Social Comparison Perfomance

Journal of Applied Psychology
© 2020 American Psychological Association 2021, Vol. 106, No. 3, 422– 438
ISSN: 0021-9010 http://dx.doi.org/10.1037/apl0000519
Referents or Role Models? The Self-Efficacy and Job Performance Effects

of Perceiving Higher Performing Peers
Patrick E. Downes Eean R. Crawford

Texas Christian University University of Iowa
Scott E. Seibert Adam C. Stoverink

Rutgers University University of Arkansas
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
Elizabeth M. Campbell
University of Minnesota
What are the effects of perceiving peers’ higher performance? Social– cognitive theory emphasizes the
positive influence that perceiving higher performers can have on observer task and job performance
(because observational learning from role models enhances self-efficacy). Social comparison theory
emphasizes the negative self-evaluations that accompany perceiving higher performers, which should
under many circumstances reduce self-efficacy and subsequent task and job performance. To more fully
understand the effects of perceiving higher performance, we argue the effects of perceived higher
performers on observer task and job performance depend on individuals’ disposition in how they
cognitively process coworkers’ performance. Drawing on goal orientation theory, we suggest individuals
with higher levels of performance prove goal orientation (PPGO) primarily interpret perceived higher
performers as comparative referents rather than as instructive role models, inhibiting social learning and
reducing self-efficacy. Results from a 2 studies (a field study of 110 corporate employees as well as an
experimental study with 107 undergraduate students) support these ideas: Individuals with higher levels
of PPGO have decreased self-efficacy and performance when observing higher performing coworkers,
and individuals with lower levels of PPGO have increased self-efficacy and performance when observing
higher performing coworkers.
Keywords: goal orientation, self-efficacy, social– cognitive theory, social comparison theory
Supplemental materials: http://dx.doi.org/10.1037/apl0000519.supp
Organizations desire high performing employees, but how are inger, 1954; Tesser, 1988), which can spark detrimental responses
individuals affected by their own perceptions of higher performing like envy and threat (Campbell, Liao, Chuang, Zhou, & Dong,
coworkers? Research converges on the general view that one’s 2017; Kim & Glomb, 2014) as well as harmful behaviors that
own performance is affected by perceiving others to be higher could negatively impact performance (Jensen, Patel, & Raver,
performers (Call, Nyberg, & Thatcher, 2015; Kehoe & Tzabbar, 2014; Lam, Van der Vegt, Walter, & Huang, 2011). However,
2015; Lockwood & Kunda, 1997), but there is little consensus on perceiving higher performers can benefit observers’ learning, mo-
the direction and contingencies of these effects. Perceiving higher tivate pursuit of higher goals, and enhance self-image through
performers leads observers to make “upward comparisons” (Fest- affiliation (Bandura, 1977; Bandura & Locke, 2003; Cialdini et al.,
1976; Stajkovic & Luthans, 1998). Based on these outcomes,
should organizations encourage employees to interact with those
This article was published Online First May 28, 2020. they see as higher performers, or would such practices invite
Patrick E. Downes, Department of Management and Leadership, Nee- unfavorable comparisons that negatively affect motivation and
ley School of Business, Texas Christian University; Eean R. Crawford, De- performance?
partment of Management and Entrepreneurship, University of Iowa; Scott E. Two theoretical perspectives explain these reactions. On the one
Seibert, Department of Human Resource Management, Rutgers University; hand, social comparison theory (Festinger, 1954) portrays higher
Adam C. Stoverink, Department of Management, University of Arkansas; performers as referents against whom observers evaluate their own
Elizabeth M. Campbell, Department of Work and Organizations, University of
levels of ability and performance. At work, social comparisons are
Minnesota.
Correspondence concerning this article should be addressed to Patrick E. difficult to avoid (Shah, 1998) because employees work closely
Downes, Department of Management and Leadership, Neeley School of with one another and often share similarities that stem from
Business, Texas Christian University, Box 298530, Fort Worth, TX 76129. working in the same organization, industry, and geography. In
E-mail: [email protected] essence, the social environment of workplaces lays the stage for
422
REFERENTS OR ROLE MODELS? 423
workers to compare themselves against perceptions of higher (e.g., appraisal of threat or benefit to resources; Campbell et al.,
performers. Comparisons in which individuals perceive them- 2017) that drive observers’ attitudes and behaviors toward higher-
selves to have lower relative performance can negatively impact performing peers. Yet more nascent is research exploring how
observers’ self-concept (Wood, 1989), which could adversely im- perceptions of higher performance influence observers’ own job
pact future performance.1 On the other hand, social– cognitive performance. For example, Lee and Duffy (2019) identified envy
theory (Bandura, 1986) generally portrays higher performers as as an emotional mechanism through which peers’ own perfor-
role models from whom observers learn effective performance mance changes after perceiving a higher performer. Building on
strategies. In this view, higher performers should elevate observ- this emotional mechanism, we examine self-efficacy as a cognitive
ers’ self-efficacy (i.e., confidence in one’s own abilities in a job or mechanism through which comparisons to higher performers
task), leading to setting higher goals, pursuing goals with more should impact observers’ own subsequent job performance. We
vigor, and ultimately achieving higher performance (Bandura,
contend that perceiving coworkers as higher performing employ-
2012; Bandura & Locke, 2003; Stajkovic & Luthans, 1998).
ees (whether or not they meet the more substantial criteria requisite
Contrasting the predictions of social comparison and social–
to qualify as a star; Call et al., 2015) can spark self-evaluative
cognitive theories in this way raises an important theoretical and

cognitions that can influence future performance.
practical question: When do higher performers function as refer-

ents, and when do they function as role models in the workplace? A second contribution lies in extending social comparison and
That is, when are higher performers likely to have a negative social– cognitive theories by jointly applying them to observers’
impact on the performance of those who observe them, and when responses to perceptions of higher performance. Our key research
do they have a positive impact on the performance of their co- question highlights that observing higher performance could have
workers? We research this question to determine whether the either positive or negative effects on observers’ future perfor-
answer depends on a key dispositional characteristic that influ- mance. In social– cognitive theory, role models enable learning,
ences how an observer interprets performance information in a which is generally seen as beneficial (e.g., Myers, 2018). Under
social comparison environment. Conducting two studies (one field certain adverse circumstances, learning does not occur and there is
study and one experiment), we find that an observer’s performance no effect. Yet if learning is impeded as a result of social compar-
prove goal orientation (PPGO) determines whether observing ison processes, observing higher performance should be nega-
higher performing coworkers harms or helps the observer’s own tive—and not simply neutral—for observers’ future performance.
self-efficacy and performance. PPGO describes individuals’ desire Despite Bandura (1997) raising the issue, this potential negative
to display competence and outperform others (Elliot & Church, effect on observer performance has gone underdeveloped in re-
1997), which influences how individuals make sense of perfor- search employing social– cognitive theory. In contrast, social com-
mance information from multiple simultaneous social comparisons parison theorists have articulated this duality and pointed to sev-
(Levy, Kaplan, & Patrick, 2004). In the context of social compar- eral situational contingencies that determine whether effects
isons to perceived higher performers, we argue that individuals should be positive or negative (e.g., identification; Kim & Glomb,
with higher levels of PPGO view others primarily as evaluative
2014; assimilation/contrast, Lam et al., 2011; similarity, Tesser,
referents (Button, Mathieu, & Zajac, 1996) against whom they
1988). We move beyond these situational explanations by high-
compare their own standing. Because self-evaluative thinking can
lighting individuals’ predisposition to differentially interpret per-
thwart learning (Bandura, 1997), we expect those with higher
ceptions of peers’ higher performance (i.e., PPGO). Cognitively
levels of PPGO to have decreased self-efficacy and performance
when observing higher performing coworkers. In contrast, indi- processing situational contingencies across multiple peers at work
viduals with lower levels of PPGO are not inclined toward com- is complex, leading individuals to develop systematic heuristics,
petitive self-evaluations against others’ higher performance (Elliot, such as PPGO, that simplify interpretation of social environments
2005). Without threat to self-evaluation, low-PPGO individuals (Mischel & Shoda, 1995).
are free to view others as instructive role models and attend to the Finally, our work extends literature on goal orientation by more
performance strategies that enable their higher performance. deeply connecting social comparisons, learning, and PPGO. Re-
Social– cognitive theory describes that this observational learning search on PPGO in educational contexts has suggested that PPGO
process should lead to increased self-efficacy and job performance affects the frequency of social comparisons (Régner, Escribe, &
(Bandura, 1997). Dupeyrat, 2007), but it remains an open question whether PPGO
This integration of social comparison, social– cognitive, and affects responses to social comparisons once they are made. Our
goal orientation theories offers a new depiction of observers’ research offers new theory describing how PPGO affects the
responses to perceptions of peers’ higher performance that con- development of self-efficacy. It also contributes to research em-
tributes to several relevant literatures. First, our theory extends phasizing the role of PPGO in social situations (e.g., Dietz, van
research on observer responses to higher performers, particularly Knippenberg, Hirst, & Restubog, 2015; Dragoni, Tesluk, Russell,
those focusing on observers’ feelings of envy and victimization of & Oh, 2009).
higher performers (e.g., Jensen et al., 2014; Kim & Glomb, 2014;
Lam et al., 2011) as well as stars, who are higher performers but
1
also have greater visibility, social capital, and perform dispropor- Although social comparison theory has recognized that in certain
tionately better than peers (e.g., Aguinis & O’Boyle, 2014; Call et specific circumstances this reduced self-concept can be functional in pro-
ducing an internal drive toward higher performance (e.g., Buunk & Gib-
al., 2015; Kehoe, Lepak, & Bentley, 2018). Researchers in these bons, 2007), the effect often observed in the literature is that perceiving
areas have highlighted the emotional reactions (e.g., envy; Camp- others as better performers adversely affects one’s self-concept (Dineen et
bell et al., 2017; Kim & Glomb, 2014) and cognitive mechanisms al., 2017; Suls et al., 2002).
424 DOWNES, CRAWFORD, SEIBERT, STOVERINK, AND CAMPBELL
Theoretical Background Social Cognitive Theory

According to Bandura (1997, p. 90), observers interpret higher Social– cognitive theory suggests that peers influence observers’
performers through a “complex set of interrelated [cognitive] sub- job performance (Bandura, 1977) through self-efficacy, which
functions.” Two subfunctions that are germane to our research are refers to “people’s sense of personal efficacy to exercise some
the instructive function that oversees social learning, and the control over events that affect their lives” (Bandura, 1986, p. 391).
self-evaluative function that oversees social comparison (Bandura, Self-efficacy is not the presence of skills or competencies per se,
1997). Self-efficacy is instilled and strengthened by “maximizing but rather one’s own perception of one’s capabilities to organize
modeling’s instructive function and minimizing its comparative skills into a course of action to achieve goals. Although Bandura
evaluation function” (Bandura, 1997, p. 92). Thus, although (1997) identified a number of mechanisms that influence self-
social– cognitive theory recognizes multiple cognitive functions, efficacy, observational learning is most germane to our hypothe-
researchers examining the theory have focused primarily on the ses. Observational learning occurs when individuals, having ob-
conditions under which the instructive function dominates, as well served a role model’s behavior or performance strategies, process
that behavior as personally relevant, and subsequently add it to
as the outcomes of that instructive function. Social comparison

their own behavioral response repertoire (Bandura, 1986; Myers,
theory (Festinger, 1954) recognizes similar cognitive processes

(e.g., Tesser, 1988), referring to them sometimes as motives for 2018). Observers’ vicarious experiences allow them to evaluate
self-improvement or self-evaluation (see Wood, 1989). In contrast courses of action and determine which performance strategies will
to social– cognitive theory, research on social comparison typically be effective in a given domain (Bandura, 1997; Gist & Mitchell,
focuses on the conditions under which the self-evaluative function 1992), enhancing self-efficacy and performance.
dominates, as well as the outcomes of that self-evaluation. In Social– cognitive theory recognizes that observational learning
Wood’s (1989) view, social comparison theory “emphasized ac- can be inhibited by certain characteristics of the environment,
curate self-evaluation as the purpose of social comparison” (p. observer, or role model. Bandura (1986) refers to these inhibitory
232) and lamented that “very little research has been aimed at features of social learning as “attentional deficits” (p. 54), which
discovering the ways in which social comparisons may be used for prevent observers from noticing and integrating the information
self-improvement” (p. 238). Both theories also agree that achieve- role models provide. Observational learning requires that observ-
ment contexts regularly confront individuals with perceptions of ers pay attention to role models, mentally rehearse what they have
higher-performing others, and in doing so initialize the self- observed, and then incorporate their new understanding of perfor-
evaluative and instructive functions in the minds of observers. mance into their expectations about their future performance (Ban-
Bandura (1997) describes this as workers being “continually con- dura, 1997). Provided that observers devote sufficient cognitive
fronted with comparative appraisals whether they seek them or resources to the performance strategies and outcomes of role
not” (p. 457). Social comparison theorists describe the phenome- models, observers should learn and enjoy higher levels of self-
non as upward social comparisons that “occur automatically” as efficacy (Bandura, 1997).
part of everyday life (Wood, 1989, p. 233).
One important point is that both social comparison and social– Social Comparison Theory
cognitive theories base their propositions on perceptions of higher According to Festinger’s (1954) social comparison theory, in
performers in the work environment rather than veridical perfor- situations where standards are ambiguous, people compare them-
mance. Veridical performance reflects peers’ true underlying per- selves with salient others to gauge their own standing or worth on
formance that can be verified as factual against some data, such as some comparison dimension. Although some work activities pro-
number of widgets produced, sales volume, or rating in the per- vide objective performance feedback, the true performance value
formance management system. In the veridical performance envi- of most kinds of work contributions are ambiguous. Thus, at work,
ronment, the highest performer has no one with whom they can employees use social comparison information to construct a sense
make an upward comparison. In contrast, the perceived perfor- of what adequate performance looks like (Tesser, 1988). This
mance environment assumes that individuals must cognitively comparative information can be valuable even when objective
make sense of their social environment, including (a) choosing performance feedback is provided, as given performance levels
which coworkers are salient for comparison (i.e., Kulik & Am- take on richer meaning through social comparison. For example,
brose, 1992; Shah, 1998), (b) considering which dimensions of selling 15 units is objective performance feedback for a salesper-
broader job performance are relevant to their own jobs, and (c) son, but selling 15 units shifts in meaning in a context where others
forming perceptions of salient coworkers’ performance levels. sell 20 units compared with a context where others sell 10 units.
This means that even the individual with the highest veridical Social comparisons are likely even more important in uncertain
performance may perceive other salient coworkers as higher per- contexts that are common to the workplace where performance is
formers in the perceived performance environment. These percep- defined along multiple criteria, includes subjective components,
tions are prominent in both social– cognitive and social compari- and may not be transparent or easily discernable among workgroup
son theories because they are proximal to individuals’ motivation members.
and behavior. This is because observers must perceive higher The effects of upward social comparisons in the workplace are
performance for it to have an effect on self-evaluation and learn- generally described as detrimental for the observer in organiza-
ing; a veridically higher performer in an individual’s work envi- tional research (Buunk, Zurriaga, Gonzalez-Roma, & Subirats,
ronment can only initiate comparison and learning cognitions if the 2003; Dineen, Duffy, Henle, & Lee, 2017; Duffy, Scott, Shaw,
observer (a) views the coworker as salient and (b) perceives the Tepper, & Aquino, 2012). Although scholars have noted that
coworker to be a higher performer. upward social comparisons can have positive outcomes for observ-
ers under specific circumstances (Lockwood & Kunda, 1997; Suls, performance-avoid goal orientation (PAGO). LGO reflects indi-
Martin, & Wheeler, 2002), many organizational researchers posit viduals’ desire to improve their knowledge and master skills
that upward social comparisons generally highlight individuals’ (Dweck, 1986), PPGO describes individuals’ desire to display
shortfalls relative to others (Hogg, 2000), which threatens self- their competence and outperform others (Elliot & Church, 1997),
esteem (Buunk & Gibbons, 2007; Lockwood, Jordan, & Kunda, and PAGO represents individuals’ desire to avoid appearing in-
2002) and incites feelings of inferiority and envy (Cohen-Charash competent in achievement pursuits. In the current study, we focus
& Mueller, 2007; Dineen et al., 2017; Duffy et al., 2012). For these on PPGO rather than LGO or PAGO as the key moderator of the
reasons, upward comparisons are often cast as negative for ob- outcomes of upward social comparisons. This is because high-
servers’ self-concepts and well-being. LGO individuals are typically self-focused rather than others-
focused; they have a general “tendency to evaluate performance
relative to past episodes of performance” (Button et al., 1996, p.
Applying the Theories to Higher Performers’
40). High-LGO individuals define competence using an “absolute-
Influence on Observer’s Performance
intrapersonal” (Elliot, 2005, p. 62) standard, which reflects indi-
Both social– cognitive theory and social comparison theory offer viduals’ improvement over time at completing the requirements of
predictions about how perceiving higher performers influences the task. In contrast, high-PPGO individuals look to others (i.e.,
observers. The theories are similar in positioning the self-concept they have a general “tendency to evaluate performance by norma-
as the mechanism through which peers shape observers’ job per- tive” standards; Button et al., 1996, p. 40), representing an “inter-
formance. However, they often depict the self-concept differently; personal” (Elliot, 2005, p. 62) standard. High-PPGO individuals
social comparison theory emphasizes self-esteem whereas social– rely heavily upon the normative standard, answering the question
cognitive theory focuses on self-efficacy. Self-efficacy and self- “how well am I performing?” by comparing their own perfor-
esteem differ in their breadth. Self-efficacy represents beliefs mance to others’ performance. Thus, although high-LGO individ-
about one’s ability to perform specific tasks, whereas self-esteem uals may be more motivated toward learning, they spend much of
represents broad feelings of self-worth across domains (Harter, their cognitive energy focused on self-improvement rather than
1990). Researchers have positioned both constructs as aspects of evaluating and observing referent others. Finally, we focus on
the self-concept (Judge, Locke, & Durham, 1997), noting that as PPGO instead of PAGO given that PAGO has been referred to as
the focus of the self-esteem construct becomes narrower, it be- the “dysfunctional branch” of GO (Payne, Youngcourt, &
comes more closely related to self-efficacy (Brockner, 1988; Gist Beaubien, 2007), and is posited to lead to a helpless motivational
& Mitchell, 1992). process rather than one of adaptation and achievement (Elliot &
Church, 1997). For these reasons, high-PAGO individuals are not
likely either to compare themselves or aspire to resemble higher
Hypothesis Development
performers, but instead observe coworkers in an effort to achieve
Both social– cognitive and social comparison theories recognize the minimum level of performance necessary to avoid appearances
that the instructive cognitive function, which enables learning and of incompetence.
should elevate self-efficacy, and the self-evaluative cognitive func- High-PPGO individuals, because they pay close attention to
tion, which focuses on observers’ relative poor performance and others and use social information for self-evaluation, are primed
should reduce self-efficacy, are at odds with one another (Bandura, for self-evaluative cognitive process that forestall learning. Indi-
1997; Wood, 1989). This is because cognitive resources are lim- viduals with higher levels of PPGO tend to view others’ perfor-
ited, and those spent on the self-evaluation are unavailable for mance as self-evaluative threats rather than instructive role models
instruction. Accordingly, we suggest that the number of higher (Darnon, Dompnier, Delmas, Pulfrey, & Butera, 2009). They
performers one observes relates positively to self-efficacy for evaluate their own performance in terms of demonstrating com-
individuals who view higher performers through an instructive petence and outperforming others (Elliot & Church, 1997), using
lens, and negatively for individuals who view higher performers normative rather than absolute standards. Owing to the tendency
through a self-evaluative lens. for high-PPGO individuals to interpret social situations as oppor-
Core to this proposition is the notion that interpreting uncertain tunities for self-evaluative comparison, PPGO should explain
social environments requires the use of cognitive heuristics that whether observers benefit from social learning. This proposition is
simplify one’s understanding of his or her social world. People depicted in Figure 1.
differ in how they form and access these heuristics, leading to When high-PPGO individuals perceive that coworkers have
variation in how they process the informational cues available in higher performance, they are more likely to view the coworkers as
social situations (Mischel, 1973; Mischel & Shoda, 1995). People
derive different meaning of the same events and behaviors; two
different observers could consider the same set of peers as either
evaluative referents or instructive role models. These cognitive Number of Perceived Job-Based Self-
Job Performance
Higher Performers Efficacy
patterns of processing social information should qualify the impact
of perceived higher performance on observers’ self-efficacy.
Goal orientation (GO) is a major dispositional influence on how
individuals view and process information in achievement-related Performance-Prove
environments such as the workplace (Lee, Sheldon, & Turban, Goal Orientation
2003; Poortvliet & Darnon, 2010). Goal orientations are expressed
in three dimensions: learning goal orientation (LGO), PPGO, and Figure 1. Hypothesized model of relationships.
evaluative referents than instructive role models (VandeWalle, Although this prediction is consistent with social– cognitive
Cron, & Slocum, 2001). These views activate social comparison theory (Bandura, 1986), recent research on the causal ordering of
cognitions that thwart learning. Because high-PPGO individuals self-efficacy and performance is in a state of debate. Meta-analytic
weight self-evaluation over learning, the number of higher per- estimates demonstrate that individuals with higher levels of self-
formers they perceive in the environment should negatively relate efficacy are generally higher performers (Stajkovic & Luthans,
to their self-efficacy. Perceiving more higher performers present 1998). Yet other research shows that self-efficacy can be just as
stronger threat, reinforcing negative social comparisons and pull- much— or even more—a reflection of past performance as it is a
ing cognitive resources from learning. Owing to a focus on predictor of future performance (Heggestad & Kanfer, 2005; Sitz-
self-evaluation over learning, the number of perceived higher mann & Yeo, 2013). Other voices in the debate have argued that
performers should negatively relate to job-based self-efficacy for self-efficacy is antecedent to performance, but in a negative direc-
high-PPGO individuals. In contrast, when low-PPGO individuals tion under certain circumstances (Vancouver, More, & Yoder,
perceive that coworkers have higher performance, they are more 2008). There appears to be agreement that self-efficacy can affect
likely to view coworkers as role models rather than evaluative performance (Bandura, 2015; Vancouver & Purl, 2017) under
referents. Low-PPGO individuals have a reduced motivational certain circumstances. Sitzmann and Yeo’s (2013) meta-analysis
drive to engage in self-evaluative cognitions, and thus have more reports a positive relationship between self-efficacy and perfor-
cognitive resources available to learn from role models (Bandura, mance when the trajectory of performance is positive (i.e., learning
1965). Thus, for low-PPGO individuals we expect a positive is possible; see also Bandura, 2012). Because job and task perfor-
relationship between the number of perceived higher performers mance in organizations is often a context in which performance
and self-efficacy. can improve over time (Quiñones, Ford, & Teachout, 1995), we
One important note is that social– cognitive and social compar- expect a positive relationship between self-efficacy and subse-
ison theories can be applied in the lower-bound case when indi- quent performance.
viduals do not perceive any higher performers in their environ-
ment. Our theory posits that high-PPGO individuals at this lower Hypothesis 2: Individual self-efficacy positively relates to
bound experience higher levels of self-efficacy, and that self- performance.
efficacy decreases as the number of perceived upward compari-
sons rises. For the high-PPGO individual, zero perceived higher Integrated Conditional Indirect Effects Model
performers signifies that she has succeeded in displaying her Taken as a whole, Hypotheses 1 and 2 suggest a conditional
competence and outperforming others. Because there are no ad- indirect effects model where the effect of the number of perceived
verse social comparisons, the high-PPGO individual should expe- higher performers on performance is mediated by self-efficacy,
rience high levels of self-efficacy. In contrast, for the low-PPGO and this indirect effect is moderated by PPGO. Hypothesis 1 is the
individual, zero perceived higher performers signifies that she has first stage of our model, whereby the relationship between the
no instructive role models from which she can learn. Because there number of perceived higher performers and self-efficacy is nega-
are no beneficial social comparisons, the low-PPGO individual tive for individuals high in PPGO and is positive for individuals
should experience lower self-efficacy than he or she would in the low in PPGO. Hypothesis 2 is the second stage of our model,
presence of instructive role models to elevate his or her aspirations. where self-efficacy positively relates to performance. This first-
Thus, we suggest that low-PPGO individuals at this lower bound stage conditional indirect effects model reflects the core proposi-
experience lower levels of self-efficacy, and that self-efficacy tions of our theory: for individuals with higher levels of PPGO,
increases as the number of perceived higher performers. higher performers are perceived primarily through the self-
evaluative cognitive function that weakens the self-efficacy gains
Hypothesis 1: PPGO moderates the relationship between the
from learning and strengthens the self-efficacy losses from com-
number of perceived higher performers and self-efficacy such
parison, resulting in lower self-efficacy and performance. For
that a negative relationship exists for high-PPGO individuals
individuals with lower levels of PPGO, higher performers can be
and a positive relationship exists for low-PPGO individuals.
perceived as role models, initiating the instructive cognitive func-
tion and enhancing observers’ self-efficacy. We thus formally
Self-Efficacy and Performance hypothesize this conditional indirect effect.
We expect self-efficacy positively relates to job performance. Hypothesis 3: The indirect effect of the number of perceived
Individuals with higher levels of self-efficacy set more challenging higher performers on performance via self-efficacy is positive
goals and demonstrate more commitment toward achieving those for low PPGO and negative for high PPGO.
goals (Wofford, Goodwin, & Premack, 1992). By setting their
sights on more challenging (relative to easier) goals, individuals Study 1
aspire to higher levels and are likely to perform at higher levels.
Further, individuals with higher levels of self-efficacy are likely to
Method
persist to higher levels of job performance beyond individuals with
lower self-efficacy (Bandura, 2012). This is particularly true when Sample and procedure. Data were collected from employees
individuals encounter obstacles and negative feedback, as those in corporate functions of a food production company located in the
with higher levels of self-efficacy remain steadfast in their original United States (Approved by University of Iowa Institutional Re-
goal (Bandura, 1997; Locke & Latham, 2002). For these reasons, view Board [Project Number 201410770, Title: “Social Work
self-efficacy should positively relate to performance. Context”]). Two hundred fifty-six individuals in supervisory, pro-
fessional, and clerical roles were eligible to participate. The sam- In the second step of the procedure, respondents reported their
ple was chosen because employees were colocated where they had perceptions of the relative performance of each salient coworker.
ample opportunity to observe coworkers’ job performance. Em- For each person listed, respondents were asked to rate the peer’s
ployees worked in well-defined yet interconnected jobs across performance relative to their own using the item, “For each name
multiple business functions. This was ideal in that employees’ that you entered earlier, please indicate this person’s performance
roles required them to work independently but also to interact with relative to yours” (1 ⫽ This person is generally a higher performer
coworkers across the company. These interactions provided a basis than me, 3 ⫽ We are generally similar in our performance, and
for employees to observe others’ performance. 5 ⫽ I am generally a higher performer than this person). We used
Two surveys were administered online using Qualtrics survey the two network items to compute the number of higher performers
software. The first survey asked respondents to identify the num- for each respondent. Specifically, number of perceived higher
ber of perceived higher performers in their social environment performers was computed by summing the number of salient
(described below). The second survey was administered one month coworkers whom observers also perceived to be higher performers
after the first survey. On the second survey, respondents indicated (i.e., rated 1 or 2 on the relative performance item). The mean
their self-efficacy, goal orientation, and demographic information. number of higher performers was 2.71 with a standard deviation of
Unless otherwise indicated, survey items were based on a 5-point 3.29.

scale (1 ⫽ strongly disagree, 5 ⫽ strongly agree). Employees’ Job-based self-efficacy was measured using Wilk and Moyni-
supervisors were asked to rate job performance on a separate han’s (2005) three-item job-based self-efficacy scale that was
survey one week after the second survey was administered. designed to measure individuals’ self-efficacy, or beliefs in their
Of the 256 eligible participants, 139 employees responded to own abilities, to meet ongoing job demands. A sample item is, “I
Survey 1 (response rate: 54.3%). About one month later, these 139 am certain I can meet the performance standards of this job” (␣ ⫽
individuals received an e-mail inviting them to respond to the .92).
second survey, and 120 individuals completed Survey 2 (response Performance-prove goal orientation was assessed on the second
rate: 86.3%). One case was removed based on an extreme value in survey with Vandewalle’s (1997) four-item scale. A sample item
which the respondent reported performance comparisons 6.4 stan- is, “I’m concerned with showing that I can perform better than my
dard deviations above the mean. For the remaining analyses, peers” (␣ ⫽ .70).
pairwise deletion was used to examine as much usable data as Control variables. Our measure of the number of perceived
possible. As a robustness check, we conducted the analysis using higher performers introduces a potential confound in that people
listwise deletion; substantive conclusions did not vary. We ob- who have more salient coworkers overall may have more per-
tained performance ratings for 110 respondents, for a final usable
ceived higher performers. We controlled for this by including the
response rate of 43.0%.
total number of salient coworkers (i.e., number of names listed in
Measures.
step one) throughout analyses.
Dependent variable. Job performance was rated by direct
An alternative explanation for how higher performers influence
supervisors using the 7-item performance scale from Williams and
their coworkers is that they control resources disproportionately,
Anderson (1991). A sample item is “this person meets the formal
making it difficult for focal individuals to access the things they
performance requirements of his/her job.” The scale exhibited high
need to effectively complete their work (Kehoe & Tzabbar, 2015).
reliability (␣ ⫽ .94). Eleven supervisors rated the 110 participants
To account for this possible alternative, we measured information
(on average 10 employees per supervisor, with a range of 7 to 19).
and resource access using two 3-item scales from Spreitzer (1996).
Because supervisors rated more than one participant, we conducted
A sample item for resource access (␣ ⫽ .89) is “I have access to
ANOVA to determine whether respondent performance differed
the resources I need to do my job well.” A sample item for
by supervisor; results suggested employee performance did not
information access (␣ ⫽ .91) is “I have access to the strategic
significantly vary between supervisors (F ⫽ .58, ns). Multilevel
techniques accounting for the nesting of supervisory performance information I need to do my job well.”
ratings were therefore not necessary.2 Because scholars have debated whether self-efficacy studies
Independent variables. Number of perceived higher perform- should control for prior performance (Heggestad & Kanfer, 2005),
ers was measured using a two-step name generation procedure we conducted analyses both ways. Specifically, we retrieved ar-
based on commonly used techniques in social network analysis chival records from the company’s performance appraisals that
(e.g., Shah, 1998). Each respondent first listed salient coworkers occurred approximately 10 months before the beginning of the
from within the company. They then indicated up to 20 individuals study. The company used a 3-point scale that coded low perform-
(M ⫽ 8.18; SD ⫽ 6.37) to whom they looked as they considered ers as 2 (2%), middle performers as 3 (52%), and high performers
their own performance at work. The following text was used as a as 4 (36%). This measure positively and significantly correlated
prompt: with later performance, r ⫽ .78, p ⬍ .01; conclusions were
unchanged with prior performance included or excluded as a
Sometimes, when we think about our own level of effort or level of covariate.
performance at work, we compare ourselves to other people in the Finally, because social comparisons and self-efficacy may vary
company to evaluate how well we are doing. These people might be across tenure or gender, we controlled for the number of years that
coworkers we look up to or aspire to be like, coworkers at about our
same level, or coworkers who are newer or more junior than our-
selves. Please list below the people at work with whom you compare 2
As a robustness check, we analyzed regressions using Mplus 7 to
yourself when you evaluate your own level of effort and performance account for clustering of standard errors by supervisor. Substantive find-
at work. ings were identical; results are available by request of the first author.
participants had worked for the company and participant gender a negative relationship would exist for high-PPGO individuals and
(1 ⫽ male, 2 ⫽ female) throughout our analyses. Based on a positive relationship would exist for low-PPGO individuals. The
recommendations from goal orientation research, we also con- interaction term for PPGO and the number of perceived higher
trolled for learning goal orientation and performance-avoid goal performers was significantly related to job-based self-efficacy
orientation using Vandewalle’s (1997) measures. The learning (Table 2; b ⫽ ⫺.09, p ⬍ .05). Figure 2 displays this interaction for
goal orientation scale has 5 items; a sample item is “I am willing a low number of perceived higher performers at 0 and a high
to select a challenging work assignment that I can learn a lot from” number of perceived higher performers at 6 (representing 1 stan-
(␣ ⫽ .87). The performance-avoid goal orientation scale consists dard deviation above and below the mean). A test of simple slopes
of 4 items; a sample item is “I prefer to avoid situations at work showed the number of perceived higher performers negatively
where I might perform poorly” (␣ ⫽ .89). related to job-based self-efficacy for individuals high in PPGO
(b ⫽ ⫺.04, p ⬍ .05) and positively related to job-based self-
Results efficacy for low-PPGO individuals (b ⫽ .07, p ⬍ .01), which
offered support for Hypothesis 1.
We first examined the proposed measurement model using

Hypothesis 2 suggested that self-efficacy would positively relate
confirmatory factor analysis, entering the two psychometric vari-

to performance. Job-based self-efficacy was positively and signif-
ables (PPGO and job-based self-efficacy). Results indicated mar-
icantly related to individual performance (Table 2; b ⫽ .81, p ⬍
ginal fit (␹2 ⫽ 43.04, df ⫽ 13, CFI ⫽ .93, RMSEA ⫽ .14,
.01), which offered support for Hypothesis 2.
SRMR ⫽ .08). We observed one item on the PPGO scale that had
a low factor loading (standardized estimate of .45). We thus Hypothesis 3 posited a conditional indirect effect model
analyzed a model without that item (i.e., PPGO measured by three whereby the interactive effects of the number of perceived higher
items), and results indicated a stronger fit (␹2 ⫽ 17.67, df ⫽ 8, performers and goal orientation would influence performance
CFI ⫽ .98, RMSEA ⫽ .10, SRMR ⫽ .04). We therefore conducted through its effect on job-based self-efficacy. We examined this
the remaining analyses using both the three- and four-item mea- model using procedures by Edwards and Lambert (2007) for
sures of PPGO, and substantive conclusions did not vary between testing a first stage only model (e.g., Tepper, Henle, Lambert,
the two procedures. To facilitate comparability with prior PPGO Giacalone, & Duffy, 2008). First, we standardized control vari-
research, we present results based on the published four-item ables. Then, we regressed self-efficacy on the control variables, the
PPGO measure. number of perceived higher performers, PPGO, and the interaction
Table 1 reports the descriptive statistics and correlations. Bi- term, which represents the first stage of the model and test of
variate correlations showed that the number of perceived higher Hypothesis 1. We then regressed job performance on control
performers was not associated with either job-based self-efficacy variables, the number of higher performers, and PPGO. After
(r ⫽ ⫺.01; ns) or performance (r ⫽ ⫺.13; ns), although job-based estimating the regression models, we used Edwards and Lambert’s
self-efficacy positively related to performance (r ⫽ .41; p ⬍ .01). (2007) SPSS macro to bootstrap 1,000 samples to estimate bias-
Table 2 summarizes the regression results. Following guidance corrected 95% confidence intervals (CI). Table 3 displays the
from Carlson and Wu (2012), we report models excluding control results of this analysis, presenting the first stage and indirect paths
variables (Models 1 and 4). Models 2, 3, 5, and 6 include control for the number of higher performers at high and low (⫾ 1 SD)
variables as well as main and interactive effects of the number of levels of PPGO. Because our theoretical model hypothesized a
higher performers and PPGO on job-based self-efficacy and job first-stage moderation model, we constrained the second-stage and
performance. direct effect paths to be invariant across levels of PPGO.
Hypothesis 1 proposed that PPGO moderates the effect of the Results presented in Table 3 support the predicted first-stage
number of perceived higher performers on self-efficacy such that moderation. The relationship between perceived number of higher
Table 1
Study 1 Descriptive Statistics and Correlations
Variable M SD 1 2 3 4 5 6 7 8 9 10 11
1. Prior performance 3.38 0.52

2. Organization tenure 11.32 11.38 .13
3. Gendera 1.60 0.49 .27ⴱⴱ .12
4. Information access 3.63 0.84 ⫺.07 .00 ⫺.04
5. Resource access 3.94 0.63 ⫺.05 ⫺.06 ⫺.20ⴱ .50ⴱⴱ
6. LGO 3.98 0.60 ⫺.19 ⫺.32ⴱⴱ ⫺.17 .28ⴱⴱ .40ⴱⴱ
7. PAGO 2.20 0.66 .04 ⫺.10 .03 ⫺.13 ⫺.30ⴱⴱ ⫺.36ⴱⴱ
8. PPGO 3.30 0.49 ⫺.07 ⫺.31ⴱⴱ ⫺.09 ⫺.01 ⫺.07 .10 .39ⴱⴱ
9. Number of salient coworkers 8.18 6.37 ⫺.01 .14 ⫺.10 ⫺.08 ⫺.02 .07 ⫺.16 .05
10. Number of higher performers 2.71 3.29 ⫺.15 .02 ⫺.05 ⫺.10 ⫺.06 .09 ⫺.12 .10 .62ⴱⴱ
11. Job-based self-efficacy 4.38 0.53 ⫺.04 ⫺.25ⴱⴱ ⫺.18 .20ⴱ .40ⴱⴱ .46ⴱⴱ ⫺.25ⴱⴱ .19ⴱ .03 ⫺.01
12. Job performanceb 3.06 1.05 .78ⴱⴱ ⫺.02 .12 .22ⴱ .19ⴱ ⫺.05 ⫺.05 .09 .01 ⫺.13 .41ⴱⴱ
Note. LGO ⫽ learning goal orientation; PAGO ⫽ performance avoid goal orientation; PPGO ⫽ performance prove goal orientation; No. ⫽ number. N ⫽
119.
a
1 ⫽ male, 2 ⫽ female. b N ⫽ 110.
ⴱ
p ⬍ .05 (two-tailed). ⴱⴱ p ⬍ .01 (two-tailed).
Note. Standard errors reported in parentheses. DV ⫽ dependent variable; LGO ⫽ learning goal orientation; OLS ⫽ ordinary least squares; PAGO ⫽ performance avoid goal orientation; PPGO ⫽
.75ⴱⴱ
.19ⴱⴱ
⫺.19ⴱⴱ
.38ⴱⴱ
4.2
⫺.02
.00
⫺.04
.02
⫺.03
.06
.07
⫺.02
␤
Model 6
3.8
.78ⴱⴱ
Job-Based Self-Efficacy
⫺5.19 (.85)
1.51ⴱⴱ (.11)
.00 (.01)
⫺.08 (.12)
.24ⴱⴱ (.08)
.04 (.12)
⫺.32ⴱⴱ (.12)
⫺.04 (.10)
.01 (.01)
⫺.01 (.02)
.15 (.13)
⫺.01 (.05)
.81ⴱⴱ (.13)
b (SE)
3.4
ⴱ
Low PPGO
3.0 High PPGO
.81ⴱⴱ
DV: Job performanceab
.15ⴱ
.15ⴱ
.01
⫺.08
.09
⫺.07
⫺.08
.06
.00
⫺.11
␤
2.6
Model 5
.69ⴱ
.01 (.01)
.00 (.01)
⫺2.81 (.89)
1.61ⴱⴱ (.13)
.00 (.01)
⫺.17 (.15)
.19ⴱ (.09)
.15 (.09)
⫺.12 (.14)
⫺.12 (.12)
.30ⴱ (.15)
⫺.09 (.05)
2.2
b (SE)
Low Number of Higher High Number of Higher

Performers Performers
Study 1 OLS Regressions of Individual Performance on Job-Based Self-Efficacy and the Number of Perceived Higher Performers
Figure 2. The interactive effect of the number of high performers and

participant performance prove goal orientation (PPGO) on job-based self-
.43ⴱⴱ
efficacy (Study 1). High and low levels of PPGO operationalized at ⫾ 1

⫺.16
.03
.10
␤
SD. Low and high number of higher performers are approximately ⫾ 1 SD

(0 and 6 higher performers, respectively). The simple slope for high-PPGO
Model 4
individuals is negative and significant (b ⫽ ⫺.04, p ⬍ .05); the simple

.19ⴱ
⫺.05 (.03)
(.85)
.06 (.19)
.08 (.08)
.89ⴱⴱ (.19)
slope for low-PPGO individuals is positive and significant (b ⫽ .06, p ⬍

b (SE)
.05).
⫺.83
performers and job-based self-efficacy was positive for low-PPGO

.22ⴱ
.27ⴱ
.20ⴱ
⫺.22ⴱ
.09
⫺.04
⫺.10
⫺.06
.00
⫺.11
.04
individuals (b ⫽ .07, p ⬍ .05; 95% CI [.01, .10]) and negative

␤
for high-PPGO individuals (b ⫽ ⫺.04, p ⬍ .01; 95% CI

Model 3
.37ⴱ
[⫺.06, ⫺.01]). This difference was statistically significant (b ⫽

.00 (.01)
2.87 (.62)
.09 (.09)
.00 (.01)
⫺.10 (.10)
⫺.03 (.06)
.19ⴱ (.09)
.23ⴱ (.09)
⫺.09 (.08)
.00 (.01)
.21ⴱ (.10)
⫺.09ⴱ (.04)
.11, p ⬍ .05; 95% CI [.03, .14]). The first-stage findings carried

b (SE)
ⴱ
through to impact performance such that the indirect effect on job

performance was negative for high-PPGO individuals (b ⫽ ⫺.04,
DV: Job-based self-efficacy
p ⬍ .01; 95% CI [⫺.06, ⫺.01]), but positive for low-PPGO

individuals (b ⫽ .06, p ⬍ .01; 95% CI [.01, .10]). This difference
.23ⴱ
.30ⴱ
.25ⴱ
⫺.10
.06
⫺.05
⫺.04
⫺.06
⫺.15
.07
␤
was statistically significant (b ⫽ .10, p ⬍ .01; 95% CI [.04, .16]).

Model 2
These results offered support for Hypothesis 3.

.34ⴱ
.01 (.01)
⫺.02 (.02)
2.82 (.64)
.06 (.10)
.00 (.01)
⫺.05 (.10)
⫺.04 (.06)
.20ⴱ (.10)
.25ⴱ (.10)
⫺.11 (.08)
.26ⴱ (.10)
b (SE)
Supplementary Analysis and Robustness Checks

ⴱ
performance prove goal orientation; Perfs. ⫽ performers. N ⫽ 119.
Research has pointed out the importance of ego-referent simi-

larity in how upward comparisons affect individuals (Tesser,
.07
.15
⫺.29
1988). We argued above that evaluating similarity across multiple

␤
simultaneous comparisons is cognitively complex. Although ex-

Model 1
isting research emphasizes ego-referent similarity as a key mod-

.11
.01 (.02)
.16 (.10)
4.40 (.05)
⫺.11ⴱ (.04)
erator of social comparison effects, our intent was instead to focus

b (SE)
on understanding PPGO as an individual difference in information

ⴱ
processing across perceptions of multiple higher performers. Con-

sequently, we did not treat ego-referent similarity as a focal con-
1 ⫽ male, 2 ⫽ female. b N ⫽ 110.
struct, but did account for similarity in several ways.

Number of Higher Perfs. ⫻ PPGO
Most importantly, we collected data that allowed respondents to

Number of higher performers
Number of salient coworkers
self-report their salient others at work. Given that similarity plays

a critical role in how referent others are selected (Kulik & Am-
Job-based self-efficacy
brose, 1992; Shah, 1998), the salient others in our study have
Variable
Organization tenure
Information access
already been deemed similar (or at least relevant) by the focal

Prior performance
Resource access
individual. That is, the referents selected have already overcome

the similarity hurdle at least to some degree because they were
Intercept
cognitively salient to respondents. However, we acknowledge that

Table 2
Genderb
PAGO
PPGO
among these salient coworkers, more similar peers (e.g., same job
LGO
R2
or department) might be more meaningful (and thus have a stron-

ⴱ
a
Table 3
Study 1 Conditional Indirect Effects of the Number of Perceived Higher Performers on Job Performance Through Job-Based
Self-Efficacy
Path Low PPGO High PPGO Difference
First stage effect on self-efficacy

Number of Higher Performers ⫻ PPGO ¡ Job-based self-efficacy .07ⴱ [.01, .10] ⫺.04ⴱ [⫺.06, ⫺.01] .11ⴱ [.03, .14]
Conditional indirect effect on performance
Number of Higher Performers ⫻ PPGO ¡ Job-based self-efficacy ¡ Job Performance .06ⴱ [.01, .10] ⫺.04ⴱ [⫺.06, ⫺.01] .10ⴱⴱ [.04, .16]
Note. PPGO ⫽ performance prove goal orientation. High and low levels of PPGO operationalized at ⫾ 1 SD. 95% bias-corrected confidence intervals
are represented in brackets. All control variables from regression analyses are entered. N ⫽ 110.
ⴱ
ger effect) than less similar peers. To address this, we coded the which the number of perceived higher performers could be exper-
number of perceived higher performers who were in the same job. imentally manipulated.
Averaged across respondents, about 21% of all listed coworkers
had the same job title as the focal respondent. We included this
variable as a moderator, testing a three-way interaction between Study 2
number of higher performers, same-job higher performers, and
PPGO. There were no two- or three-way interactive effects on Method
self-efficacy in this model. Third, one potential explanation for our
effects is that observer-peer similarity operates not as a moderator Sample and procedure. We recruited 128 individuals from a
but as a confounding variable. To probe this possibility, we in- behavioral business laboratory participant pool at a large public
cluded same-job peers as covariates; conclusions were unaffected U.S. university (Approved by University of Arkansas Institutional
(available in the online supplemental materials). Review Board [Protocol Number 1903181360, Title: “High Per-
former Comparison Study”]). In exchange for participating, each
received $12.00. Participants were primarily undergraduates
Discussion
(80%), but also included university staff (8.6%), graduate students
Study 1 demonstrated a negative relationship between the per- (7%), alumni (2.3%), and university faculty (2.3%). After arriving,
ceived number of higher performers and self-efficacy for individ- participants were led to a computer lab and seated at workstations,
uals high in PPGO, and a positive relationship for individuals low where we secured voluntary consent and explained that they would
in PPGO. There are several strengths of this study. First, employ- complete two product marketing tasks. The number of participants
ees in the sample worked in a context where working adults had in each session ranged from a low of five to a high of 20 (M ⫽
substantial in-person interaction with one another and could ob- 12.47; median ⫽ 13). Participants responded to the goal orienta-
serve others’ performance. Second, we employed a two-step de- tion scale and, then completed the first of two 10-min performance
sign to measure the number of perceived higher performers. Spe- rounds. In the first round, participants were asked to design an
cifically, focal individuals identified a salient subset of coworkers, advertisement to sell socks, using a single Microsoft PowerPoint
which recognizes that not all coworkers in a social context are slide. They were told that their advertisements would be rated by
equally important for the self-evaluative and instructive cognitive other study participants on five criteria: (a) eye-catching, (b)
functions. Respondents then evaluated the higher performance of creative, (c) persuasive, (d), overall quality, and (e) likelihood of
only this salient subset, ensuring that the respondent was aware of making a sale.
and cared about others’ higher performance. This measurement After the first round, they were prompted to describe the strat-
strategy corresponds with both social comparison and social– egies they used in their design in an open text entry field. They
cognitive theories, which place the perceived performance envi- were then presented with four advertisements that were feigned to
ronment as more proximal to individual outcomes than others’ be from other participants who were with them in the room, along
veridical performance. Third, variables were time-separated and with the strategies each of these peers used to complete their
obtained from different sources to reduce potential common design. The peer advertisements and strategies were in reality
method bias and strengthen the test of the causal claims of our developed by the research team (described below). Participants
theory. evaluated each advertisement (including their own), and then com-
These strengths notwithstanding, it remains plausible that high- pleted the self-efficacy scale. They were told that the second-round
PPGO individuals with high levels of self-efficacy simply perceive advertisement would be for a different product. That product
a smaller number of higher performers in their environment than (toothbrushes) was revealed to them after they completed the
do high-PPGO individuals with low levels of self-efficacy. This self-efficacy scale. Of the 128 participants, 21 failed an item
view accords with our theory that high- and low-PPGO individuals designed to catch careless responding (i.e., “if you are reading
interpret their social environments differently, but it offers a po- carefully, please select slightly disagree for this item”; Meade &
tential reverse-causal explanation where the combination of self- Craig, 2012) and were removed, resulting in an eligible sample
efficacy and PPGO impacts employees’ perceptions of the number size of 107.
of high performers. To examine this alternative and establish Manipulation. After the first round of advertisement design,
causal order for our model, we conducted an experimental study in participants were randomly assigned to one of two conditions. In
the experimental condition, participants viewed one high- the eight participants in the experimental condition who rated their
performing peer advertisement and corresponding strategies and own advertisement as better than the higher performer. We then
three low-performing peer advertisements. In the control condi- conducted a second robustness check removing 13 participants
tion, participants viewed four low-performing peer advertisements. (total across both conditions) who failed the manipulation check.
Participants were then asked to rate each design along the five Because results in either check did not differ in terms of substan-
criteria listed above (used later as a manipulation check). Prior to tive conclusions, results reported below include the full sample
beginning the study, the peer advertisements were pilot tested with (n ⫽ 107).
a similar sample to ensure designs reflected intended levels of high
and low performance.
Results
Measures.
Goal orientations. We again measured PPGO, LGO, and We again conducted a CFA of PPGO and self-efficacy to
PAGO using Vandewalle’s (1997) scales (␣ ⫽ .73, .83, and .82, evaluate our measurement model. Results indicated a strong fit
respectively). (␹2 ⫽ 14.91, df ⫽ 13, CFI ⫽ .99, RMSEA ⫽ .04, SRMR ⫽ .05).
Task-based self-efficacy. Because job- and task-based self- Tables 4, 5, and 6 present experimental results in parallel to Study
efficacy differ in scope, we assessed self-efficacy using Tierney 1. The experimental condition did not significantly increase or
and Farmer’s (2002) three-item scale, adapted to the task, to decrease task-based self-efficacy (Mc ⫽ 3.31, Me ⫽ 3.20; ns), but
anchor self-efficacy ratings appropriately to the context. A sample did significantly increase task performance (Mc ⫽ 2.69, Me ⫽
item is, “I feel confident in my ability to design the next adver- 2.87; t ⫽ 2.74; p ⬍ .01). Task-based self-efficacy positively
tisement” (␣ ⫽ .84). related to performance (r ⫽ .25 p ⬍ .01).
Task performance. Two research assistants, blind to study Table 5 shows the interaction of PPGO with the experimental
conditions and hypotheses, independently rated participants’ ad- condition was significantly related to task-based self-efficacy
vertisements using the same five criteria used by participants, (Model 3; b ⫽ ⫺.46, p ⬍ .05). Figure 3 plots this interaction
coding both the sock (Time 1) and toothbrush (Time 2) advertise- including mean differences. Working with a high-performing peer
ments. As raters achieved interrater agreement in both rounds (vs. no high-performing peer) significantly decreased task-based
(ICC[2]: T1 ⫽ .83, T2 ⫽ .98), we averaged their ratings. self-efficacy for participants high in PPGO (t ⫽ ⫺2.10; p ⬍ .05)
Manipulation check. To ensure the manipulation operated as but had no significant effect on task-based self-efficacy for par-
intended, we examined participant ratings of the high-performing ticipants low in PPGO (t ⫽ 1.31; ns). These findings offered
peer advertisements across the same five criteria as the perfor- partial support for Hypothesis 1.
mance variable. Items were averaged to form a performance index Hypothesis 2 suggested that task-based self-efficacy would pos-
for each “peer.” An independent samples t test confirmed that itively relate to performance. As Table 5 shows, task-based self-
participants in the experimental condition rated the focal high- efficacy was positively and significantly related to individual
performing peer as having a better advertisement (M ⫽ 4.19, SD ⫽ performance (Table 5, Model 6; b ⫽ .15, p ⬍ .01), which sup-
.72) than the average low-performing peer in the control condition ported Hypothesis 2.
(M ⫽ 2.62, SD ⫽ .67; t ⫽ 11.72, p ⬍ .01). This result suggests our Examination of Hypothesis 3 followed the same procedures
manipulation was successful in inducing higher perceived perfor- as Study 1 after dummy-coding conditions (0 ⫽ control; 1 ⫽
mance. As additional robustness checks, we examined partici- experimental). The first stage effect (high performers on self-
pants’ self-ratings of their own advertisements against their ratings efficacy) was negative and significant for high-PPGO individ-
of the higher performing peer. The majority of participants in the uals (b ⫽ ⫺.51, p ⬍ .05; 95% CI [⫺.95, ⫺.02]), as was the
experimental condition (46 of 54) rated the manipulated peer indirect effect on task performance through task-based self-
advertisement better than they rated their own, as we intended. efficacy (b ⫽ ⫺.07, p ⬍ .05; 95% CI [⫺.19, ⫺.02]). For low-
Further, the majority of participants in the control condition (48 of PPGO individuals, the first stage effect was positive, but not
53) rated the manipulated peer advertisement as worse than they significant (b ⫽ .28, ns; 95% CI [⫺.25, .85]), although the indirect
rated their own. We conducted a first robustness check removing effect for low-PPGO individuals was positive and approached
Table 4
Study 2 Descriptive Statistics and Correlations
Variable M SD 1 2 3 4 5 6
1. Time 1 task performance 3.07 0.62

2. PAGO 2.78 0.89 .04
3. LGO 4.06 0.63 ⫺.08 ⫺.12
4. PPGO 3.65 0.83 .06 .46ⴱⴱ .17
5. Higher performer conditiona 0.50 0.50 .03 ⫺.14 .11 .05
6. Self-efficacy 3.25 0.98 ⫺.06 .20ⴱ .11 .27ⴱⴱ ⫺.06
7. Time 2 task performance 2.78 0.45 .18 ⫺.05 ⫺.15 .00 .20ⴱ .25ⴱⴱ
Note. LGO ⫽ learning goal orientation; PAGO ⫽ performance avoid goal orientation; PPGO ⫽ performance
prove goal orientation. N ⫽ 107.
a
1 ⫽ higher performer present, 0 ⫽ no higher performer present.
ⴱ
Table 5
Study 2 OLS Regressions of Study Condition on Task-Based Self-Efficacy and Subsequent Task Performance
DV: Task-based self-efficacy DV: T2 task performance

Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
Variable b (SE) ␤ b (SE) ␤ b (SE) ␤ b (SE) ␤ b (SE) ␤ b (SE) ␤
ⴱ ⴱ ⴱ ⴱ ⴱ ⴱ
Intercept 3.33 (.13) 2.79 (.94) 2.78 (.93) 2.25 (.16) 2.96 (.43) 2.55 (.43)
T1 perf. ⫺.11 (.15) ⫺.07 ⫺.07 (.15) ⫺.04 .12 (.07) .16 .13 (.07) .18
PAGO .12 (.12) .11 .11 (.10) .10 ⫺.04 (.06) ⫺.08 ⫺.06 (.05) ⫺.11
LGO .13 (.15) .08 .11 (.15) .07 ⫺.13 (.07) ⫺.18 ⫺.15ⴱ (.07) ⫺.21ⴱ
Number of higher perfsa ⫺.14 (.18) ⫺.07 ⫺.12 (.19) ⫺.06 ⫺.12 (19) ⫺.06 .20ⴱ (.08) .22ⴱ .18ⴱ (.09) .20ⴱ .20ⴱ (.08) .22ⴱ
PPGO .54ⴱⴱ (.15) .45ⴱⴱ .25 (.13) .21 .46ⴱⴱ (.17) .39ⴱⴱ ⫺.06 (.07) ⫺.11 .06 (.08) .11 ⫺.01 (.08) ⫺.02
Number of Higher Perfs. ⫻ PPGO ⫺.49ⴱ (.22) ⫺.27ⴱ ⫺.46ⴱ (.23) ⫺.26ⴱ .03 (.10) .04 ⫺.07 (.11) ⫺.09 .00 (.10) .00
Task-based self-efficacy .13ⴱ (.05) .29ⴱ .15ⴱⴱ (.04) .32ⴱⴱ
ⴱ ⴱ ⴱ
R 2
.12 .09 .13 .12 .10 .19ⴱ
Note. Standard errors reported in parentheses. DV ⫽ dependent variable; LGO ⫽ learning goal orientation; OLS ⫽ ordinary least squares; PAGO ⫽
performance avoid goal orientation; PPGO ⫽ performance prove goal orientation; Perf. ⫽ performance; Perfs. ⫽ performers. N ⫽ 107.
a
1 or 0.
ⴱ
significance (b ⫽ .04, ns; 90% CI [.00, .14]). Both the first stage learning and integrating peer strategies into their own repertoire
(b ⫽ .78, p ⬍ .05; 95% CI [.02, 1.50]) and the indirect effect (b ⫽ (Myers, 2018).
.12, p ⬍ .05; 95% CI [.03, .30]) were significantly different across Results of this analysis, presented in Table 7, demonstrate an
low and high levels of PPGO. These results offered partial support interaction of PPGO on the relationship between the high per-
for Hypothesis 3. former condition and the amount of time participants reviewed
peer strategies (b ⫽ ⫺7.46, p ⬍ .01). The nature of this interaction
Supplementary Analysis Related to Learning is portrayed in Figure 4, which shows that for low-PPGO individ-
uals, the presence of a high performer increased the amount of time
Although we did not hypothesize the mechanisms through participants spent reviewing peer strategies (b ⫽ 7.33, p ⬍ .05). In
which higher performers influence self-efficacy, our theory— contrast, high-PPGO participants spent less time reviewing peer
drawing from Bandura (1997) and Wood (1989)—implies that strategies in the high performer condition, although this negative
high-performing peers initiate two cognitive functions: instruction simple slope was not significant (b ⫽ ⫺5.05, ns). The fact that
and self-evaluation. We could not perceptually measure these high-PPGO individuals did not spend more time reviewing higher
mechanisms in Study 2 without creating a demand effect (Orne, performers work products and strategies is consistent with our
1962) in which participants would be prompted for learning before theory that they devote fewer cognitive resources to observational
evaluating their self-efficacy. Such a procedure would have threat- learning and therefore miss out on the self-efficacy gains that
ened the internal validity of our study by potentially sensitizing accompany learning from role models.
participants toward our hypotheses. We did, however, include a
passive measurement of the amount of time participants spent
Discussion
viewing peer strategies. Time spent on an activity has frequently
been used by researchers as an indicator of the amount of personal The primary purpose of Study 2 was to offer a stronger test of
resources (i.e., effort, attention) expended toward some task the causal ordering of our hypothesized model. Because partici-
(Mitchell & Nebeker, 1973; Northcraft, Schmidt, & Ashford, pants were randomly assigned to conditions, there is less concern
2011; Weiss & Sherman, 1973). In our particular application, (relative to Study 1) that the number of perceived higher perform-
spending more time reviewing peer strategies provides a reason- ers were chosen on the basis of participants’ self-efficacy and
able indicator of participants’ attention and effort devoted to PPGO. Study 2 also explicitly tests the hypotheses with a focus on
Table 6
Study 2 Conditional Indirect Effects of Number of Higher Performers Manipulation on Task Performance Through Task Self-Efficacy
Path Low PPGO High PPGO Difference
First stage effect on self-efficacy

Number of Higher Performers ⫻ PPGO ¡ Task-based self-efficacy .28 [⫺.25, .85]a ⫺.51ⴱ [⫺95, ⫺.02]a .78ⴱ [.02, 1.50]b
Conditional indirect effect on performance
Number of Higher Performers ⫻ PPGO ¡ Task-based self-efficacy ¡ Task performance .04 [⫺.02, .15]b ⫺.07ⴱⴱ [⫺.19, ⫺.02]b .12ⴱ [.03, .30]b
Note. Coefficients are unstandardized. PPGO ⫽ performance prove goal orientation. High and low levels of PPGO operationalized at ⫾ 1 SD. N ⫽ 107.
a
First stage effects significance tests are based on simple slope standard errors. b Indirect and difference effects tested using 95% bias-corrected
confidence intervals from Edwards & Lambert bootstrap. All control variables from regression analyses are entered into both equations.
ⴱ
4.2 45
Time Spent on Peer Strategies

3.8 40
Task-Based Self-Efficacy
3.4 35
(seconds)
Low PPGO Low PPGO
3.0 High PPGO High PPGO
30
2.6
25
2.2
Low Number of Higher High Number of Higher 20

Performers Performers No High Performer High Performer Present
Figure 3. The interactive effect of the number of high performers and Figure 4. Supplementary analysis of first stage moderation of the number
participant performance prove goal orientation (PPGO) on task-based of higher performers on time spent reviewing peer strategies (Study 2).
self-efficacy (Study 2). High and low levels of PPGO operationalized at PPGO ⫽ performance-prove goal orientation. High and low levels of
⫾1 SD. For participants high in PPGO, task-based self-efficacy was PPGO operationalized at ⫾ 1 SD.
significantly lower when they worked in a group with a higher performer
compared with those who did not work with a higher performer.
Finally, Study 2 manipulated the veridical performance envi-
ronment such that the true underlying quality of coworkers’ per-
a specific task, complementing the field sample in Study 1 where formance differed across conditions. This manipulation carried
our hypotheses were assessed at the broader level of the job as a set through to participants’ perceived performance environment (as
of tasks that may differ across employees. One additional benefit evidenced by the manipulation check). This design feature com-
to this approach is that it enabled us to test our hypotheses in the plements Study 1 in that it allows us to verify that the perceived
context of a knowledge work task, where role modeling occurs as performance environment in Study 2 is rooted in veridical differ-
performance strategies are explicitly articulated rather than di- ences in performance rather than bias in how individuals with
rectly observed. As social– cognitive theory points out, observa- different combinations of PPGO and self-efficacy perceive others’
tional learning is not merely a function of “behavioral mimicry” performance.
(Bandura, 1997, p. 93), but instead can occur as observers’ witness
the product of higher performance and hear (or in our study, read) General Discussion
the verbalized strategies that led to this higher performance. This Perceiving higher performers could be detrimental or beneficial
is consistent with Bandura’s premise that “models do not behave to the motivation and performance of observers. Existing theories
like mute automations” (p. 88). In this way, one advantage of offer different predictions about how these effects emerge in the
Study 2 was that it utilized a task that enabled us to explicitly test workplace. Through the lens of social comparison theory, per-
this verbalized form of observational learning beyond rote emula- ceived higher performers represent self-evaluative referents, lead-
tion of behaviors. ing to reduced job-based self-efficacy and job performance.
Through the lens of social– cognitive theory, perceived higher
performers represent instructive role models who can elevate job-
Table 7 based self-efficacy and job performance. We developed theory
Study 2 Cognitive Mechanism Supplementary Analysis linking these perspectives by describing self-evaluation and learn-
ing as alternative cognitive processes individuals adopt when
DV: Time spent on peer interpreting their environments of higher performers. Because
strategies PPGO plays a role in these cognitive processes, we hypothesized
Variable b (SE) ␤ it would moderate the relationship between perceived higher per-
formers and self-efficacy.
Intercept 49.89ⴱⴱ (9.88)
PAGO ⫺3.00 (1.55) ⫺.21 Our results were consistent with this general proposition. In
LGO ⫺1.98 (1.93) ⫺.10 Study 1, we found that for high-PPGO individuals, the number of
Number of higher perfs.a 1.14 (2.37) .05 perceived higher performers was negatively related to job-based
PPGO 2.62 (2.11) .17 self-efficacy and job performance. In contrast, for low-PPGO
Number of Higher Perfs. ⫻ PPGO ⫺7.46ⴱⴱ (2.85) ⫺.33ⴱⴱ
R2 .12ⴱ
individuals, the number of perceived higher performers was pos-
itively related to job-based self-efficacy and job performance. In
Note. Standard errors in parentheses. Perf. ⫽ performers; LGO ⫽ learn- Study 2, we again examined the model using an experimental
ing goal orientation; PAGO ⫽ performance avoid goal orientation;
PPGO ⫽ performance prove goal orientation. N ⫽ 107.
design that allowed for a stronger test of this causal ordering.
a
1 higher performer present, 0 no higher performer present. Consistent with Study 1, the presence of a higher performing peer
ⴱ
p ⬍ .05 (two-tailed). ⴱⴱ p ⬍ .01 (two-tailed). had a significant, negative influence on the self-efficacy of high-
PPGO participants. For low-PPGO participants, the presence of a learning and hamper self-efficacy for certain kinds of individuals.
higher performing peer had a positive influence on self-efficacy; Our findings are consistent with this proposition.
however, this relationship failed to reach significance. Although it Our research also contributes to understanding of PPGO in work
is difficult to diagnose exactly why the low PPGO relationship was contexts. Scholars have noted that PPGO should be positively
not fully replicated in Study 2, several features of the Study 2 task related to social comparisons in general (Régner et al., 2007), and
and sample may be muting effects in the lab that have time to that PPGO influences the quality of relationships individuals form
emerged in the field. First, compared with the field, participants in at work (Janssen & Van Yperen, 2004). These conform to the
the lab have only a brief performance episode to engage in obser- existing narrative that PPGO influences individuals’ motivation
vational learning and the richness of those learning opportunities and actions across many situations (Payne et al., 2007; Phillips &
are restricted, which may also restrict growth in their self-efficacy. Gully, 1997). We extend this view by articulating how and why
Because Study 1 took place in the field, participants had numerous PPGO shapes individual responses to the performance environ-
opportunities and an extended time period to learn from their ment. In doing so, we position PPGO not as an antecedent of
peers. In contrast, the nature of the Study 2 laboratory experiment individuals’ social behaviors, but rather as a pattern of information
offered only a few minutes to learn from higher performing peers. processing impacting how they respond to social cues that arise.
Second, we were also likely advantaged to unearth a stronger Viewing PPGO as a condition for how individuals respond to
positive relationship in the field because, as theorized, beyond just social situations may be valuable in explaining the nuanced and
the presence of a high performer (v. none) a linear relationship sometimes contradictory effects of PPGO on work performance
exists such that the more high performers that employees are (e.g., Dietz et al., 2015; Yeo, Loft, Xiao, & Kiewitz, 2009).
around, the more they are likely to learn from and model—if they
are low on PPGO. Nevertheless, in both studies social comparison
Practical Implications
theory better described outcomes of perceiving higher perfor-
mance for individuals high in PPGO, and social– cognitive theory Our research carries a number of implications for practice. Most
better described outcomes for individuals low in PPGO. The importantly, managers should tamp down performance-prove goal
findings offer several implications for organizational research and mindsets in situations where individuals are likely to make self-
practice. evaluative comparisons. For example, managers should avoid pre-
senting relative performance feedback to high-PPGO individuals,
or make efforts to present relative performance feedback in ways
Theoretical Implications
that do not threaten employees’ desires to be seen as competent. As
Our research directly extends social comparison theory. First, negative relative performance feedback can be particularly detri-
we tighten the connection between social comparisons and job mental to high-PPGO individuals, managers could consider simul-
performance focusing on job-based self-efficacy. A great deal of taneously delivering additional information that can boost those
research on social comparison theory has examined the broader individuals’ self-efficacy. This might include jointly presenting a
self-concept (e.g., Suls et al., 2002), which is distal to job perfor- review of goals they have achieved or highlighting ways that they
mance in comparison to job-based self-efficacy. Our research exceeded expectations based on their own previous performance
advances social comparison theory by linking comparisons to job level. Another potential strategy to manage high-PPGO individu-
performance through job-based self-efficacy, tightening the con- als’ self-efficacy is to rely on absolute, rather than relative, per-
nection between social comparisons and job performance. Second, formance information when delivering feedback. Absolute feed-
we conceptualize social comparisons as stemming from the num- back could focus on how actual performance deviated from
ber of salient higher performers a person has at work. This is a expected performance, or more simply on the raw achievements of
notable departure from existing social comparison research, which the high-PPGO individual. Importantly, delivering feedback to
has often measured or experimentally manipulated a single social high-PPGO individuals in ways that avoid relative performance
comparison in isolation. This perspective leads our theory away information may mitigate the adverse effect of upward compari-
from dyadic moderators of the outcomes of upward comparisons sons on employees’ self-efficacy.
(i.e., similarity between the observer and the referent, Tesser, These recommendations are consistent with the feedback liter-
1988) and toward a more general cognitive pattern of processing ature (Kluger & DeNisi, 1996, 1998), which advises managers to
social comparison information: goal orientation. By considering avoid normative comparisons and focus feedback on tasks rather
general patterns across multiple comparisons, our theory offers an than on social comparisons with others. However, our study offers
extension to social comparison outcomes. two implications beyond these recommendations. First, our study
Our study also contributes to social– cognitive theory by iden- points out that normative feedback is actually beneficial for self-
tifying conditions in which higher performers reduce others’ self- efficacy and performance when individuals have lower levels of
efficacy. This extends social– cognitive theory, which has tradi- PPGO. Our study suggests that for these low-PPGO individuals,
tionally argued that role modeling should promote self-efficacy, normative feedback can initiate learning the performance strate-
provided the observer pays enough attention to integrate observed gies for enhanced performance. Second, we note that employees
performance strategies into their behavioral repertoire. When ob- self-administer feedback through their observations of others. In
servers do not pay attention, social– cognitive theory suggests feedback intervention theory (Kluger & DeNisi, 1996), employees
there should be no learning, and there should be no effect on are often considered passive; they wait for the organization to
self-efficacy (Bandura, 1997). Our theory extends this by explain- initiate feedback and then act. Our account highlights employees’
ing how the effect not only can vary, but that self-evaluative proactive attunement to coworkers to determine how they are
cognitions stemming from observing higher performers thwart performing. This is notably different from a formal feedback
intervention, because every coworker represents an opportunity to Second, we did not collect data about how managers influence
learn or compare. Practically speaking, this means that managers comparative and learning processes as salient role models or
need to go further to guide employees’ interpretations of the referents. To examine the extent to which our findings were the
performance environment even outside of a formal feedback con- result of managers being listed as salient coworkers, we identified
text. whether the focal individual listed the supervisor as a higher
Because this process appears to be self-managed without orga- performer in Study 1. We excluded this comparison for these 39
nizational prompting, a practical implication for individuals is that respondents (33% of the sample) and reanalyzed the data. Al-
they should be mindful when comparing themselves to better though substantive conclusions were identical, it seems plausible
performers. If they are low-PPGO, individuals could benefit from that perceiving a supervisor as higher-performing is fundamentally
seeking out a large number of higher performers to observe, as different than perceiving peers as higher-performing.
those observations can serve as catalysts for higher self-efficacy Finally, we applied a relatively simple operationalization of
and performance. In contrast, if they are high-PPGO, such self- higher performers by summing the number of salient higher per-
evaluative comparisons are likely to deflate their own sense of formers an observer named. Given the complex nature of social
self-efficacy. To the extent possible, high-PPGO individuals contexts, future research might offer alternative approaches that
should avoid relative comparisons to preserve their self-efficacy. capture salient features of employees’ social environments. One
Finally, our research suggests an opportunity for organizations potential way to represent an individual’s pattern of higher per-
to design staffing plans that capitalize on the motivational effects formers would be to measure the proportion of total salient co-
of upward comparisons. Processes could be designed with sensi- workers that are higher performing. We tested this operationaliza-
tivity toward workers’ goal orientations. For example, organization but found no main or moderated effects on self-efficacy.
tions could schedule shiftwork in ways that facilitate low-PPGO However, it remains plausible that the proportion, rather than the
employees to cowork with as many top performers as possible. In raw number, of perceived higher performers influences self-
contrast, schedules might be designed to limit high-PPGO employ- efficacy perceptions. Perhaps maintaining a balanced portfolio of
ees’ exposure to top performers. The findings could also be applied comparisons is important to self-efficacy. Furthermore, it may be
in the areas of team composition, rotational training programs, and that some comparisons with specific individuals carry more influ-
the layout of physical spaces in managing the appropriate oppor- ence than other comparisons. For example, individuals’ other types
tunities for interaction with higher performers at work given em- of relationships (e.g., rivals, friends, advisors) may weight social
ployees’ goal orientation. comparisons. More generally, future research might identify the-
Although such initiatives may be difficult to manage on an oretically relevant (and potentially complex) ways people make
individual-employee basis, certain departments, units, or organi- sense of a higher performing set of peers.
zations may have employees with generally lower or higher goal The social context plays a pivotal role in how individuals
orientations. For example, sales units may have generally higher evaluate themselves and learn to perform their jobs at work. For
levels of PPGO than training or operational units. Such differences low-PPGO individuals, the social context is an environment for
can be considered in programs that highlight top performers. Units learning from role models to improve their own performance.
with primarily high-PPGO individuals (e.g., sales organizations) For high-PPGO individuals, the social context is an environment
should tread lightly when planning opportunities that encourage of competitive referents for whom performance is on display.
employees to compare themselves to the very top performers. In Social comparison and social– cognitive theories describe similar
this context, the average performer, because of her performance aspects of workers’ social context, depending on how employees
goal orientation, is likely to experience a reduced sense of self- interpret and respond to these social worlds. Our research shows
efficacy and motivation from such comparisons. These initiatives how the social context of work influences employees differently
may be more successful in units with lower levels of PPGO. by shaping their internal beliefs about themselves and ultimately
their performance on tasks and jobs.
Limitations and Future Research
Although there are a number of strengths of the present research, References
including data collected over multiple time points and sources Aguinis, H., & O’Boyle, E., Jr. (2014). Star performers in twenty-first
using experimental and field designs, there are limitations. First, century organizations. Personnel Psychology, 67, 313–350. http://dx.doi
Study 1 may have prompted social comparisons which respondents .org/10.1111/peps.12054
may not otherwise have conducted. If our survey unduly prompted Bandura, A. (1965). Influence of models’ reinforcement contingencies on
social comparisons, our findings may be overstated. We mitigated the acquisition of imitative responses. Journal of Personality and Social
this concern by separating the referent identification from refer- Psychology, 1, 589 –595. http://dx.doi.org/10.1037/h0022070
ents’ performance evaluation into different steps in the first Study Bandura, A. (1977). Social learning theory. Englewood Cliffs, NJ: Prentice
Hall.
1 survey. While recruiting participants, we described the purpose
Bandura, A. (1986). Social foundations of thought and action. Englewood
of the study was to understand the ways employees work with
Cliffs, NJ: Prentice Hall.
coworkers to avoid artificially cuing comparison processes. How- Bandura, A. (1997). Self-efficacy: The exercise of control. New York, NY:
ever, we cannot rule out the possibility that our survey was a cause Freeman and Company.
of, as well as an evaluation of, workers’ social comparisons. Future Bandura, A. (2012). On the functional properties of perceived self-efficacy
research may benefit from using passive measurement to avoid revisited. Journal of Management, 38, 9 – 44. http://dx.doi.org/10.1177/
these issues. 0149206311410606
Bandura, A. (2015). On deconstructing commentaries regarding alternative Dweck, C. S. (1986). Motivational processes affecting learning. American
theories of self-regulation. Journal of Management, 41, 1025–1044. Psychologist, 41, 1040 –1048. http://dx.doi.org/10.1037/0003-066X.41
http://dx.doi.org/10.1177/0149206315572826 .10.1040
Bandura, A., & Locke, E. A. (2003). Negative self-efficacy and goal effects Edwards, J. R., & Lambert, L. S. (2007). Methods for integrating moder-
revisited. Journal of Applied Psychology, 88, 87–99. http://dx.doi.org/ ation and mediation: A general analytical framework using moderated
10.1037/0021-9010.88.1.87 path analysis. Psychological Methods, 12, 1–22. http://dx.doi.org/10
Brockner, J. (1988). Self-esteem at work. Lexington, MA: Lexington .1037/1082-989X.12.1.1
Books. Elliot, A. J. (2005). A conceptual history of the achievement goal construct.
Button, S. B., Mathieu, J. E., & Zajac, D. M. (1996). Goal orientation in In A. J. Elliot & C. S. Dweck (Eds.), Handbook of competence and
organizational research: A conceptual and empirical foundation. Orga- motivation (pp. 52–72). New York, NY: Guilford Press.
nizational Behavior and Human Decision Processes, 67, 26 – 48. http:// Elliot, A. J., & Church, M. A. (1997). A hierarchical model of approach
dx.doi.org/10.1006/obhd.1996.0063 and avoidance achievement motivation. Journal of Personality and
Social Psychology, 72, 218 –232. http://dx.doi.org/10.1037/0022-3514
Buunk, A. P., & Gibbons, F. X. (2007). Social comparison: The end of a
.72.1.218
theory and the emergence of a field. Organizational Behavior and
Festinger, L. (1954). A theory of social comparison processes. Human

Human Decision Processes, 102, 3–21. http://dx.doi.org/10.1016/j
Relations, 7, 117–140. http://dx.doi.org/10.1177/001872675400700202

.obhdp.2006.09.007
Gist, M. E., & Mitchell, T. R. (1992). Self-efficacy: A theoretical analysis
Buunk, B. P., Zurriaga, R., Gonzalez-Roma, V., & Subirats, M. (2003).
of its determinants and malleability. The Academy of Management
Engaging in upward and downward comparisons as a determinant of
Review, 17, 183–211. http://dx.doi.org/10.5465/amr.1992.4279530
relative deprivation at work: A longitudinal study. Journal of Vocational
Harter, S. (1990). Causes, correlates, and the functional role of global
Behavior, 62, 370 –388. http://dx.doi.org/10.1016/S0001-8791(02) self-worth: A life-span perspective. In R. J. Sternberg & J. Kolligan
00015-5 (Eds.), Competence considered (pp. 67–97). New Haven, CT: Yale
Call, M. L., Nyberg, A. J., & Thatcher, S. M. (2015). Stargazing: An University Press.
integrative conceptual review, theoretical reconciliation, and extension Heggestad, E. D., & Kanfer, R. (2005). The predictive validity of self-
for star employee research. Journal of Applied Psychology, 100, 623– efficacy in training performance: Little more than past performance.
640. http://dx.doi.org/10.1037/a0039100 Journal of Experimental Psychology: Applied, 11, 84 –97. http://dx.doi
Campbell, E. M., Liao, H., Chuang, A., Zhou, J., & Dong, Y. (2017). Hot .org/10.1037/1076-898X.11.2.84
shots and cool reception? An expanded view of social consequences for Hogg, M. A. (2000). Social identity and social comparison. In J. Suls & L.
high performers. Journal of Applied Psychology, 102, 845– 866. http:// Wheeler (Eds.), Handbook of social comparison: Theory and research
dx.doi.org/10.1037/apl0000183 (pp. 401– 421). New York, NY: Kluwer/Plenum Press. http://dx.doi.org/
Carlson, K. D., & Wu, J. (2012). The illusion of statistical control: Control 10.1007/978-1-4615-4237-7_19
variable practice in management research. Organizational Research Janssen, O., & Van Yperen, N. W. (2004). Employees’ goal orientations,
Methods, 15, 413– 435. http://dx.doi.org/10.1177/1094428111428817 the quality of leader-member exchange, and the outcomes of job per-
Cialdini, R. B., Borden, R. J., Thorne, A., Walker, M. R., Freeman, S., & formance and job satisfaction. Academy of Management Journal, 47,
Sloan, L. R. (1976). Basking in reflected glory: Three (football) field 368 –384.
studies. Journal of Personality and Social Psychology, 34, 366 –375. Jensen, J. M., Patel, P. C., & Raver, J. L. (2014). Is it better to be average?
http://dx.doi.org/10.1037/0022-3514.34.3.366 High and low performance as predictors of employee victimization.
Cohen-Charash, Y., & Mueller, J. S. (2007). Does perceived unfairness Journal of Applied Psychology, 99, 296 –309. http://dx.doi.org/10.1037/
exacerbate or mitigate interpersonal counterproductive work behaviors a0034822
related to envy? Journal of Applied Psychology, 92, 666 – 680. http://dx Judge, T. A., Locke, E. A., & Durham, C. C. (1997). The dispositional
.doi.org/10.1037/0021-9010.92.3.666 causes of job satisfaction: A core evaluations approach. Research in
Darnon, C., Dompnier, B., Delmas, F., Pulfrey, C., & Butera, F. (2009). Organizational Behavior, 19, 151–188.
Achievement goal promotion at university: Social desirability and social Kehoe, R. R., Lepak, D. P., & Bentley, F. S. (2018). Let’s call a star a star:
utility of mastery and performance goals. Journal of Personality and Task performance, external status, and exceptional contributors in orga-
nizations. Journal of Management, 44, 1848 –1872. http://dx.doi.org/10
Social Psychology, 96, 119 –134. http://dx.doi.org/10.1037/a0012824
.1177/0149206316628644
Dietz, B., van Knippenberg, D., Hirst, G., & Restubog, S. L. D. (2015).
Kehoe, R. R., & Tzabbar, D. (2015). Lighting the way or stealing the
Outperforming whom? A multilevel study of performance-prove goal
shine? An examination of the duality in star scientists’ effects on firm
orientation, performance, and the moderating role of shared team iden-
innovative performance. Strategic Management Journal, 36, 709 –727.
tification. Journal of Applied Psychology, 100, 1811–1824. http://dx.doi
http://dx.doi.org/10.1002/smj.2240
.org/10.1037/a0038888
Kim, E., & Glomb, T. M. (2014). Victimization of high performers: The
Dineen, B. R., Duffy, M., Henle, C., & Lee, K. (2017). Green by compar-
roles of envy and work group identification. Journal of Applied Psy-
ison: Deviant and normative transmutations of job search envy in a chology, 99, 619 – 634. http://dx.doi.org/10.1037/a0035789
temporal context. Academy of Management Journal, 60, 295–320. http:// Kluger, A. N., & DeNisi, A. (1996). The effects of feedback interventions
dx.doi.org/10.5465/amj.2014.0767 on performance: A historical review, a meta-analysis, and a preliminary
Dragoni, L., Tesluk, P. E., Russell, J. E., & Oh, I. S. (2009). Understanding feedback intervention theory. Psychological Bulletin, 119, 254 –284.
managerial development: Integrating developmental assignments, learn- http://dx.doi.org/10.1037/0033-2909.119.2.254
ing orientation, and access to developmental opportunities in predicting Kluger, A. N., & DeNisi, A. (1998). Feedback interventions: Toward the
managerial competencies. Academy of Management Journal, 52, 731– understanding of a double-edged sword. Current Directions in Psycho-
743. http://dx.doi.org/10.5465/amj.2009.43669936 logical Science, 7, 67–72. http://dx.doi.org/10.1111/1467-8721
Duffy, M. K., Scott, K. L., Shaw, J. D., Tepper, B. J., & Aquino, K. (2012). .ep10772989
A social context model of envy and social undermining. Academy of Kulik, C. T., & Ambrose, M. L. (1992). Personal and situational determi-
Management Journal, 55, 643– 666. http://dx.doi.org/10.5465/amj.2009 nants of referent choice. The Academy of Management Review, 17,
.0804 212–237. http://dx.doi.org/10.5465/amr.1992.4279534
Lam, C. K., Van der Vegt, G. S., Walter, F., & Huang, X. (2011). Harming performance goals. Current Directions in Psychological Science, 19,
high performers: A social comparison perspective on interpersonal 324 –328. http://dx.doi.org/10.1177/0963721410383246
harming in work teams. Journal of Applied Psychology, 96, 588 – 601. Quiñones, M. A., Ford, J. K., & Teachout, M. S. (1995). The relationship
http://dx.doi.org/10.1037/a0021882 between work experience and job performance: A conceptual and meta-
Lee, F. K., Sheldon, K. M., & Turban, D. B. (2003). Personality and the analytic review. Personnel Psychology, 48, 887–910. http://dx.doi.org/
goal-striving process: The influence of achievement goal patterns, goal 10.1111/j.1744-6570.1995.tb01785.x
level, and mental focus on performance and enjoyment. Journal of Régner, I., Escribe, C., & Dupeyrat, C. (2007). Evidence of social com-
Applied Psychology, 88, 256 –265. http://dx.doi.org/10.1037/0021-9010 parison in mastery goals in natural academic settings. Journal of Edu-
.88.2.256 cational Psychology, 99, 575–583. http://dx.doi.org/10.1037/0022-0663
Lee, K., & Duffy, M. K. (2019). A functional model of workplace envy and .99.3.575
job performance: When do employees capitalize on envy by learning Shah, P. P. (1998). Who are employees’ social referents? Using a network
from envied targets? Academy of Management Journal, 62, 1085–1110. perspective to determine referent others. Academy of Management Jour-
http://dx.doi.org/10.5465/amj.2016.1202 nal, 41, 249 –268.
Levy, I., Kaplan, A., & Patrick, H. (2004). Early adolescents’ achievement Sitzmann, T., & Yeo, G. (2013). A meta-analytic investigation of the
goals, social status, and attitudes towards cooperation with peers. Social within-person self-efficacy domain: Is self-efficacy a product of past
Psychology of Education, 7, 127–159. http://dx.doi.org/10.1023/B: performance or a driver of future performance? Personnel Psychology,
SPOE.0000018547.08294.b6 66, 531–568. http://dx.doi.org/10.1111/peps.12035
Locke, E. A., & Latham, G. P. (2002). Building a practically useful theory Spreitzer, G. M. (1996). Social structural characteristics of psychological
of goal setting and task motivation. A 35-year odyssey. American empowerment. Academy of Management Journal, 39, 483–504.
Psychologist, 57, 705–717. http://dx.doi.org/10.1037/0003-066X.57.9 Stajkovic, A. D., & Luthans, F. (1998). Self-efficacy and work-related
.705 performance: A meta-analysis. Psychological Bulletin, 124, 240 –261.
Lockwood, P., Jordan, C. H., & Kunda, Z. (2002). Motivation by positive http://dx.doi.org/10.1037/0033-2909.124.2.240
or negative role models: Regulatory focus determines who will best Suls, J., Martin, R., & Wheeler, L. (2002). Social comparison: Why, with
inspire us. Journal of Personality and Social Psychology, 83, 854 – 864. whom, and with what effect? Current Directions in Psychological Sci-
http://dx.doi.org/10.1037/0022-3514.83.4.854 ence, 11, 159 –163. http://dx.doi.org/10.1111/1467-8721.00191
Tepper, B. J., Henle, C. A., Lambert, L. S., Giacalone, R. A., & Duffy,
Lockwood, P., & Kunda, Z. (1997). Superstars and me: Predicting the
M. K. (2008). Abusive supervision and subordinates’ organization de-
impact of role models on the self. Journal of Personality and Social
viance. Journal of Applied Psychology, 93, 721–732. http://dx.doi.org/
Psychology, 73, 91–103. http://dx.doi.org/10.1037/0022-3514.73.1.91
10.1037/0021-9010.93.4.721
Meade, A. W., & Craig, S. B. (2012). Identifying careless responses in
Tesser, A. (1988). Toward a self-evaluation maintenance model of social
survey data. Psychological Methods, 17, 437– 455. http://dx.doi.org/10
behavior. Advances in Experimental Social Psychology, 21, 181–227.
.1037/a0028085
http://dx.doi.org/10.1016/S0065-2601(08)60227-0
Mischel, W. (1973). Toward a cognitive social learning reconceptualiza-
Tierney, P., & Farmer, S. M. (2002). Creative self-efficacy: Its potential
tion of personality. Psychological Review, 80, 252–283. http://dx.doi
antecedents and relationship to creative performance. Academy of Man-
.org/10.1037/h0035002
agement Journal, 45, 1137–1148.
Mischel, W., & Shoda, Y. (1995). A cognitive-affective system theory of
Vancouver, J. B., More, K. M., & Yoder, R. J. (2008). Self-efficacy and
personality: Reconceptualizing situations, dispositions, dynamics, and
resource allocation: Support for a nonmonotonic, discontinuous model.
invariance in personality structure. Psychological Review, 102, 246 –
Journal of Applied Psychology, 93, 35– 47. http://dx.doi.org/10.1037/
268. http://dx.doi.org/10.1037/0033-295X.102.2.246
0021-9010.93.1.35
Mitchell, T. R., & Nebeker, D. M. (1973). Expectancy theory predictions Vancouver, J. B., & Purl, J. D. (2017). A computational model of self-
of academic effort and performance. Journal of Applied Psychology, 57, efficacy’s various effects on performance: Moving the debate forward.
61– 67. http://dx.doi.org/10.1037/h0034181 Journal of Applied Psychology, 102, 599 – 616. http://dx.doi.org/10
Myers, C. G. (2018). Coactive vicarious learning: Toward a relational .1037/apl0000177
theory of vicarious learning in organizations. The Academy of Man- Vandewalle, D. (1997). Development and validation of a work domain goal
agement Review, 43, 610 – 634. http://dx.doi.org/10.5465/amr.2016 orientation instrument. Educational and Psychological Measurement,
.0202 57, 995–1015. http://dx.doi.org/10.1177/0013164497057006009
Northcraft, G. B., Schmidt, A. M., & Ashford, S. J. (2011). Feedback and VandeWalle, D., Cron, W. L., & Slocum, J. W., Jr. (2001). The role of
the rationing of time and effort among competing tasks. Journal of goal orientation following performance feedback. Journal of Applied
Applied Psychology, 96, 1076 –1086. http://dx.doi.org/10.1037/ Psychology, 86, 629 – 640. http://dx.doi.org/10.1037/0021-9010.86.4
a0023221 .629
Orne, M. T. (1962). On the social psychology of the psychological exper- Weiss, H., & Sherman, J. (1973). Internal-external control as a predictor
iment: With particular reference to demand characteristics and their of task effort and satisfaction subsequent to failure. Journal of
implications. American Psychologist, 17, 776 –783. http://dx.doi.org/10 Applied Psychology, 57, 132–136. http://dx.doi.org/10.1037/
.1037/h0043424 h0037054
Payne, S. C., Youngcourt, S. S., & Beaubien, J. M. (2007). A meta-analytic Wilk, S. L., & Moynihan, L. M. (2005). Display rule “regulators”: The
examination of the goal orientation nomological net. Journal of Applied relationship between supervisors and worker emotional exhaustion.
Psychology, 92, 128 –150. http://dx.doi.org/10.1037/0021-9010.92.1 Journal of Applied Psychology, 90, 917–927. http://dx.doi.org/10.1037/
.128 0021-9010.90.5.917
Phillips, J. M., & Gully, S. M. (1997). Role of goal orientation, ability, Williams, L. J., & Anderson, S. E. (1991). Job satisfaction and organiza-
need for achievement, and locus of control in the self-efficacy and tional commitment as predictors of organizational citizenship and in-role
goal—Setting process. Journal of Applied Psychology, 82, 792– 802. behaviors. Journal of Management, 17, 601– 617. http://dx.doi.org/10
http://dx.doi.org/10.1037/0021-9010.82.5.792 .1177/014920639101700305
Poortvliet, P. M., & Darnon, C. (2010). Toward a more social understand- Wofford, J. C., Goodwin, V. L., & Premack, S. (1992). Meta-analysis of
ing of achievement goals: The interpersonal effects of mastery and the antecedents of personal goal level and of the antecedents and
consequences of goal commitment. Journal of Management, 18, 595– function of task demands. Journal of Applied Psychology, 94, 710 –726.
615. http://dx.doi.org/10.1177/014920639201800309 http://dx.doi.org/10.1037/a0015044
Wood, J. V. (1989). Theory and research concerning social comparisons of
personal attributes. Psychological Bulletin, 106, 231–248. http://dx.doi
.org/10.1037/0033-2909.106.2.231 Received May 17, 2017
Yeo, G., Loft, S., Xiao, T., & Kiewitz, C. (2009). Goal orientations and Revision received February 11, 2020
performance: Differential relationships across levels of analysis and as a Accepted April 8, 2020 䡲
Members of Underrepresented Groups:

Reviewers for Journal Manuscripts Wanted
If you are interested in reviewing manuscripts for APA journals, the APA Publications and
Communications Board would like to invite your participation. Manuscript reviewers are vital to the
publications process. As a reviewer, you would gain valuable experience in publishing. The P&C
Board is particularly interested in encouraging members of underrepresented groups to participate
more in this process.
If you are interested in reviewing manuscripts, please write APA Journals at [email protected].
Please note the following important points:
To be selected as a reviewer, you must have published articles in peer-reviewed journals. The
experience of publishing provides a reviewer with the basis for preparing a thorough, objective
review.
To be selected, it is critical to be a regular reader of the five to six empirical journals that are most
central to the area or journal for which you would like to review. Current knowledge of recently
published research provides a reviewer with the knowledge base to evaluate a new submission
within the context of existing research.
To select the appropriate reviewers for each manuscript, the editor needs detailed information.
Please include with your letter your vita. In the letter, please identify which APA journal(s) you
are interested in, and describe your area of expertise. Be as specific as possible. For example,
“social psychology” is not sufficient—you would need to specify “social cognition” or “attitude
change” as well.
Reviewing a manuscript takes time (1– 4 hours per manuscript reviewed). If you are selected to
review a manuscript, be prepared to invest the necessary time to evaluate the manuscript
thoroughly.
APA now has an online video course that provides guidance in reviewing manuscripts. To learn
more about the course and to access the video, visit http://www.apa.org/pubs/journals/resources/
review-manuscript-ce-video.aspx.

Social Comparison Perfomance

Uploaded by

Copyright:

Available Formats

Social Comparison Perfomance

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Social Comparison Perfomance

Uploaded by

Copyright:

Available Formats

Journal of Applied Psychology

Referents or Role Models? The Self-Efficacy and Job Performance Effects

Patrick E. Downes Eean R. Crawford

Scott E. Seibert Adam C. Stoverink

Supplemental materials: http://dx.doi.org/10.1037/apl0000519.supp

cognitive theories in this way raises an important theoretical and

practical question: When do higher performers function as refer-

Theoretical Background Social Cognitive Theory

as the outcomes of that instructive function. Social comparison

theory (Festinger, 1954) recognizes similar cognitive processes

Unless otherwise indicated, survey items were based on a 5-point 3.29.

We first examined the proposed measurement model using

confirmatory factor analysis, entering the two psychometric vari-

1. Prior performance 3.38 0.52

Low Number of Higher High Number of Higher

Figure 2. The interactive effect of the number of high performers and

efficacy (Study 1). High and low levels of PPGO operationalized at ⫾ 1

SD. Low and high number of higher performers are approximately ⫾ 1 SD

individuals is negative and significant (b ⫽ ⫺.04, p ⬍ .05); the simple

slope for low-PPGO individuals is positive and significant (b ⫽ .06, p ⬍

performers and job-based self-efficacy was positive for low-PPGO

individuals (b ⫽ .07, p ⬍ .05; 95% CI [.01, .10]) and negative

for high-PPGO individuals (b ⫽ ⫺.04, p ⬍ .01; 95% CI

[⫺.06, ⫺.01]). This difference was statistically significant (b ⫽

.11, p ⬍ .05; 95% CI [.03, .14]). The first-stage findings carried

through to impact performance such that the indirect effect on job

p ⬍ .01; 95% CI [⫺.06, ⫺.01]), but positive for low-PPGO

was statistically significant (b ⫽ .10, p ⬍ .01; 95% CI [.04, .16]).

These results offered support for Hypothesis 3.

Supplementary Analysis and Robustness Checks

performance prove goal orientation; Perfs. ⫽ performers. N ⫽ 119.

Research has pointed out the importance of ego-referent simi-

1988). We argued above that evaluating similarity across multiple

simultaneous comparisons is cognitively complex. Although ex-

isting research emphasizes ego-referent similarity as a key mod-

erator of social comparison effects, our intent was instead to focus

on understanding PPGO as an individual difference in information

processing across perceptions of multiple higher performers. Con-

struct, but did account for similarity in several ways.

Most importantly, we collected data that allowed respondents to

self-report their salient others at work. Given that similarity plays

already been deemed similar (or at least relevant) by the focal

individual. That is, the referents selected have already overcome

cognitively salient to respondents. However, we acknowledge that

or department) might be more meaningful (and thus have a stron-

Path Low PPGO High PPGO Difference

First stage effect on self-efficacy

1. Time 1 task performance 3.07 0.62

DV: Task-based self-efficacy DV: T2 task performance

Path Low PPGO High PPGO Difference

First stage effect on self-efficacy

Time Spent on Peer Strategies

Low Number of Higher High Number of Higher 20

Performers Performers No High Performer High Performer Present

Festinger, L. (1954). A theory of social comparison processes. Human

Relations, 7, 117–140. http://dx.doi.org/10.1177/001872675400700202

Members of Underrepresented Groups:

You might also like