7, and 10, using five 5-point Likert-type items (e.g., “I am strongly committed to pursuing this
goal”) taken from Klein, Wesson, Hollenbeck, Wright, and DeShon (2001). Scale scores could
range from 1 (strongly disagree) to 5 (strongly agree).

Self-Efficacy

Measures of self-efficacy were taken after decision making Rounds 2, 7, and 10. Thirteen
levels were assessed, ranging from attaining 1% to 37% market share at the end of the task
(e.g., “I am able to achieve 16 percent or more market share”). Consistent with the rec-
ommendations of Locke and Latham (1990), both self-efficacy magnitude and strength were
measured. Self-efficacy magnitude was operationalized as the total number of “Yes” answers
to the 13 questions (Yes = 1; No = 0). Self-efficacy strength was the sum of the rating scores
across the 13 levels. The ratings were made in terms of a continuous 100-point scale ranging
from 0 (no confidence at all) to 100 (complete confidence). The self-efficacy magnitude and
strength measures were converted to z scores and then summed to derive a total self-efficacy
score.

Task Complexity

The extent to which the participants perceived the task as complex was measured using
three 5-point Likert-type items, after they had completed the task. These items (e.g., “The
task required me to coordinate many different things at the same time”) were adapted from
Winters and Latham (1996). Scale scores could range from 1 (strongly disagree) to 5 (strongly
agree).
 196 SEIJTS AND LATHAM

Self-Set Performance Goal

The last item included in the questionnaire was, “I had a specific personal goal for attaining
a market share of _____%.” This item was included to determine whether the participants were
pursuing both a performance and a learning goal.

RESULTS

Descriptive Statistics
Downloaded by [University of Toronto Libraries] at 14:33 06 October 2011

Table 1 shows the means and standard deviations, as well as the intercorrelations, of the variables
measured.
The grand mean scale score (M = 4.04, SD = .59) indicated that the participants perceived
the task as complex. A two-tailed independent t test indicated no significant difference between
the low and high learning goal difficulty conditions, t(125) = 1.82, p > .05, on perceived task
complexity. In addition to the participants’ perceptions, there were objective data on task com-
plexity. Twenty-three percent of the participants (n = 29) bankrupted the organization. Only 20%
of them (n = 26) increased market share beyond 7%. Thus there is convergent validity that the
task was indeed complex for the participants.
The Cronbach’s alpha coefficient of goal commitment across the three measurements was .90.
The test–retest reliability between Measures 1 and 2, Measures 2 and 3, and Measures 1 and 3 was
.76, .83, and .70, respectively. The average of the three commitment measures was taken as the
overall commitment score to be used in the regression analysis to test H2. The Pearson correlation
between goal commitment and performance at the end of the simulation was .47 (p < .001); the
correlations between commitment and performance during decision-making rounds 2, 7, and 10
were .32, .59, and .59 (all ps < .001, respectively).
The Cronbach’s alpha coefficient of self-efficacy across the three measurements was .83. The
test–retest reliability between Measures 1 and 2, Measures 2 and 3, and Measures 1 and 3 was
.53, .90, and .43, respectively. The average of the three self-efficacy measures was taken as the
overall self-efficacy score to be used in the regression analysis to test H2. The Pearson correla-
tion between self-efficacy and performance at the end of the simulation was .38 (p < .001); the

TABLE 1
Descriptive Statistics and Correlations Among Variables Measured

M SD 1. 2. 3. 4. 5.

1. Assigned learning goal 0.50 0.50

2. Performance 4.15 5.87 .13
3. Self-efficacy 0.06 1.52 −.13 .38∗∗∗
4. Commitment 3.28 0.72 −.17 .47∗∗∗ .51∗∗∗
5. Task complexity 4.04 0.59 .16 .20∗ .13 .11
6. Self-set performance goal 15.68 11.75 .05 .31∗∗ .60∗∗∗ .42∗∗∗ .01

Note. N = 126–128. Assigned learning goal: 0 = low goal and 1 = high goal.
∗ p < .05. ∗∗ p < .05. ∗∗∗ p < .001.
 LEARNING GOAL 197

correlations between self-efficacy and performance during decision-making rounds 2, 7, and 10

were .12 (ns), .65 (p < .001), and .70 (p < .001), respectively.

Hypotheses

The correlations just reported support H1. The relationship between commitment to the learning
goal and performance is significant and positive.
H2 was tested using multiple hierarchical regression analysis of the data. The main effect for
learning goal difficulty level was entered in Step 1. The effects for goal commitment and self-
efficacy were entered in Steps 2 and 3. The Learning Goal Difficulty Level × Goal Commitment
Downloaded by [University of Toronto Libraries] at 14:33 06 October 2011

interaction was entered in Step 4. Goal difficulty level was a dichotomous variable (0, 1). Goal
commitment and self-efficacy were mean-centered. Centering was done to reduce multicollinear-
ity between the component measures and their associated higher order term (e.g., Cohen, Cohen,
West, & Aiken, 2003). The interaction was formed by multiplying together the predictor vari-
ables. We examined the significance of incremental variance explained in performance at each
step. The results, displayed in Table 2, show that Steps 2, 3 and 4 provided additional explained
variance.
The Goal Difficulty Level × Goal Commitment interaction was significant. The interaction,
shown graphically in Figure 1, was plotted using the resulting regression equation and inputting
the dummy-coded values for the goal difficulty manipulation and the low and high values (defined
as the mean plus and minus 1 standard deviation) for goal commitment. The nature of the
interaction is such that participants who reported high commitment to the high-learning goal
outperformed participants who reported low commitment to the high-learning goal. Commitment
made less difference on the performance of participants in the low-learning goal condition. Low
commitment hurt the effect of a challenging learning goal on performance more so than low

TABLE 2
Results of Hierarchical Regression Analysis for Variables Predicting Performance

B SE B R2 F

Step 1
Assigned learning goal 1.43 1.04 .02 (1, 125) = 1.90
Step 2
Assigned learning goal 2.45∗∗ 0.92
Commitment 4.14∗∗∗ 0.64 .25 (1, 124) = 42.28∗∗∗
Step 3
Assigned learning goal 2.56∗∗ 0.90
Commitment 3.34∗∗∗ 0.72
Self-efficacy 0.76∗ 0.34 .03 (1, 123) = 4.96∗
Step 4
Assigned learning goal 2.50∗∗ 0.88
Commitment 1.54 1.00
Self-efficacy 0.77∗ 0.33
Assigned learning goal × commitment 3.13∗∗ 1.24 .04 (1, 122) = 6.40∗∗

Note. N = 95.
∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001.
 198 SEIJTS AND LATHAM

10
High commitment
9
Low commitment
8
7

Performance
6
5
4
3
2
Downloaded by [University of Toronto Libraries] at 14:33 06 October 2011

1
0
Low goal High goal

FIGURE 1 Graphic depiction of the Learning Goal Level × Goal Commitment interaction on performance (color
figure available online).

commitment to an easier learning goal. The pattern of results is consistent with H2. The results
also revealed that the correlation between goal commitment and performance in the high goal
condition was .59 (p < .001); it was .37 (p < .001) in the low goal condition. The difference in
these two correlations was not significant.
The results in Table 2 also show that the difficulty level of the learning goal was a sig-
nificant predictor of performance. This result should be interpreted with caution. The results
in Tables 1 and 2 suggest, in fact, a suppressor effect that may have no psychological real-
ity. Self-efficacy and goal commitment were negatively correlated with learning goal level.
Neither of these correlations was significant, but they added up in the regression analysis so
that learning goal level emerged as a significant predictor of performance in Steps 2, 3, and 4.
There was no significant first-order correlation between learning goal level and performance.
These results suggest that the setting of a learning goal per se is not important; rather there
must be high commitment to a high learning goal for there to be a meaningful effect on
performance.
To test the third hypothesis, a mediation analysis was conducted in accordance with the
recommendations by Baron and Kenny (1986). Three regression equations were estimated:
(a) regressing goal commitment (mediator) on self-efficacy (independent variable), (b) regressing
performance (dependent variable) on self-efficacy, and (c) regressing performance on commit-
ment and self-efficacy. To establish mediation, three conditions must hold. First, self-efficacy
must affect commitment in the first equation. Second, self-efficacy must be shown to affect
performance in the second equation. Finally, self-efficacy and commitment must affect perfor-
mance in the third equation. Commitment mediates the relationship between self-efficacy and
performance if the aforementioned conditions all hold in the predicted direction and the effect
of self-efficacy on performance is less in the third equation than in the second equation. The
results, shown in Table 3, indicate that goal commitment partially mediated the effect of self-
efficacy on performance. The Sobel test statistic (3.53, p < .001) was significant, suggesting
partial mediation by commitment of the effect of self-efficacy on performance. H3 was therefore
supported.
 LEARNING GOAL 199

TABLE 3
Results of the Three-Step Mediated Regression Analysis for Variables Predicting Performance

Dependent Variable Independent Variables β t R2 F

Step 1 Commitment Self-efficacy .51∗∗ 6.60 .26 (1, 125) = 43.57∗∗

Step 2 Performance Self-efficacy .38∗∗ 4.53 .14 (1, 125) = 20.49∗∗
Step 3 Performance Self-efficacy .18∗ 2.02
Commitment .38∗∗ 4.20 .25 (2, 124) = 20.41∗∗

Note. N = 127.
∗ p < .01. ∗∗ p < .001.
Downloaded by [University of Toronto Libraries] at 14:33 06 October 2011

TABLE 4
Results of the Three-Step Mediated Regression Analysis for Variables Predicting Self-Efficacy

Dependent Variable Independent Variables β t R2 F

Step 1 Performance Commitment .47∗∗∗ 6.03 .22 (1, 126) = 36.35∗∗∗

Step 2 Self-efficacy Commitment .51∗∗∗ 6.60 .26 (1, 125) = 43.57∗∗∗
Step 3 Self-efficacy Commitment .43∗∗∗ 4.94
Performance .17∗ 2.02 .28 (2, 124) = 24.35∗∗∗

Note. N = 127.
∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001.

The results in Table 4 show that performance partially mediated the effect of goal commit-
ment on self-efficacy. The Sobel test statistic (1.94, p ≤ .05) was significant. H4 was therefore
supported.

Longitudinal Analysis

Figure 2 shows the results of a path analysis—we modeled the cyclical effects between self-
efficacy, commitment, and performance. The results show reciprocal relationships. For example,
an increase in market share had a positive effect on both self-efficacy and commitment; and
self-efficacy and commitment in turn facilitated subsequent performance.

Self-Set Personal Goal

Seventy-five percent (N = 96) of the participants self-set a specific performance goal ranging
from 1% to 60% of market share. Thus these people had both performance and learning goals.
The correlation between self-set performance goals for market share and actual performance was
significant (r = .31, p < .01). A chi-square test revealed that the number of people who self-
set a performance goal was significantly greater than those who did not do so, χ 2 (1) = 32.0,
p < .001). There was no significant difference between the level of the self-set performance goal
between the low (M = 15.11, SD = 12.89) and high (M = 16.30, SD = 10.49) learning goal
conditions, t(94) = 0.50, p > .05. We reran the multiple hierarchical regression analysis with
self-set performance goal for market share entered in Step 1. The direction and significance of
 200 SEIJTS AND LATHAM

.58*** 75***

.30*** .24*** .06

.12 .23*** .31*** .56 *** .25 *** .42 *** .27*** .07 .34***
Performance SE Commit Performance SE Commit Performance SE Commit Performance
trial 2 trial 7 trial 10 trial 13

.12 .34*** .44***

Downloaded by [University of Toronto Libraries] at 14:33 06 October 2011

.42*** .75***

FIGURE 2 Path analysis of relationships among self-efficacy, goal commitment, and performance across time.

the results reported in Table 2 did not change. The self-set performance goal was a significant
predictor in Step 1 (B = .16, p < .01) but not in Steps 2, 3, and 4. This finding is likely the
result of the high covariation between a self-set performance goal and self-efficacy (r = .60,
p < .001).

DISCUSSION

The present experiment overcame several methodological concerns with previous experiments
on the effect of learning goals. First, the present task was highly complex for the participants.
The participants had difficulty in increasing market share, and bankruptcies occurred throughout
the simulation. In previous studies on learning goals, there was a ceiling effect on performance.
Consequently, no improvement in performance could occur after the initial trials because nearly
everyone learned the task-relevant strategies (e.g., Latham et al., 2008). The use of a highly
complex task, coupled with a highly difficult learning goal, induced variance in goal commit-
ment, thus allowing for a strong test of its moderating effect on performance. Consistent with
goal-setting theory, the results are the first to show that a specific high-learning goal affects per-
formance only if goal commitment too is high. Specifically, commitment to a high-learning goal
is crucial for acquiring the knowledge necessary to perform a task that is complex for people.
The relationship between learning goal level and performance was stronger among participants
with high rather than low goal commitment. This finding supports the moderator effect of com-
mitment to a goal. The finding that goal commitment partially mediated the relationship between
self-efficacy and performance when a learning goal was set is also a new finding in the goal-
setting literature. The reciprocal relationships found between self-efficacy, goal commitment,
and performance are also new in the literature on goal setting. In short, on a task that one has yet
to master, it appears that performance affects and is affected by self-efficacy and commitment to
a high learning goal.
A further contribution that this experiment made to the goal-setting literature is that sev-
eral components of goal-setting theory were assessed simultaneously, namely, self-efficacy, goal
commitment, and an assigned learning goal. The search for moderating and mediating variables
allows a knowledge base to become more complete and predictions more accurate. Programmatic
 LEARNING GOAL 201

research with replications and extensions enables academics and practitioners to understand why,
when, and how specific high-learning goals affect performance. The examination of learning
goals is important because individuals have to continuously master new responsibilities and tasks
as many organizations face a highly complex environment with a multitude of new realities.
Finally, the present results revealed that the majority of the participants self-set a performance
goal even though they were committed to attaining a learning goal. This has never before been
shown. That a goal for performance was self-set is likely the result of three factors. First, the
feedback the participants received was an objective measure of their performance, namely, mar-
ket share. Second, the importance of market share, an outcome variable, was explicitly stressed
to the participants. Third, the measure of self-efficacy was a measure of confidence in attaining a
Downloaded by [University of Toronto Libraries] at 14:33 06 October 2011

percentage of market share. This measure, along with the feedback provided, likely cued partici-
pants to set a specific performance goal. The theoretical and practical significance of this finding
suggests that people are able to keep in mind their desired performance level even when their
primary focus is on a learning goal, and they can do so without hurting task mastery.

LIMITATIONS AND FUTURE RESEARCH

A drawback of using a simulation and business students as participants is that there are no tan-
gible outcomes associated with performance success or failure. Such factors as coaching and the
multiple demands on time that often occur in field settings were not operating in the present set-
ting. Despite the fact that the simulation is based on real-world events in an industrial sector, field
studies are needed to replicate the effect of learning goals on job performance and the mediating
and moderating variables.
Future field research is also needed to examine whether the effect of an assigned learning goal
and a self-set performance goal vary across time. Do individuals commit to both a learning goal
and a performance goal at the same time? Are there effects on both motivational and cognitive
outcomes? For example, Pintrich (2000) explained that

mastery and performance goals could set up and foster different patterns of motivation, affect, strat-
egy use, and performance over time. . . . Students who adopt different goals might follow different
pathways, or trajectories, over time, with some of them ending up in the “same” place in terms of
actual achievement or performance but having a very different experience on the way to this overall
outcome. (p. 545)

Ideally a self-set performance goal should be measured at the beginning of the task and again
repeatedly rather than only at the end, as was done in the present experiment. The present experi-
ment yields no clues as to when in the experiment the performance goals were self-set. Were they
set immediately? Were they set toward the end of the experiment after knowledge acquisition
had taken place? Is a self-set performance goal irrelevant when there is high self-efficacy and
high commitment to a high learning goal? In seeking answers to these questions, creativity will
be required of the experimenter to minimize sensitization effects (Campbell & Stanley, 1966)
whereby the participant is cued by the questions asked that a self-set performance goal is desired
by the experimenter.
 202 SEIJTS AND LATHAM

A third limitation was the ordering and timing of the variables and associated measures. The
mediation effect that was discovered is based on the assumption that the causal ordering pro-
ceeds from self-efficacy to goal commitment to performance. Self-efficacy was not manipulated
as a factor in the present experimental design. Self-efficacy was assessed after the participants
had received performance feedback. The conclusions from this experiment would have been
stronger had we measured initial self-efficacy at the outset before participants worked on the
task. Nevertheless, the present results are consistent with goal setting and social cognitive the-
ories as well as prior studies on performance goals. They thus bolster the conclusion that both
self-efficacy and goal commitment are antecedents of performance following the setting of a
specific learning goal. But we also acknowledge that self-efficacy and goal commitment are the
Downloaded by [University of Toronto Libraries] at 14:33 06 October 2011

result of performance through feedback. This finding of course is consistent with social cognitive
theory.
Research is also needed to address feedback, a moderator in goal-setting theory. Are there dif-
ferences in affective reactions in how people with learning versus performance goals respond
to negative feedback? Is performance enhanced on tasks that require knowledge acquisition
when learning goals include process feedback as opposed to performance feedback, or are both
optimal?
Finally, it should be noted that we did not include a simple task condition in this experiment,
nor did we include a comparison with an experimental manipulation of a performance goal to
examine differences between the effect of a performance and a learning goal. This was not done
for two reasons. First, more than 1,000 studies have used tasks that are straightforward for partic-
ipants to examine the motivational effect of performance goals on task performance (Mitchell &
Daniels, 2003). Second, studies have already shown that it is only on tasks that require knowl-
edge acquisition individuals assigned a specific learning goal significantly outperform those with
a specific high performance goal (e.g., Brown & Latham, 2002; Winters & Latham, 1996).

CONCLUSIONS

Of the 1,000 or more studies on goal setting, only 9 to the authors’ knowledge have investigated
the relationship of a specific learning as opposed to a performance goal on performance. All
have been conducted in the laboratory to ensure rigorous experimental control. The results of the
present experiment suggest that the axiom of goal-setting theory regarding the necessity for goal
commitment for a specific high goal to have a beneficial effect on performance appears to be as
true for a learning goal as it is for a performance goal.

REFERENCES

Anderson, J. R. (1985). Cognitive psychology and its implications. New York: Freeman.
Audia, P. G., Locke, E. A., & Smith, K. G. (2000). The paradox of success: An archival and a laboratory study of strategic
persistence following radical environmental change. Academy of Management Journal, 43, 837–853.
Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice
Hall.
Bandura, A. (1997). Self-efficacy: The exercise of control. New York: Freedman.
 LEARNING GOAL 203

Bandura, A., & Cervone, D. (1986). Differential engagement of self-reactive influences in cognitive motivation.
Organizational Behavior and Human Decision Processes, 38, 92–113.
Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social psychological research:
Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182.
Brown, T. C., & Latham, G. P. (2002). The effects of behavioral outcome goals, learning goals, and urging people to do
their best on an individual’s teamwork behavior in a group problem-solving task. Canadian Journal of Behavioural
Science, 34, 276–285.
Campbell, D. T., & Stanley, J. C. (1963). Experimental and quasi-experimental designs for research. Boston: Houghton
Mifflin.
Cianci, A. M., Klein, H. J., & Seijts, G. H. (2010). The effect of negative feedback on tension and subsequent perfor-
mance: The main and interactive effects of goal content and conscientiousness. Journal of Applied Psychology, 95,
Downloaded by [University of Toronto Libraries] at 14:33 06 October 2011

618–630.
Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple regression/correlation analysis for the
behavioral sciences. Mahwah, NJ: Erlbaum.
Drach-Zahavy, A., & Erez, M. (2002). Challenge versus threat effects on the goal – performance relationship.
Organizational Behavior and Human Decision Processes, 88, 667–682.
Durham, C., Knight, D., & Locke, E. A. (1997). Effects of leader role, team-set goal difficulty, efficacy, and tactics on
team effectiveness. Organizational Behavior and Human Decision Processes, 72, 203–231.
Earley, P. C., Connolly, T., & Ekegren, G. (1989). Goals, strategy development and task performance: Some limits on the
efficacy of goal setting. Journal of Applied Psychology, 74, 24–33.
Erez, M., & Zidon, I. (1984). Effect of goal acceptance on the relationship of goal difficulty to performance. Journal of
Applied Psychology, 69, 69–78.
Frayne, C. A., & Latham, G. P. (1987). Application of social learning theory to employee self-management of attendance.
Journal of Applied Psychology, 72, 387–392.
Kanfer, R., & Ackerman, P. L. (1989). Motivation and cognitive abilities: An integrative/aptitude-treatment interaction
approach to skill acquisition. Journal of Applied Psychology, 74, 657–690.
Klein, H. J., Wesson, M. J., Hollenbeck, J. R., Wright, P. M., & DeShon, R. P. (2001). The assessment of goal
commitment: A measurement model meta-analysis. Organizational Behavior and Human Decision Processes, 85,
32–55.
Kozlowski, S. W. J., & Bell, B. S. (2006). Disentangling achievement orientation and goal setting: Effects on
self-regulatory processes. Journal of Applied Psychology, 91, 900–916.
Latham, G. P., & Budworth, M. H. (2006). The effect of training in verbal self-guidance on the self-efficacy and
performance of Native North Americans in the selection interview. Journal of Vocational Behavior, 68, 516–523.
Latham, G. P., & Frayne, C. A. (1989). Self-management training for increasing job attendance: A follow-up and a
replication. Journal of Applied Psychology, 74, 411–416.
Latham, G. P., & Lee, T. (1986). Goal setting. In E. A. Locke (Ed.), Generalizing from laboratory to field settings.
Lexington, MA: Lexington Books.
Latham, G. P., & Locke, E. A. (2007). New developments and directions for goal setting. European Psychologist, 12,
290–300.
Latham, G. P., Seijts, G. H., & Crim, D. (2008). The effects of learning goal difficulty level and cognitive ability on
strategy development and performance. Canadian Journal of Behavioural Sciences, 40, 220–229.
Locke, E. A. (1986). Generalizing from laboratory to field settings. Lexington, MA: Lexington Books.
Locke, E. A. (2007). The case for inductive theory building. Journal of Management, 33, 867–890.
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.
Locke, E. A., Latham, G. P., & Erez, M. (1988). The determinants of goal commitment. Academy of Management Review,
1, 23–39.
Mitchell, T. R., & Daniels, D. (2003). Motivation. In W. C. Borman, D. R. Ilgen, & R. J. Klimoski (Eds.), Handbook of
psychology, Vol. 12: Industrial organizational psychology (pp. 225–254). New York: Wiley & Sons.
Mone, M. A., & Shalley, C. E. (1995). Effects of task complexity and goal specificity on change in strategy and
performance over time. Human Performance, 8, 249–262.
 204 SEIJTS AND LATHAM

Noel, T. W., & Latham, G. P. (2006). The importance of learning goals versus outcome goals for entrepreneurs.
International Journal of Entrepreneurship and Innovation, 7, 213–220.
Pintrich, P. R. (2000). Multiple goals, multiple pathways: The role of goal orientation in learning and achievement.
Journal of Educational Psychology, 92, 544–555.
Seijts, G. H., & Latham, G. P. (2001). The effect of distal learning, outcome, and proximal goals on a moderately complex
task. Journal of Organizational Behavior, 22, 291–302.
Seijts, G. H., & Latham, G. P. (2005). Learning versus performance goals: When should each be used? Academy of
Management Executive, 19, 124–131.
Seijts, G. H., Latham, G. P., Tasa, K., & Latham, B. W. (2004). Goal setting and goal orientation: An integration of two
different yet related literatures. Academy of Management Journal, 47, 227–239.
Yanar, B., Budworth, M. H., & Latham, G. P. (2009). The effect of verbal self-guidance training for overcoming
Downloaded by [University of Toronto Libraries] at 14:33 06 October 2011

employment barriers: A study of Turkish women. Applied Psychology: An International Review, 58, 586–601.
Winters, D., & Latham, G. P. (1996). The effect of learning versus outcome goals on a simple versus a complex task.
Group and Organization Management, 2, 235–250.
Wood, R. E. (1986). Task complexity: Definition of the construct. Organizational Behavior and Human Decision
Processes, 37, 60–82.
Wood, R. E., & Bandura, A. (1989). Social cognitive theory of organizational management. Academy of Management
Review, 14, 361–384.

View publication stats