Group identity and positive deviance in work groups

Abstract

This study examines why and how identity cognitions, including group identification and individual differentiation, influence the positive deviance of employees. We identify the risk-taking intention of employees as a critical psychological mechanism to overcome stigma-induced identity threat of positive deviance. The analysis of data collected from 293 members comprising 66 work teams reveals that the relationship between individual differentiation and positive deviance is partially mediated by risk-taking intention. The indirect effect of group identification on positive deviance through risk-taking intention is also significant and positive in groups with low conformity pressure, whereas the same indirect effect is neutralized in groups with high conformity pressure. The current analysis offers new insights into the way the group context and the identity cognition of members explain the development of positive deviance and workplace creativity.

Full text

Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=vsoc20
The Journal of Social Psychology
ISSN: 0022-4545 (Print) 1940-1183 (Online) Journal homepage: http://www.tandfonline.com/loi/vsoc20
Group identity and positive deviance in work
groups
Moon Joung Kim & Jin Nam Choi
To cite this article: Moon Joung Kim & Jin Nam Choi (2018) Group identity and positive
deviance in work groups, The Journal of Social Psychology, 158:6, 730-743, DOI:
10.1080/00224545.2017.1412931
To link to this article: https://doi.org/10.1080/00224545.2017.1412931
Accepted author version posted online: 05
Dec 2017.
Published online: 26 Dec 2017.
Submit your article to this journal
Article views: 212
View Crossmark data

Group identity and positive deviance in work groups
Moon Joung Kim
a
and Jin Nam Choi
b
a
Department of Business Administration, Cheongju University, Cheongju si, South Korea;
b
Graduate School of
Business, Seoul National University, Seoul, South Korea
ABSTRACT
This study examines why and how identity cognitions, including group
identification and individual differentiation, influence the positive deviance
of employees. We identify the risk-taking intention of employees as a critical
psychological mechanism to overcome stigma-induced identity threat of
positive deviance. The analysis of data collected from 293 members com-
prising 66 work teams reveals that the relationship between individual
differentiation and positive deviance is partially mediated by risk-taking
intention. The indirect effect of group identification on positive deviance
through risk-taking intention is also significant and positive in groups with
low conformity pressure, whereas the same indirect effect is neutralized in
groups with high conformity pressure. The current analysis offers new
insights into the way the group context and the identity cognition of
members explain the development of positive deviance and workplace
creativity.
ARTICLE HISTORY
Received 21 November 2016
Revised 7 November 2017
Accepted 18 November 2017
KEYWORDS
Group conformity pressure;
group identification;
individual differentiation;
risk-taking intention;
positive deviance
As creativity and innovation are critical to a successful performance in knowledge-based society,
recent studies underscored the role of deviant behavior in organizations (Krau, 2008). Deviance is
viewed as a propitious basis of creativity and innovation rather than merely a harmful behavior,
because creative and innovative processes often require individuals to deviate from existing norms
and status quo (Appelbaum, Iaconi, & Matousek, 2007). By focusing on the positive side of deviant
behavior, scholars have labeled the act of disregarding significant norms of the referent group to
achieve socially desirable ends such as positive deviance (Spreitzer & Sonenshein, 2004).
Despite the increasing importance of positive deviance, research on the psychological and group
processes underlying this phenomenon remains lacking. By drawing on social identity theory, we
examine the positive deviance in relation to the identity cognition of individual members, including
group identification and individual differentiation (Van Knippenberg, 2000). Group members may
act either in terms of their group membership or as independent individuals; these distinct
identification processes of individual members may significantly affect the likelihood for them to
engage in positive deviance (Janssen & Huang, 2008). Considering that deviation from norms may
generate stigma or negative sanctions in work groups, it is important to elaborate on the necessary
cognitive beliefs or psychological states to overcome the stigma-induced identity threat from groups
and to successfully initiate constructive deviant behaviors (Major & O’Brien, 2005; Spreitzer &
Sonenshein, 2004).
Rather than assuming the direct, main effects of identity cognition on positive deviance, we
explore a potential intermediate psychological state and group context to further clarify the relation-
ship between identity cognition and positive deviance. Specifically, we identify risk-taking intention
as a mediator of the relationship, because the most evident barrier to positive deviance in a group
setting involves social risks associated with violating group norms (Mertens, Recker, Kohlborn, &
CONTACT Jin Nam Choi jnchoi@snu.kr Graduate School of Business, Seoul National University, 1 Gwanak-ro, Gwanak-gu,
Seoul 08826, South Korea.
© 2018 Taylor & Francis
THE JOURNAL OF SOCIAL PSYCHOLOGY
2018, VOL. 158, NO. 6, 730–743
https://doi.org/10.1080/00224545.2017.1412931

Kummer, 2016; Warren, 2003). In addition, we isolate group conformity pressure as a group-level
contingency that attenuates the effect of risk-taking intention in mediating the relationship between
identity cognitions and positive deviance. Individuals will unlikely take risks in a group, which
pressures its members to conform to other members (Stangor, 2004).
The current study makes the following contributions. First, we theoretically identify and empiri-
cally examine the psychological mechanism (i.e., risk-taking intention) through which identity
cognitions increase positive deviance. Second, we further propose a boundary condition of such
mediated relationship by proposing that the mediating role of risk-taking intention may vary across
teams with different levels of conformity pressure on members. The investigation of this contextual
contingency may elaborate the understanding on when identity cognition facilitates or hampers
positive deviance. Finally, the current effort to include group conformity pressure extends the
positive deviance literature by employing a multilevel approach (Morrison, Wheeler-Smith, &
Kamdar, 2011; Vadera, Pratt, & Mishra, 2013). In the present study, we broaden the focus of
previous research, which is mostly confined on the individual-level relationship involving positive
deviance, by examining the role of group-level conformity pressure in shaping the individual-level
relationship between identity cognition and risk-taking intention and subsequent positive deviance.
Theoretical background and hypotheses
Effects of identity cognition on positive deviance
Positive deviance is an “umbrella term”(Warren, 2003) encompassing a diverse group of related
behaviors, such as principled organizational dissent (Graham, 1986), whistle-blowing (Near &
Miceli, 1985), exercising voice (Morrison et al., 2011), prosocial behaviors (O’Reilly & Chatman,
1986), and functional or creative disobedience (Brief, Buttram, & Dukerich, 2001). Although positive
deviance is broadly defined to include various behaviors with constructive intention to improve
performance, its key elements are rule-breaking, norm-violation, and breaking the “red tape,”which
put them in the same category (Vadera et al., 2013).
Given that deviation from norms may generate stigma or negative sanctions in work groups,
positive deviance may be critically affected by the identity cognition of members. Thus, the present
study examines positive deviance in relation to identity cognition by drawing on social identity
theory. We focus on the two complementary types of identity cognition, namely, group identification
and individual differentiation. The former refers to the self-perceptions of individuals that they are
part of a group (Ashforth & Mael, 1989), whereas the latter is the extent to which individuals
distinguish themselves from other group members in terms of their thoughts, feelings, and behaviors
(Janssen & Huang, 2008). The empirical results of the effects of individual differentiation on positive
deviance consistently show that individual differentiation is positively associated with positive
deviance (Janssen & Huang, 2008; Tse & Chiu, 2014). However, the empirical results of group
identification in relation to constructive deviance are quite mixed with positive (Morrison et al.,
2011; Venkataramani & Tangirala, 2010), insignificant (Janssen & Huang, 2008; Tse & Chiu, 2014),
and negative relationships (Umphress, Bingham, & Mitchell, 2010).
Based on their divergence and idiosyncratic perspectives, people with a strong sense of individual
differentiation are unlikely to be constrained by established paradigms and norms, and thus, they
can question the status quo of their organization and spontaneously consider alternatives (Janssen &
Huang, 2008). Individual differentiation may drive group members to engage in change-oriented,
risk-taking behaviors. Members with strong individual differentiation can challenge the assumptions
behind established frameworks of group thoughts and routines rather than take them for granted,
and they feel free to take risks associated with novel solutions (Tse & Chiu, 2014). Thus, individual
differentiation is expected to facilitate positive deviance. Our first hypothesis is as follows:
Hypothesis 1. Individual differentiation is positively related with positive deviance.
THE JOURNAL OF SOCIAL PSYCHOLOGY 731

Individuals with strong group identification cognitively categorize themselves as group members
(Ashmore, Deaux, & Mclaughlin-Volpe, 2004) and perceive themselves in terms of the values, goals,
attitudes, and behaviors they share with other members (Swaab, Postmes, & Spears, 2004). The
controversy in this identification rests on whether identified members are motivated toward either
norm-following or -breaking. When group identification strongly influences the social motives of
members, they tend to seek cooperative and pro-social alternatives, thereby adhering to group norms
and status quo (De Cremer, 2001; Desivilya & Eizen, 2005). Meanwhile, highly identifying members
may be motivated to “bend the rules”(Vadera et al., 2013, p. 1264) or deviate from norms as far as
the group benefits by changing a troubling situation (Venkataramani & Tangirala, 2010). By
considering these ambivalent implications of group identification toward positive deviance, we do
not expect any direct main effect but propose that the relationship is moderated by the nature of a
given group context.
Mediating role of risk-taking intention
The inconsistent findings regarding the relationship between identity cognition and positive
deviance highlight the importance of investigating mediating mechanisms. By focusing on the
psychological state that develops positively deviant behaviors, we examine the role of risk-taking
intention of group members as a plausible mediating mechanism. The willingness to confront social
risk reflects a psychological readiness to initiate positive deviance (Spreitzer & Sonenshein, 2004).
For example, Madjar, Greenberg, and Chen (2011) demonstrated a positive association between
willingness to take risks and radical creativity. Morrison (2006) also determined a positive relation-
ship between risk-taking propensity and likelihood to engage in prosocial rule-breaking. Positive
deviance often involves significant risks because it may not be necessarily rewarded but often
punished for going against established rules or practices (Jones, 1998). Individuals must move
outside their comfort zone beyond the boundaries of their psychological safety net when engaging
in positive deviance (Morrison, 2006). Hence, risk-taking intention may be a key driver of positive
deviance.
We further propose that risk-taking intention mediates the relationship between identity cogni-
tion and positive deviance. People generally prefer to maintain congruent identity and behavior.
Thus, individuals may approach risk in ways fitting their self-identity. For example, the group
identification level of a member, which indicates the psychological prominence of group member-
ship, may determine whether a person takes risks and violates current norms (Tyler & Blader, 2003).
However, the aforementioned debate exists on whether members acting in terms of membership are
motivated toward either norm-following or -breaking to benefit their group (Vadera et al., 2013).
Thus, we expect an overall neutral effect of group identification on risk-taking without specific
contingencies.
By contrast, members with strong individual differentiation may approach risk differently to
achieve their goals or cultivate their own judgment regardless of compatibility with group values or
norms (Damaraju, Barney, & Dess, 2010). Risk-taking involves deviation from status quo and group
expectations or norms, thereby, engendering the stigma of failure (Fu, Tsui, & Dess, 2006). Members
with strong individual differentiation are less sensitive and tend to accept such stigma from a group
compared with members with strong group identification because they place higher importance on
their own judgment and beliefs. Therefore, members with individual differentiation may develop
strong risk-taking intention and undertake proactive and risk-taking behaviors rather than resorting
to defensive and risk-averse strategies. Thus, we propose the following mediation hypothesis:
Hypothesis 2. Risk-taking intention mediates the positive relationship between individual differen-
tiation and positive deviance.
732 M. J. KIM AND J. N. CHOI

Moderating role of group conformity pressure
The relationship between identity cognition and risk-taking intention is likely affected by group
conformity pressure. Group conformity pressure is the extent of social influence on members,
which involves a forced change in their belief or behavior to fit in with a group (Janis, 1972).
Groups enforce conformity norms by rewarding conforming members with acceptance and
approval and by threatening non-conformers with rejection and sanctions. However, the efficacy
and potency of group conformity pressure differ among members with various identity
cognitions.
Highly identified members tend to be more susceptible to group contexts and care more about
avoiding negative social stigma during in-group interactions than those with strong individual differ-
entiation (Mandel, 2003). The former tend to cherish group membership and prioritize group con-
formity pressure, and thus, they will likely avoid social risks associated with contradicting group norms
(Brewer & Silver, 2000). Thus, members with strong group identification may comply with rules under
high group conformity pressure to serve a group. By contrast, members of groups with low conformity
pressure who have strong group identification become willing to take risks and bend the rules to benefit
their group. Thus, we identify group conformity pressure as a critical boundary condition that affects the
willingness of group-identified members to take risks and their resulting choices between rule-complying
and -bending behaviors. However, those with strong individual differentiation have a sense of self-
determination and stand up for what they believe is right regardless of the group norms and conformity
pressure. Thus, we propose that group conformity pressure moderates the effect of group identification
but not that of individual differentiation:
Hypothesis 3. Group conformity pressure negatively moderates the individual-level relationship
between group identification and risk-taking intention such that group identification
is more negatively related to risk-taking intention when group conformity pressure is
high than when low.
Mediation moderated by group conformity pressure
We assume an overall neutral main effect of groupidentification on positive deviance without
specific contingencies. By extending Hypothesis 3, which isolates group conformity pressure as a
contingency that changes the relationship between group identification and risk-taking intention,
we further propose a moderated mediation hypothesis. As group conformity pressure accentuates
a potentially negative association between group identification and risk-taking intention
(Hypothesis 3), this pressure may negatively moderate the indirect effect of group identification
on positive deviance through risk-taking intention.
When group conformity pressure is low, members with strong group identification tend to be less
restrained by negative social stigma and are willing to take risks and bend rules as they believe that
such deviant behaviors generate positive consequences for a group (Vadera et al., 2013). Members
with strong identification will likely make alternative assessments “in terms of the consequences for
the specified group”(Venkataramani & Tangirala, 2010, p. 585), thereby actively changing a
troubling situation beyond the status quo to facilitate the achievement of a group goal, particularly
without conformity pressure. Conversely, with high group conformity pressure, members with
strong group identification become concerned about the negative social stigma and act in line
with their group norms. We proposed our fourth hypothesis:
Hypothesis 4. Group conformity pressure negatively moderates the indirect effect of group identi-
fication on positive deviance through risk-taking intention, such that the indirect
effect is more negative when group conformity pressure is high than when low.
THE JOURNAL OF SOCIAL PSYCHOLOGY 733

Methods
Research setting, participants, and procedures
To test the current propositions, we collected data from four companies (semiconductor equipment
manufacturing, flat panel display equipment manufacturing, vacuum technology, and marine and
fire insurance companies) in Korea. The current data collection was supported by the top manage-
ment of these companies with between 100 and 300 employees. The human resource (HR) managers
of these companies identified participating work teams and encouraged their participation. All
companies employed a team-based structure and team-level, performance-based incentives.
Members of the same team were physically collocated and interacted daily. The participants
performed various functions, including sales, HR, finance, research and development, production,
and quality control. We visited these companies and asked participants to fill out the questionnaires
during their regular work hours and collected the completed survey in person.
Of the 68 supervisor surveys and 459 employee surveys distributed, 66 supervisor (97% response
rate) and 359 employee questionnaires (78% response rate) were returned. After excluding ques-
tionnaires with missing information and those failing to match supervisor ratings, the final analysis
sample was comprised of 293 employees from 66 work teams (64% response rate). The final sample
included 186 employees from 41 groups in Company 1, 30 employees from 6 groups in Company 2,
62 employees from 14 groups in Company 3, and 19 employees from 5 groups in Company 4. The
average team size is 4.4, and team size ranges from 2 to 10.
For the employee sample, the education levels of participants were high school (10.9%), 2-year
college (41%), bachelor’s degree (41.3%), and graduate degree (5.1%). Their job positions were staff
(21.8%), senior staff (21.2%), assistant manager (28.3%), department manager (24.6%), and deputy
general manager or higher (4.1%). The average organizational tenure of the participating employees
was 4.63 years (SD = 3.70) with an average age of 33.03 years (SD = 5.13), and 12.6% were women.
For the supervisor sample, the average organizational tenure was 10.22 years (SD = 5.84), and the
average age was 41.19 years (SD = 4.18). Supervisor education levels were high school (13.4%), two-
year college (28.4%), bachelor’s degree (44.8%), and graduate degree (11.9%). Their job positions
were department manager (26.2%) and deputy general manager or higher (72.3%), and 3% of the
supervisors were women.
Measures
Data were collected from two sources to reduce concerns related to similar source bias (Podsakoff,
MacKenzie, Lee, & Podsakoff, 2003). Employees completed a questionnaire, which included group
identification, individual differentiation, risk-taking intention, and group conformity pressure mea-
sures, as well as demographic information. Supervisors completed a separate survey wherein they
rated the positive deviance of their subordinates. Study variables were assessed using multi-item
scales with acceptable reliability. All items were measured on a Likert-type scale ranging from one
(strongly disagree) to six (strongly agree).
Group identification. We assessed the group identification of team members using a three-item
measure (α= .91) developed by Sheldon and Bettencourt (2002). The scale included the following
items: “I feel included in this group,”“I feel well-integrated into this group,”and “I feel a sense of
belongingness within the group.”
Individual differentiation. To assess individual differentiation, we adopted the three-item measure
(α=.88) developed by Sheldon and Bettencourt (2002). The items were “I feel like I stand out within this
group,”“I feel unique as I participate in this group,”and “I feel distinct and separate within this group.”
Risk-taking intention. By adopting items from previous studies (Andrews & Smith, 1996;Weber,
Blais, & Betz, 2002), we used a three-item scale (α= .81) to measure the risk-taking intention of
734 M. J. KIM AND J. N. CHOI

members. The items included “In my work, I prefer to play safe when I develop ideas (reverse),”“I
am a risk-taker when it comes to proposing ideas,”and “I prefer to think conservatively when I
develop ideas (reverse).”
Group conformity pressure. The groupthink measure by Henningsen, Henningsen, Eden, and Cruz
(2006)includes two subscales, namely, pressure to conform and self-censorship, and each subscale
comprised four items. We adopted the first subscale to assess group conformity pressure. The four
items (α= .84) are “In my group, (a) members feel the pressure to go along with other group
members, (b) members feel that they need to conform to the other members, (c) each member is
pressured to make the decision everyone else favors, and (d) members seem pressured by other
members to make a specific decision.”
Positive deviance. To assess positive deviance, we used four items (α=.85) developed by
Galperin (2012), which include “This employee did not follow my orders to improve work
procedures,”“This employee disagreed with others in the work group to improve the current
work procedures,”“This employee disobeyed my instructions to perform efficiently,”and “This
employee reported a wrong-doing of another person in the company to bring about positive
organizational change.”
Control variables. In addition to the study variables described above, we included several control
variables in the statistical analyses. Specifically, we controlled for company, gender, age, and tenure
at the individual level, and female proportion and group size at the group level of analysis because
these variables may significantly influence the interpersonal and task-related processes of individuals
in a group setting (Amabile, 1996; Becker, 2005).
Data aggregation
We investigated multilevel dynamics that involve a group-level variable as a cross-level moderator. We
checked various aggregation statistics before aggregating the individual-level ratings of members on
group conformity pressure to the group level. Within-group agreement among members was evaluated
by computing r
wg
(James, Demaree, & Wolf, 1984). The mean r
wg
value for group conformity pressure
was .82, which indicated a strong within-group agreement (LeBreton & Senter, 2008). We then estimated
the intra-class correlation (ICC) statistics. The ICC1 and ICC2 statistics for group conformity pressure
were .23 and .51, respectively, with a significant group effect (F[65, 227] = 2.05, p<.001).Overall,these
group-level psychometric properties of the group conformity pressure scale provide empirical justifica-
tions for group-level aggregation. All other study variables were operationalized and analyzed at the
individual level.
Results
Confirmatory factor analysis
To examine the empirical distinctness of the study variables, a confirmatory factor analysis (CFA)
was conducted with maximum likelihood estimation. The results confirmed the five-factor structure
(χ
2
(df = 125) = 397.18, p< .001, CFI = .90, RMSEA = .08), which fits the data better than any
alternative factor models. For example, the CFA of a four-factor model in which risk-taking
intention and positive deviance load onto a single factor produced a significantly worse fit (χ
2
(df = 129) = 624.95, p< .001, CFI = .82, RMSEA = .12). Table 1 presents the descriptive statistics and
correlations among all study and control variables. The correlations among the three company
dummy variables are quite high; thus, the three company dummies are mean-centered before their
use in further statistical analyses (Tabachnick & Fidell, 2013).
THE JOURNAL OF SOCIAL PSYCHOLOGY 735

Comparison of the hypothesized and alternative models
The present theoretical propositions were tested by performing multilevel structural equation
modeling using Mplus 6.1 (Muthen & Muthen, 1998-2007), which allowed an omnibus test of
multi-step mediation effects in two-level data (Preacher, Zhang, & Zyphur, 2011). To assess the
constructs, we used the scale means of the corresponding items employed in this study. In testing the
hypothesized effect of individual differentiation on risk-taking intention and positive deviance, we
also included the path from group identification to control (and also contrast) its effect on the
mediator and the outcome variables. The hypothesized structural model exhibited a reasonable fit to
the data: χ
2
(df = 14) = 36.18, p< .001, CFI = .83, RMSEA = .07.
To verify the possibility that a theoretically plausible alternative model provides an improved
explanation of the present data, we compared two alternative models against the hypothesized
model, as reported in Table 2. The first alternative model examined the possibility that the mediating
role of risk-taking intention is only partial rather than full. To explore this possibility, we added two
paths representing the direct effects of group identification and individual differentiation on positive
deviance. As reported in Table 2, the partial mediation model with the two additional direct effect
paths improved the model fit significantly, exhibiting a better fit to the data (χ
2
(df = 12) = 18.36,
p< .001, CFI = .95, RMSEA = .04) with a significant chi-square change (Δχ
2
(df = 2) = 17.82,
p< .001). The second alternative model tested the possibility of the non-mediated direct effects
model, in which group identification, individual differentiation, and risk-taking intention directly
predict positive deviance without the hypothesized mediated relationship. Model 2 exhibits a poor fit
to the data (χ
2
(df = 18) = 49.77, p< .001, CFI = .76, RMSEA = .078). Thus, we adopted the partial
mediation model (Alternative Model 1) as it has a better fit and provided a theoretically plausible
explanation of the current data.
Tests of hypotheses
The results of the better-fitting, partial mediation model are presented in Figure 1. Among the
control variables that were included, gender indicated a highly significant relationship with risk-
taking intention (β= -.21, p< .001), suggesting that female employees reported lower risk-taking
Table 1. Means, standard deviations, and correlations: Individual level (N= 293).
Variables MSD1234567891011
1. Company 1 .63 .48 –
2. Company 2 .09 .29 −.41** –
3. Company 3 .21 .41 −.68** −.16** –
4. Gender .13 .33 −.01 −.08 .00 –
5. Age 33.03 5.13 −.22** .24** .07 −.28** –
6. Tenure 4.64 3.70 −.20** .48** −.21** −.05 .47** –
7. Group Identification 4.53 .97 −.07 −.00 .08 −.11 .04 −.02 –
8. Individual Differentiation 3.45 1.09 .05 −.14* .10 −.02 −.02 −.17** .19** –
9. Risk-taking Intention 3.52 .70 −.01 .04 .02 −.26** .16** −.03 .18** .30** –
10. Group Conformity
Pressure
3.07 .54 −.19** .26** −.02 16** .05 .14* −.14* −.05 −.05 –
11. Positive Deviance 2.77 .92 .02 −.11 .04 −.10 .26** .09 .15* .24** .23** .02 –
Note.*p< .05; ** p< .01.
Table 2. Comparison of alternative structural models.
Structural Models χ2df p CFI RMSEA AIC
Hypothesized structural model (Full mediation model) 36.183 14 .001 .830 .074 3012.270
Alternative model 1: Partial mediation model 18.366 12 .105 .951 .043 2998.200
Alternative model 2: Direct effect model
(Group Identification, Individual Differentiation, and Risk-taking Intention
predicting Positive Deviance)
49.772 18 .001 .757 .078 3011.909
736 M. J. KIM AND J. N. CHOI

tendency than their male counterparts. Age showed significant positive relationships with risk-taking
intention (β= .12, p< .05) and positive deviance (β= .25, p< .001). Company tenure also has a
significant positive relationship with positive deviance (β= .21, p< .05).
Hypothesis 1 suggests a direct, positive effect of individual differentiation on positive deviance. As
expected, the analysis showed that individual differentiation exerted a significant, positive effect on
positive deviance (β= .21, p< .001), confirming Hypothesis 1. Hypothesis 2 posits the mediating role
of risk-taking intention in the relationship between individual differentiation and positive deviance.
Risk-taking intention was positively related to positive deviance (β= .14, p< .05). Moreover, the
indirect effect of individual differentiation on positive deviance through risk-taking intention was
also significant (b = .039, p < .05). Thus, Hypothesis 2 was supported. Although we did not
hypothesize, group identification has a significant direct effect on positive deviance (β= .11,
p< .05). However, its indirect effect through risk-taking was insignificant.
Hypothesis 3 proposes that group-level conformity pressure negatively moderates the individual-level
relationship between group identification and risk-taking intention. We estimated a slope-as-outcome
model in HLM to test this cross-level moderation hypothesis. Model 3 in Table 3 reports that the cross-
level interaction between group identification and group conformity pressure was significant (γ=-.25,
p< .05), supporting Hypothesis 3. However, individual differentiation failed to show any significant
interaction with group conformity pressure (γ= .06, ns.). We further probed the significant cross-level
interaction by comparing the slopes associated with high and low group conformity pressure conditions
(Aiken & West, 1991). Consistent with our expectation, Figure 2 shows that group identification is
negatively related to risk-taking intention with high group conformity pressure (b = -.07, p<.05)but
positively related to risk-taking intention with low group conformity pressure (b = .23, p<.05).
Hypothesis 4 suggests distinct conditional indirect effects of group identification on positive
deviance through risk-taking intention at different levels of group conformity pressure. To test this
hypothesis, we used the PROCESS procedure based on a bootstrapping procedure (Hayes, 2013),
which provides a test for the entire moderated mediation model in an integrated analysis instead of
testing it in a piecemeal fashion. The current analytic procedure to test the conditional indirect
Figure 1. Results of structural equation modeling.
*p< .05; ** p< .01; *** p< .001.
THE JOURNAL OF SOCIAL PSYCHOLOGY 737

effects cannot accommodate the nested, multi-level data structure; thus, we performed the PROCESS
analysis using the individual scores of group conformity pressure as reported by members and found
a positive indirect effect of group identification on positive deviance with either low or medium
group conformity pressure, but not when it was high (see Table 4). To further validate the
moderated mediation pattern, we assigned the same group-level mean of the aggregated group
conformity pressure to members from the same group and performed the same analysis. Table 4
reports results consistent in that the indirect effect of group identification was positive with either
low or medium group conformity pressure, but not when it was high. These alternative moderated
mediation analyses of the current multi-level data supported Hypothesis 4.
Discussion
Positive deviance represents innovative behavior for change and improvement targeted to identify
and solve problems regarding existing work practices and processes (Pascale, Sternin, & Sternin,
Table 3. Hierarchical linear models predicting risk-taking intention.
Variables Model 1 Model 2 Model 3
Individual-level Process
Company1 .14 .15 .09
Company2 .26 .23 .17
Company3 .14 .15 .08
Age .02* .02* .03*
Gender −.59*** −.61*** −.43*
Tenure −.01 −.01 .01
Group Identification (GI) .07 .08
Individual Differentiation (ID) .24*** .22**
Group-level Process
Gender Proportion −.09
Group Size .04
Group Conformity Pressure (GCP) .01
Cross-level Moderation
GI * GCP .24*
ID * GCP .07
Pseudo R2.09 .10
Note.*p< .05; ** p< .01; *** p< .001.
2.5
3.0
Low High
Risk-Taking
Intention
Group Identification
High Conformity Pressure
Low Conformity Pressure
Figure 2. Cross-level moderation by group conformity pressure.
738 M. J. KIM AND J. N. CHOI

2010; Spreitzer & Sonenshein, 2004). Nonetheless, this pivotal process toward challenging status quo
and changing existing behaviors may be hindered by the possible stigma-induced identity threat that
group members often experience in taking social risks and initiating positive deviance in work teams.
By drawing on social identity theory, we isolate the identity cognition of members as a fundamental
determinant of positive deviance. We further elaborate this process by identifying risk-taking
intention as an intermediate psychological state and group-level conformity pressure as a cross-
level, contextual moderator. Our empirical analysis based on multi-source field data considerably
confirmed the current theoretical propositions. Below, we discuss the theoretical contributions of
this study and identify the limitations that can guide future investigations.
Theoretical contributions
This study contributes to the emerging literature on positive deviance by examining the relationship
between identity cognition and positive deviance. We specifically investigated group identification and
individual differentiation, which constitute the two complementary types of identity cognitions (Brewer,
1991). By extending prior studies that focused on the main effect of identity cognition on positive
deviance (Janssen & Huang, 2008;Vignoles,2011), we theorized and empirically demonstrated how
identity cognitions influenced positive deviance by examining risk-taking intention as a significant
psychological state required for positive deviance. Similarly, our analysis confirmed that the individual
differentiation of group members (but not group identification) exerted a significant positive effect on
positive deviance by increasing the members’risk-taking intention.
The current results confirmed that when individuals perceive themselves as different in a group, they
engage in positively deviant behaviors accompanied by their willingness to take social risks. Creativity
and innovation are viewed as positive deviances, especially when focusing on their social underpinnings
(Kusa, 2006). To perform creatively in work teams and organizations, individuals should tolerate the
sense of marginality or social isolation when breaking away positively from the established norms and
pursue unconventional solutions (Appelbaum et al., 2007). In this regard, appreciating individual
differences is a critical resource for creativity in a group; moreover, collective engagement in discovering
and appreciating positive deviants may be a critical first step toward identifying and adapting unique
solutions for desirable social and behavioral changes (Pascale et al., 2010, pp. 3–11). To transform social
inertia toward creativity and innovation, change agents should consider the emergence of positive
deviance from complementary types of identity cognition in social settings.
For group identification, unlike our expectation, the direct effect of group identification on
positive deviance was positive but not mediated by risk-taking intention. The current analysis also
demonstrated that the effect of group identification on positive deviance is contingent on the group
context. Individuals with strong identification with their group were susceptible to group conformity
pressure (Ellis & Fisher, 1994). Thus, they are willing to take risks and exhibit positive deviance with
low group conformity pressure. A low level of group conformity pressure operated as a situational
Table 4. Moderated indirect effects of group identification on positive deviance.
Independent
Variable Mediator
Dependent
Variable Moderator Level Effect
Boot
SE
95% bias-corrected
CI
Group Identification Risk- taking
Intention
Positive Deviance Group conformity
pressure
(individual perception)
Low .0447 .0213 (.0123 .0993)
Medium .0246 .0139 (.0037 .0603)
High .0045 .0150 (-.0234 .0384)
(aggregated)
Low .0447 .0214 (.0120 .0985)
Medium .0246 .0139 (.0042 .0603)
High .0045 .0151 (-.0236 .0373)
Note. Bootstrap sample = 10,000.
THE JOURNAL OF SOCIAL PSYCHOLOGY 739

contingency signaling that a group and other members tolerate deviant behaviors instead of forcing
them to conform to the decision of others or the group. By contrast, the effect of individual
differentiation on risk-taking intention was unaffected by group conformity pressure. The current
findings highlight the importance of investigating contextual contingencies and boundary conditions
that can liberate members from group constraints and normative pressures, thus increasing their
positive deviance. This approach also enriches the theoretical understanding of the emergence of
positive deviance and creativity by applying a multi-level framework (Mertens et al., 2016).
This study offers new theoretical insights by specifying “how”and “when”positive deviance can
be enhanced in work groups by members with different identity cognition. Positive deviance is the
behavior of individuals that departs from social and normative expectations (Pascale et al., 2010).
Hence, this construct cannot be fully understood without considering the multi-level dynamics
involving a person and his/her context. By identifying group conformity pressure as a group-level
contingency for individual-level relationships involving positive deviance, the present study responds
to the call for additional research attention to multilevel theorizing and investigation of positive
deviance (Morrison, 2011; Vadera et al., 2013). Future studies can further extend the current multi-
level approach.
One notable pattern related to the current set of control variables is the highly significant effect of
gender on risk-taking intention, which suggests that women reported considerably less willingness to
take risks and consequently lower deviant behavior in work groups. According to Brown, Olkhov, Bailey,
and Daniels (2015), gender differences incur unequal and unfair social punishments when norms for
emotional expression are violated. Specifically, compared with men, women are more harmed when they
violate emotional norms by displaying incongruent emotions; thereby, they pay greater social cost for
their affective deviance (Brown et al., 2015). Similarly, female employees can be unfairly treated and more
severely penalized when they exhibit deviant behaviors (even positive ones) than male employees; thus,
they are socialized to avoid taking social risks in the workplace. This potential gender difference involving
risk-taking and social deviance may negatively affect women and constrain their creativity. This spec-
ulation should be further investigated conceptually and empirically.
Practical implications
The current analysis also provides important practical implications for practicing managers aiming
to encourage positive deviance and creative endeavors among employees. First, managers may allow
a certain degree of individuation and psychological independence of employees to promote positive
deviance (Janssen & Huang, 2008). The present data showed that individual differentiation signifi-
cantly drives positive deviance by enabling employees to stand for their own ideas even when they
have to disagree with their coworkers, thereby accepting social risks. When group members lose
their individuality (de-individuation) and accept group mind,, to pursue normative behavior for
social approval, they will unlikely produce dissenting ideas, unique perspectives, and unconventional
solutions (Franzese, 2013).
Second, managers must identify and remove workplace barriers to positive deviance. In group
settings, particularly in cohesive groups, members are exposed to substantial psychological burden and
intimidation by negative group forces, which prevent them from realizing positively deviant behaviors
(cf., groupthink; Janis, 1972). One factor we identified is group conformity pressure, which operates
strongly among members who are loyal and identified with a group. Group conformity pressure may
intensify the fear of possible stigma-induced identity threat in groups and inhibit individual members
from taking risks, thereby diminishing constructive improvement and change efforts. By contrast, as
demonstrated by our moderated mediation analysis, positive deviance can be promoted among members
with strong group identification when group conformity pressure is maintained at a low level. Given that
positive deviance and constructive rule violation are foundations of organizational creativity and
innovation (Acharya & Taylor, 2012), the management should set systematic interventions to minimize
hindrances and build work contexts tolerant of and conducive to positive deviance.
740 M. J. KIM AND J. N. CHOI

Study limitations
The present findings should be interpreted with caution considering the following limitations of the
study. First, data were collected at a single point in time and the direction of causation remains
ambiguous. For example, increased interpersonal deviant behavior can reinforce individual differ-
entiation and further promote the willingness to take social risks. Further studies should ascertain
the causal direction or potential reciprocal causation of these constructs.
Second, owing to practical limitations in collecting field data, we could not use the full scale often
with multiple sub-dimensions for the current study variables, which compelled us to use brief measures
focusing on the core domain of each construct. The use of these abbreviated measures can simplify
potentially complex and multi-dimensional constructs as rather simple and unidimensional ones.
Further studies should replicate these current findings with a rigorous research design based on full
scales to retain the complexity of given conceptual domains and accompanying phenomena.
Third, the current data were collected from Korean organizations. Thus, the cultural values of Korea
could affect the current pattern of results. The Korean society is often called “collectivist,”meaning that
the group takes precedence over the individual (Cho, Mallinckrodt, & Yune, 2010), and the loyalty of
employees to an organization is valued in Korean firms. This collectivist tendency and distinct culture of
Korean firms could affect the current empirical findings on group identification and conformity
pressure. Further empirical studies in diverse national and cultural settings should bolster the general-
izability of the multilevel dynamics of positive deviance observed in the present study.
These limitations reveal the directions for further research efforts to enrich the understanding of
positive deviance in work teams and organizations. The present study offers meaningful theoretical and
practical contributions by isolating an intervening mechanism underlying the relationship between
identity cognition and positive deviance, as well as a group-level contingency, which modifies such
relationship. Notwithstanding, each of the construct in the current theoretical model and proposed
relationships among constructs can involve considerably complex dynamics. Their relationships can
manifest in different ways depending on a number of moderating contingencies. Further theoretical and
empirical endeavors may address this conceptual complexity and potential additional moderating
contingencies to overcome the simplification of a given phenomenon. For example, in the present
study, positive deviance largely focused on performance improvement. However, positive deviance can
take different forms toward various purposes, such as facilitating self-improvement, helping others
learn, and ensuring full consideration of issues. Future work can expand the current work to involve
these various forms of positive deviance as a promising direction. Finally, further studies should explore
alternative mediating processes and moderating contingencies across levels of analysis and extend to the
organization level to include organizational contingencies. These organizational contingencies may
include culture, strategy, and reward systems that directly shape or operate as boundary conditions of
positive deviance of employees.
Acknowledgement
This research was supported by the Institute of Management Research and the Institute of Industrial Relations at Seoul
National University in South Korea.
ORCID
Moon Joung Kim http://orcid.org/0000-0002-2774-702X
Jin Nam Choi http://orcid.org/0000-0001-7890-3195
THE JOURNAL OF SOCIAL PSYCHOLOGY 741

References
Acharya, P., & Taylor, R. (2012). Innovative deviance: An investigation of the relationships between social control,
creativity and innovation in organizations. Academy of Business Research Journal, 1. Retrieved from https://ssrn.
com/abstract=2519799
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Newbury Park, CA: Sage.
Amabile, T. M. (1996). Creativity in context. New York, NY: Westview.
Andrews, J., & Smith, D. C. (1996). In search of the marketing imagination: Factors affecting the creativity of
marketing programs for mature products. Journal of Marketing Research,33(2), 174–187. doi:10.2307/3152145
Appelbaum, S. H., Iaconi, G. D., & Matousek, A. (2007). Positive and negative deviant workplace behaviors: Causes,
impacts, and solutions. Corporate Goverance,7(5), 586–598.
Ashforth, B. E., & Mael, F. (1989). Social identity theory and the organization. Academy of Management Review,14,
20–39.
Ashmore, R. D., Deaux, K., & Mclaughlin-Volpe, T. (2004). An organizing framework for collective identity:
Articulation and significance of multidimensionality. Psychological Bulletin,130,80–114.
Becker, T. E. (2005). Potential problems in the statistical control of variables in organizational research: A qualitative
analysis with recommendations. Organizational Research Methods,8(3), 274–289. doi:10.1177/1094428105278021
Brewer, M. B. (1991). The social self: On being the same and different at the same time. Personality & Social Psychology
Bulletin,7, 475–482.
Brewer, M. B., & Silver, M. D. (2000). Group distinctiveness, social identification, and collective mobilization. In: S.
Stryker, T. J. Ownes, & R. W. White (Eds.), Self, identity, and social movements (pp. 153–171). Minneapolis, MN:
University of Minnesota Press.
Brief, A. P., Buttram, R. T., & Dukerich, J. M. (2001). Collective corruption in the corporate world: Toward a process
model. In: M. E. Turner (Ed.), Groups at work: Theory and research (pp. 471–499). Mahwah, NJ: Lawrence Erlbaum.
Brown, C. M., Olkhov, Y. M., Bailey, V. S., & Daniels, E. R. (2015). Gender differences in the social cost of affective
deviance. The Journal of Social Psychology,155(6), 535–540. doi:10.1080/00224545.2015.1018859
Cho, Y.-J., Mallinckrodt, B., & Yune, S.-K. (2010). Collectivism and individualism in bicultural coping: South Korean
undergraduates adjustment to college. Asian Journal of Counseling,17,82–104.
Damaraju, N. L., Barney, J., & Dess, G. (2010). Stigma and entrepreneurial risk taking. Paper presented at the The
Summer Conference 2010 on “Opening Up Innovation: Strategy, Organization and Technology,”Imperial College
London Business School, London, UK.
De Cremer, D. (2001). Psychology in organizations: The social identity approach. European Bulletin of Social
Psychology,13,28–30.
Desivilya, H. S., & Eizen, D. (2005). Conflict management in work teams: The role of social self-efficacy and group
identification. The International Journal of Conflict Management,16, 185–211.
Ellis, D. G., & Fisher, B. A. (1994). Small group decision making: Communication and the group process. New York
(NY): McGraw-Hill.
Franzese, A. T. (2013). Motivation, motives, and individual agency. In: J. DeLamater, & A. Ward (Eds.), Handbook of
social psychology (pp. 281–318). Madison, WI: Springer.
Fu, P. P., Tsui, A. S., & Dess, G. (2006). The dynamics of guanxi in Chinese high-tech firms: Implications for
knowledge management and decision making. Management International Review,46(3), 277–305.
Galperin, B. L. (2012). Exploring the nomological network of workplace deviance: Devloping and validating a measure
of constructive deviance. Journal of Applied Social Psychology,42(12), 2988–3025.
Graham, J. W. (1986). Principled organizational dissent: A theoretical essay. In: L. L. Cummings, & B. M. Staw (Eds.),
Research in organizational behavior (Vol. 8, pp. 1–52). Greenwich, CT: JAI Press.
Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: A regression-based
approach. New York, NY: The Guilford Press.
Henningsen, D. D., Henningsen, M. L. M., Eden, J., & Cruz, M. G. (2006). Examining the symptoms of groupthink and
retrospective sensemaking. Small Group Research,37(1), 36–64. doi:10.1177/1046496405281772
James, L. R., Demaree, R. G., & Wolf, G. (1984). Estimating within-group inter-rater reliability with and without
response bias. Journal of Applied Psychology,69,86–98.
Janis, I. L. (1972). Victims of groupthink. Boston, MA: Houghton Mifflin.
Janssen, O., & Huang, X. (2008). Us and me: Team identification and individual differentiation as complementary
drivers of citizenship and creative behaviors of members of middle management teams. Journal of Management,34,
69–88.
Jones, A. L. (1998). Random acts of kindness: A teaching tool for positive deviance. Teaching Sociology,26(3), 179–189.
Krau, E. (2008). Work, creativity, inventions and society. Man and Work,16,46–54.
Kusa, D. (2006). Social benefits of being creative: Creativity as a positive deviance. Studia Psychologica,48(3), 229–240.
LeBreton, J. M., & Senter, J. L. (2008). Answers to 20 questions about interrater reliability and interrater agreement.
Organizational Research Methods,11(4), 815–852. doi:10.1177/1094428106296642
742 M. J. KIM AND J. N. CHOI

Madjar, N., Greenberg, E., & Chen, Z. (2011). Factors for radical creativity, incremental creativity, and routine,
noncreative performance. Journal of Applied Psychology,96(4), 730–743. doi:10.1037/a0022416
Major, B., & O’Brien, L. T. (2005). The social psychology of stigma. Annual Review of Psychology,56(1), 393–421.
Mandel, N. (2003). Shifting selves and decision making: The effects of self-construal priming on consumer risk-taking.
Journal of Consumer Research,30(1), 30–40.
Mertens, W., Recker, J., Kohlborn, T., & Kummer, T. F. (2016). A framework for the study of positive deviance in
organizations. Deviant Behavior,1–20. doi:10.1080/01639625.2016.1174519
Morrison, E. W. (2006). Doing the job well: An investigation of pro-social rule breaking. Journal of Management,32
(1), 5–28. doi:10.1177/0149206305277790
Morrison, E. W. (2011). Employee voice behavior: Integration and directions for future research. The Academy of
Management Annals,5, 373–412.
Morrison, E. W., Wheeler-Smith, S. L., & Kamdar, D. (2011). Speaking up in groups: A cross-level study of group
voice climate and voice. Journal of Applied Psychology,96(1), 183–191. doi:10.1037/a0020744
Muthen, L. K., & Muthen, B. O. (1998-2007). Mplus user’s guide (Vol. 5). Los Angeles, CA: Muthen & Muthen.
Near, J. P., & Miceli, M. P. (1985). Organizational dissidence: The case of whistle-blowing. Journal of Business Ethics,4,
1–16.
O’Reilly, C. A., & Chatman, J. (1986). Organizational commitment and psychological attachment: The effects of
compliance, identification and internalization on prosocial behavior. Journal of Applied Psychology,71, 492–499.
Pascale, R., Sternin, J., & Sternin, M. (2010). The power of positive deviance. Boston, MA: Harvard Business Press.
Podsakoff, P., MacKenzie, S., Lee, J., & Podsakoff, N. (2003). Common method biases in behavioral research: A critical
review of the literature and recommended remedies. Journal of Applied Psychology,88, 879–903.
Preacher, K. J., Zhang, Z., & Zyphur, M. J. (2011). Alternative methods for assessing mediation in multilevel data: The
advantages of multilevel SEM. Structural Equation Modeling: A Multidisciplinary Journal,18(2), 161–182.
doi:10.1080/10705511.2011.557329
Sheldon, K. M., & Bettencourt, B. A. (2002). Psychologial needs and subjective well-being in social groups. British
Journal of Social Psychology,41,25–38.
Spreitzer, G. M., & Sonenshein, S. (2004). Toward the construct definition of positive deviance. American Behavioral
Scientist,47(6), 828–847. doi:10.1177/0002764203260212
Stangor, C. (2004). Social groups in action and interaction. New York, NY: Psychology Press.
Swaab, R., Postmes, T., & Spears, R. (2004, July). Attraction, identification and motivation in small group negotiations.
Paper presented at the IACM 17th Annual Conference.
Tabachnick, B. G., & Fidell, L. S. (2013). Using multivariate statistics (6th ed.). Boston, MA: Pearson.
Tse, H. H. M., & Chiu, W. C. K. (2014). Transformational leadership and job performance: A social identity
perspective. Journal of Business Research,67(1), 2827–2835. doi:10.1016/j.jbusres.2012.07.018
Tyler, T. R., & Blader, S. L. (2003). The group engagement model: Procedural justice, social identity, and cooperative
behavior. Personality and Social Psychology Review,7, 349–361.
Umphress, E. E., Bingham, J. B., & Mitchell, M. S. (2010). Unethical behavior in the name of the company: The
moderating effect of organizational identification and positive reciprocity beliefs on unethical pro-organizational
behavior. Journal of Applied Psychology,95(4), 769–780. doi:10.1037/a0019214
Vadera, A. K., Pratt, M. G., & Mishra, P. (2013). Constructive deviance in organizations: Integrating and moving
forward. Journal of Management. doi:10.1177/0149206313475816
Van Knippenberg, D. (2000). Work motivation and performance: A social identity perspective. Applied Psychology: An
International Review,49, 357–371.
Venkataramani, V., & Tangirala, S. (2010). When and why do central employees speak up? An examination of
mediating and moderating variables. Journal of Applied Psychology,95(3), 582–591. doi:10.1037/a0018315
Vignoles, V. L. (2011). Identity motives. In S. J. Schwartz, K. Luyckx, & V. L. Vignoles (Eds.), Handbook of identity
theory and research (pp. 403–432). New York, NY: Springer Science Business Media, LLC.
Warren, D. E. (2003). Costructive and destructive deviance in organizations. Academy of Management Review,28(4),
622–632. doi:10.5465/AMR.2003.10899440
Weber, E. U., Blais, A.-R., & Betz, E. (2002). A Domain-specific risk-attitude scale: Measuring risk perceptions and risk
behaviors. Journal of Behavioral Decision Making,15, 263–290.
THE JOURNAL OF SOCIAL PSYCHOLOGY 743