Yuan & Woodman (2010)

!
Academy of Management Journal

2010, Vol. 53, No. 2, 323–342.
INNOVATIVE BEHAVIOR IN THE WORKPLACE: THE ROLE OF

PERFORMANCE AND IMAGE OUTCOME EXPECTATIONS
FEIRONG YUAN
The University of Kansas
RICHARD W. WOODMAN
Texas A&M University
Why do employees engage in innovative behavior at their workplaces? We examine

how employees’ innovative behavior is explained by expectations for such behavior to
affect job performance (expected positive performance outcomes) and image inside
their organizations (expected image risks and expected image gains). We found signif-
icant effects of all three outcome expectations on innovative behavior. These outcome
expectations, as intermediate psychological processes, were shaped by contextual and
individual difference factors, including perceived organization support for innovation,
supervisor relationship quality, job requirement for innovativeness, employee reputa-
tion as innovative, and individual dissatisfaction with the status quo.
The importance of innovation for organizational individual creativity (Amabile, 1996; Woodman et
effectiveness is widely accepted (e.g., Janssen, Van De al., 1993). Since generating creative ideas is a com-
Vliert, & West, 2004; Van de Ven, 1986; Woodman, ponent of innovation behavior, the intrinsic inter-
Sawyer, & Griffin, 1993). In particular, employee in- est factor could also be useful in explaining em-
novative behavior (e.g., developing, adopting, and ployee innovation. However, innovation is a risky
implementing new ideas for products and work endeavor. Engaging in innovative acts in a work-
methods) is an important asset that enables an organ- place brings benefits and costs for employees be-
ization to succeed in a dynamic business environ- yond a sense of intrinsic enjoyment (Janssen, 2003;
ment (Kanter, 1983; West & Farr, 1990a). A variety of Janssen et al., 2004). In their conceptual model,
factors have been studied as important antecedents to Farr and Ford (1990) identified expected payoffs as
individuals’ innovation, such as organization culture an important proximal antecedent to individual in-
and climate (e.g., Scott & Bruce, 1994), relationship novation. Yet few studies have directly theorized
with their supervisors (e.g., Janssen & Van Yperen, and tested the effects of these expectations. For
2004), job characteristics (e.g., Oldham & Cummings, example, Scott and Bruce (1994) examined how
1996), social/group context (e.g., Munton & West, contextual and individual difference variables im-
1995), and individual differences (e.g., Bunce & West, pact innovative behavior through the perceptions
1995). Yet research evidence regarding the interme- of organization climate for innovation, which was
diate psychological processes that would explain conceptualized as affecting innovative behavior be-
how and why different individual and contextual cause it signals expectations and potential out-
antecedents affect innovative behavior remains in- comes of behavior (Scott & Bruce, 1994). This im-
conclusive and underdeveloped (Shalley, Zhou, & portant study integrated a number of antecedents to
Oldham, 2004; West & Farr, 1989). individual innovation. But still, like the majority of
One relevant psychological aspect suggested by other innovation studies, Scott and Bruce’s work did
previous studies is an individual’s intrinsic interest not examine the nature of these outcome expectations
in his or her task, which would positively affect and their effects on individual innovation.
People act on the basis of consequences or, more
specifically, the expected consequences of their be-
This investigation was funded in part by the Mays Busi- havior, according to behavioral theories such as the
ness School of Texas A&M University and the University of
expectancy theory of motivation (Vroom, 1964). Al-
Kansas General Research Fund, allocation #2302044. We
thank our editor and the three anonymous reviewers for
though prior research has suggested that expected
their insightful suggestions. We thank Wendy Boswell, payoffs or outcomes of innovative behavior can be
Wendy Wood, Jing Zhou, Jim Guthrie, Haiyang Li, Shung important psychological considerations behind in-
Jae Shin, Michael West, Doug May, Jay Lee, Catherine dividual innovation, studies that directly theorize
Schwoerer, Bert Cannella, and Hui Liao for their valuable and test the effects of these outcome expectations
comments and suggestions on drafts of this article. are conspicuously missing. Scholars still lack a
323
Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyright holder’s express
written permission. Users may print, download or email articles for individual use only.
324 Academy of Management Journal April
good understanding of what consequences are im- as one type of innovative behavior because innova-
portant for innovative behavior and how the expec- tive behavior includes not only generating novel
tations for these consequences affect employee in- ideas by oneself but also adopting others’ ideas that
novation in the workplace. are new to one’s organization or work unit (Wood-
Drawing from two theoretical perspectives in the man et al., 1993). Also, creative behavior concerns
innovation literature, the efficiency-oriented perspec- new idea generation, whereas innovative behavior
tive and the social-political perspective, in this study includes both the generation and implementation
we examine how an employee’s innovative behavior of new ideas (Shalley et al. 2004; Zhou, 2003).
is affected by his or her expectations for such behav- Most innovation research has followed what we
ior’s potential influence on job performance (ex- call the efficiency-oriented perspective, in which it is
pected positive performance outcomes) and his or her assumed that organizations make rational decisions
image inside the organization (expected image risks in adopting innovation to maximize their efficiency
and expected image gains). Further, we examine how gains (see Abrahamson [1991] and Rogers [1983] for
these performance and image outcome expectations, reviews on this dominant perspective, which was the
as intermediate psychological processes, are shaped implicit assumption underlying most early studies on
by distal contextual and individual difference ante- innovation adoption and diffusion). This efficiency-
cedents. The major contributions of this study are oriented perspective is in part responsible for the
twofold. First, this study is the first attempt to directly proinnovation bias (that is, the view that innovation
theorize and test the major outcome expectations as- is beneficial for organizations and individuals) in the
sociated with innovative behavior. Revealing these existing literature (Farr & Ford, 1990; Kimberly, 1981;
expectations contributes to understanding of why em- Van de Ven, 1986). Several slightly different descrip-
ployees innovate in a workplace. Second, by testing tive labels (e.g., “rational models,” “efficiency-choice
the relationship between distal antecedents and out- perspective”) are also used in the literature for what
come expectations, this study sheds light on how we call the efficiency-oriented perspective. In each
contextual and individual difference factors could case, the underlying assumption is that innovations
affect employee innovation indirectly by shaping serve the economic function of improving efficiency
these intermediate psychological processes. Findings and that innovation decisions are based on expected
on these processes both contribute to theory develop- positive performance outcomes.
ment on individual innovation and help to suggest More recently, innovation research has started to
possible interventions to encourage employee pay attention to social-political processes (Dean,
innovation. 1987; Dyer & Page, 1988; Wolfe, 1994) that shed
light on how innovation is actually carried out in
the real world rather than how it should be done. In
EXPLAINING INNOVATIVE BEHAVIOR USING
particular, studies have shown the importance of
PERFORMANCE AND IMAGE
image or legitimacy considerations in explaining
OUTCOME EXPECTATIONS
innovation adoption decisions (e.g., Tolbert &
Drawing on West and Farr (1989, 1990b), we Zucker, 1983; Westphal, Gulati, & Shortell, 1997).
define innovative behavior as an employee’s inten- Similarly, in his theoretical framework, Abraham-
tional introduction or application of new ideas, son (1991) suggested that the innovation process
products, processes, and procedures to his or her can be understood as “fad” or “fashion” whereby
work role, work unit, or organization. Examples of innovations have sometimes been adopted for their
such behavior include searching out new technol- symbolic meaning, such as signaling innovative-
ogies, suggesting new ways to achieve objectives, ness, rather than to boost organizations’ economic
applying new work methods, and investigating and performance. We refer to this alternative view as
securing resources to implement new ideas. In the social-political perspective; it emphasizes the
keeping with Kanter (1988), Janssen (2000), and symbolic function of innovative acts and the influ-
Scott and Bruce (1994), we conceptualize innova- ence of image considerations on innovation deci-
tive behavior as complex behavior consisting of sions beyond an efficiency calculation.
activities pertaining to both the generation/intro- Wolfe (1994), in his review of the innovation
duction of new ideas (either by oneself or adopted literature, pointed out that one important barrier to
from others) and the realization or implementation knowledge accumulation in innovation research is
of new ideas. One related construct in the literature that researchers have limited their scope of inquiry
is creative behavior, which refers to behavior per- by working within a single theoretical perspective.
taining to the generation of ideas that are both He noted that “the adoption of a single perspective,
novel and useful (Amabile, 1988; Oldham & Cum- whatever that might be, limits the scope of a re-
mings, 1996). Creative behavior can be considered searcher’s inquiry and thus limits the extent to
2010 Yuan and Woodman 325
which he/she can capture the innovation process, itive performance outcomes, therefore, is subjectively
one which is complex, nonlinear, tumultuous, and defined by each particular employee.
opportunistic” (Wolfe, 1994: 416). Abrahamson Improved efficiency and job performance in-
(1991) suggested that one way to overcome this crease the competitiveness and success of an em-
barrier is to combine multiple perspectives so that ployee. Following the efficiency-oriented perspec-
each captures some aspect of the complex innovation tive in understanding innovation, we contend that
process. Following this approach, in the following employees are more likely to engage in innovative
sections we draw from both the efficiency-oriented behavior when they expect such behavior to benefit
perspective and the social-political perspective to their work. Figure 1 depicts this hypothesized re-
theorize about the outcome expectations associated lationship, along with the others comprising our
with innovative behavior. The efficiency-oriented theoretical model.
perspective provides insight into the effect of ex-
pected positive performance outcomes; the social- Hypothesis 1. Expected positive performance
political perspective provides insight into the effects outcomes are positively related to innovative
of expected image risks and expected image gains on behavior.
innovative behavior.
Expected Image Risks and Expected Image
Expected Positive Performance Outcomes: The Gains: The Social-Political Perspective
Efficiency-Oriented Perspective
People’s reality is, at least partially, socially con-
One major reason people innovate in the work- structed. Individual behaviors have both technical
place is to bring performance gains. New technol- and symbolic functions. Regardless of whether the
ogies are introduced and new work methods are introduction of new ideas or procedures will help
applied because these are “better” than the existing to improve efficiency or performance, the act of
ones and are expected to bring performance im- engaging in innovative behaviors is a signal; it con-
provement and efficiency gains. Such a belief in veys information about an actor to the social con-
performance gains reflects the dominant efficiency- text. Other people’s potential perceptions or im-
oriented perspective in the innovation literature, pressions are important determinants of individual
according to which the goal of technical efficiency behavior because such impressions influence oth-
guides innovation adoption and diffusion (Abraham- ers’ reactions to the actor and therefore the possi-
son, 1991; Rogers, 1983; Wolfe, 1994). Although the bility for the actor to get necessary resources and
efficiency-oriented perspective has been the domi- social support to achieve goals (Leary & Kowalski,
nant perspective in the field, the effect of expected 1990; Tedeschi & Riess, 1981). The literature in
performance gains on innovation has been mostly impression management supports the importance of
implicitly assumed rather than explicitly studied in image considerations in influencing a variety of be-
the literature. In particular, little research evidence haviors in organizations, such as feedback seeking
exists to explain whether and how the expectation for (Ashford & Northcraft, 1992), organizational citizen-
positive performance outcomes affects employee in- ship behaviors (Rioux & Penner, 2001), and issue
novative behavior in the workplace. selling (Ashford, Rothbard, Piderit, & Dutton, 1998).
Expected performance outcomes are positive In research on innovation, image concerns and
when employees believe that their innovative be- related social-political processes have received
haviors will bring performance improvement or ef- growing attention in organization-level studies
ficiency gains for their work roles or work units. (e.g., Arndt & Bigelow, 2000; Tolbert & Zucker,
The concept of efficiency usually refers to an input- 1983; Westphal, Gulati, & Shortell, 1997). Although
output ratio or comparison (Ostroff & Schmitt, image and impression management issues have
1993; Pennings & Goodman, 1977). Here we define been addressed less in research undertaken at the
“efficiency” broadly and use the term synony- individual level, anecdotal evidence suggests these
mously with “performance” to describe objective or issues affect employee innovative behavior at work.
actual task performance. Specific examples of pos- For example, in West’s (1989) study of community
itive performance outcomes include increased pro- nurses, participating nurses listed “other’s reaction”
ductivity and work quality, decreased error rate, as an important concern that prevented them from
increased ability to achieve goals and objectives, being innovative. In a product design firm, Sutton
and improved general job performance. Obviously, and Hargadon (1996) found that design engineers
relevant performance dimensions vary across differ- used brainstorming sessions as “prestige” or “status
ent job positions, and people weigh various aspects of auctions”—that is, as opportunities to impress their
performance in different ways. What constitutes pos- peers and establish favorable social images.
FIGURE 1
Explaining Innovative Behavior using Performance and Image Outcome Expectations:
The Theoretical Modela
Perceived organization The social-political

support for innovation perspective
Expected image
Supervisor relationship risks
quality
Expected image
gains
Innovativeness as a job
requirement Individual
The efficiency-oriented innovative
perspective behavior
Reputation as Expected positive

innovative performance
outcomes
Dissatisfaction with the

status quo
a
Solid lines represent relationships hypothesized to be positive. Dashed lines represent relationships hypothesized to be negative.
One basic distinction in the impression manage- evaluations. This tendency to avoid image risks rep-
ment literature is that between defensive and asser- resents the self-protective or defensive impression
tive impression management (Arkin, 1981; Schlen- management motive. Secondly, employees may en-
ker, 1980). Tetlock and Manstaed provided a good gage in innovative behaviors as a deliberate effort to
discussion of this distinction: “Defensive impres- improve image. An employee may suggest new ideas
sion management is designed to protect an individ- to a supervisor in order to appear competent and
ual’s established social image; it is triggered by conscientious. Engaging in innovative behavior to
negative affective states (e.g., embarrassment, pursue image gain represents the acquisitive or asser-
shame). Assertive impression management is de- tive impression management motive.
signed to improve an individual’s social image; it is
Hypothesis 2. Expected image risks are nega-
triggered by self-enhancing motives activated by
tively related to innovative behavior.
perceived opportunities for creating favorable im-
pressions on others” (1985: 61). The difference be- Hypothesis 3. Expected image gains are posi-
tween avoiding image risks and pursuing image tively related to innovative behavior.
gains, therefore, is not a matter of degree. They
represent different affective states and individual
Distal Antecedents Related to Performance and
motives (“getting along” vs. “getting ahead”) (Wolfe,
Image Outcome Expectations
Lennox, & Cutler, 1986).
Following the social-political perspective in under- Both the organizational context an employee is
standing innovation, and in keeping with the impres- embedded in and the individual characteristics of
sion management literature, we contend that both this employee affect his or her belief about what
types of impression management may affect em- consequences will result from his or her innovative
ployee innovative behavior. First, potential image behavior. As proximal psychological processes,
risks will constrain employee innovativeness. An em- performance and image outcome expectations are
ployee may choose to “play it safe” and avoid “rock- shaped by distal contextual and individual differ-
the-boat” innovative behaviors in order to look so- ence antecedents. Examining the relationship be-
cially appropriate and to prevent negative social tween distal antecedents and these outcome expec-
tations is important, because it explains the sources for these relationships by allowing these distal vari-
of variance among individuals in their outcome ables to covary when testing our model.
expectations and sheds light on how these anteced- Perceived organization support for innovation.
ent factors may affect employee innovation indi- Organization climate is an important contextual
rectly by shaping these intermediate processes. factor that signals expectations for behavior and
West and Farr’s (1989) theoretical framework of potential outcomes of these behaviors (James, Hart-
individual innovation presents five major types of man, Stebbins, & Jones, 1977). From a social-polit-
factors as important for understanding individual ical perspective, organization support for inno-
innovation at work: organizational factors, relation- vation, which can manifest as a proinnovation
ships at work with one’s supervisor, job character- climate or culture (Amabile, 1988; Kanter, 1988;
istics, group or social factors, and individual char- Scott & Bruce, 1994), delivers organizational values
acteristics. In keeping with these conceptual and norms that affect the potential image gains and
angles, we examine how the following contextual image risks associated with employee innovative
and individual difference factors shape an employ- behavior. If an organization’s norms favor change,
ee’s performance and image outcome expectations: rather than tradition for its own sake, its members
perceived organization support for innovation, su- will seek to initiate change to be culturally appro-
pervisor relationship quality, innovativeness as a priate (Farr & Ford, 1990). An organization climate
job requirement, reputation as innovative, and in- for innovation delivers “expectancies” and “instru-
dividual dissatisfaction with the status quo. From mentalities” (Scott & Bruce, 1994) so that organiza-
the efficiency-oriented perspective, these five faction members understand that being innovative is a
tors capture, from different angles, how organiza- desirable image and engaging in innovative behavior
tional values and beliefs, supervisor support, the
will make them look good (that is, higher expecta-
specific nature of an employee’s job, his or her
tions for image gains exist). Moreover, a proinnova-
personal reputation, and his or her dissatisfaction
tion climate encourages innovative behavior because
with the performance condition of an organization
it legitimates experimentation (West & Wallace,
could affect the employee’s belief that his or her
1991), creates psychological safety for trial and error,
innovative behavior will bring performance gains.
and reduces the image risk involved in innovation
From a social-political perspective, these anteced-
attempts (Ashford et al., 1998).
ents depict major social and political factors in the
Members in an organization with strong support
employee’s work environment that define what is
for innovation will also perceive their innovative
appropriate, what is desirable, and what resources
and opportunities the individual has to both pro- behavior as more beneficial in bringing perfor-
tect and advance him- or herself in the organiza- mance gains. From an efficiency-oriented perspec-
tion, all of which influence the employee’s assess- tive, a favorable organization climate for innova-
ment of the potential image risks and image gains tion communicates the need for change and
associated with innovative behavior. demonstrates the belief that innovation will make
It is important to note that many contextual and the organization more efficient and successful.
individual difference factors conceptualized as dis- These values and beliefs, ingrained in the culture of
tal antecedents in our model have been examined the organization, will be transmitted to and become
in prior research with regard to their relationships internalized by employees through the organiza-
with each other. For example, Scott and Bruce’s tion’s socialization processes (Chatman, 1991; Har-
(1994) model of innovative behavior tested how rison & Carroll, 1991). Employees working in or-
leader-member exchange affects support for inno- ganizations with strong support for innovation,
vation, which in turn affects innovative behavior. therefore, are more likely than those not in such
Two similar constructs, supervisor relationship organizations to share the belief that innovation is
quality and perceived organization support for in- valuable and will bring performance gains.
novation, are examined in our study. The major
focus of the current study, however, was not theo- Hypothesis 4a. Perceived organization support
rization about and testing of the relationships for innovation is positively related to expected
among these antecedents. Instead, we focused on positive performance outcomes of innovative
how these distal antecedents relate to performance behavior.
and image outcome expectations, which was what
was missing from the existing literature. At the Hypothesis 4b. Perceived organization support
same time, we recognized that relationships exist for innovation is negatively related to expected
among these distal antecedents, and we accounted image risks in innovative behavior.
Hypothesis 4c. Perceived organization support Hypothesis 5a. Supervisor relationship quality
for innovation is positively related to expected is positively related to expected positive per-
image gains from innovative behavior. formance outcomes of innovative behavior.
Hypothesis 5b. Supervisor relationship quality
Supervisor relationship quality. In addition to
is negatively related to expected image risks in
the organization context, an employee’s relation-
innovative behavior.
ship with his or her supervisor represents an im-
portant aspect of the immediate work environment Hypothesis 5c. Supervisor relationship quality
that influences the employee’s belief in the pos- is positively related to expected image gains
sible performance and image outcomes of his or from innovative behavior.
her innovative attempts. Leader-member exchange
Innovativeness as a job requirement. An em-
(LMX) theory (Graen, 1976; Graen, Novak, & Som-
ployee’s official work role is another contextual
merkamp, 1982; Graen & Uhl-Bien, 1995) suggests
factor that could affect the potential consequences
that subordinates who have high-quality relation-
of the employee’s innovative behavior. Kanter
ships with their supervisors are given greater re-
(1988) suggested that the obligations of one’s posi-
sources, decision latitude, and freedom in return tion can serve as an initial impetus that activates
for greater loyalty and commitment. Contemplating innovation. From an efficiency-oriented perspec-
and experimenting with innovative ideas to im- tive, a job requirement for innovation explicitly
prove existing products and processes often require specifies the relevance of innovative behavior to
additional time, resources, and freedom at work successful performance. Compared with others,
(Kanter, 1988). Greater resources and support from a employees who perceive innovativeness as part of
supervisor increase the odds that innovative behavior their job requirements are therefore more likely to
will be successful. From an efficiency-oriented per- believe that generating, adopting, and implement-
spective, therefore, employees with high-quality rela- ing innovative ideas will benefit their work. At the
tionships with their supervisors are likely to be more same time, from a social-political perspective, a job
confident that their innovative behavior will result in requirement for innovativeness also represents ex-
performance and efficiency gains. ternal demand and expectations for innovativeness,
A high-quality relationship with a supervisor, which legitimize the job incumbent’s innovative be-
characterized by mutual trust and respect (Graen & havior. Moreover, research has suggested that an au-
Uhl-Bien, 1995), also constitutes a valuable politi- dience tends to evaluate change-initiated behaviors
cal resource that increases potential image gain and (e.g., issue selling, innovation) more favorably when
reduces potential image loss for an innovative em- they are conducted by people whose functional back-
ployee. The desire and motivation of an individual ground or job position supports their behavior (Ash-
influence what he or she perceives (Gilbert, 1998; ford et al., 1998; Daft, 1978). Therefore, when perceiv-
Markus & Zajonc, 1985). Studies in human resource ing innovativeness as part of their job requirements,
management have observed that supervisors tend employees will both feel it is more appropriate to
to evaluate the employees they like and trust in a engage in innovative behavior (i.e., have less concern
more positive way (Judge & Ferris, 1993; Wayne & for image risks) and feel more confident that manag-
Liden, 1995). Therefore, when a supervisor trusts ers and coworkers will consider their new ideas valid
and respects an employee, she or he is more likely and well grounded (i.e., have higher expectations for
to evaluate this employee’s new ideas favorably image gains).
(Zhou & Woodman, 2003) and believe these ideas
Hypothesis 6a. Innovativeness as a job require-
are meaningful and significant, resulting in greater
ment is positively related to expected positive
possibilities of image gain. Moreover, research on
performance outcomes of innovative behavior.
attribution biases suggests that when a supervisor
likes or empathizes with an employee, he or she is Hypothesis 6b. Innovativeness as a job require-
more likely to attribute positive behavior outcomes ment is negatively related to expected image
to the employee’s disposition and negative out- risks in innovative behavior.
comes to the employee’s situation (Green & Mitch-
Hypothesis 6c. Innovativeness as a job require-
ell, 1979; Regan & Totten, 1975). Therefore, em-
ment is positively related to expected image
ployees who are trusted and liked by their gains from innovative behavior.
supervisors will feel more secure when engaging in
innovative behavior and expect less image risk be- Reputation as innovative. In addition to an em-
cause their supervisors are less likely to hold them ployee’s organization, supervisor, and official work
responsible for a failed innovative attempt. role, his or her informal social reputation may also
affect the image and performance outcomes ex- dissatisfaction with the current performance condi-
pected from innovative behavior. The impression tion of his work unit or organization. Such dissatis-
management literature suggests that the impres- faction could arise for a variety of reasons, such as
sions people try to create are affected by how they comparisons with competitors, environmental
think they are currently regarded by others (cf. changes, personality traits (e.g., neuroticism), and the
Leary & Kowalski, 1990). In particular, people are discovery of potential improvement opportunities.
inclined to present themselves in ways that are From an efficiency-oriented perspective, dissatisfac-
consistent with their existing social image (Schlen- tion with the status quo undermines the value of
ker, 1980) because people are considered as more maintaining the current condition and strengthens
socially appropriate and legitimate when their be- people’s beliefs that new ideas, products, or pro-
haviors match others’ categorizations and expecta- cesses will bring performance gains, resulting in more
tions (Zelditch, 2001). From a social-political per- innovative behaviors.
spective, therefore, when an employee regarded as At the same time, from a social-political perspec-
innovative engages in innovative behavior, image tive, a less satisfactory performance condition of a
risk is lower, because the behavior is consistent work unit or organization also serves to justify in-
with the employee’s existing social image, which novative action, reducing image risks and increas-
serves to legitimize the behavior and reduce con- ing the chances of image gain. Change is often more
cerns about inappropriateness. A reputable innova- legitimized when performance is below a targeted
tive person, though, may not necessarily expect level or perceived as a failure (Lant & Mezias,
being innovative to further improve her or his im- 1992). Poor performance is a strong force for coun-
age because behaviors that are consistent with teracting persistence in an established mode of op-
one’s existing image are likely to maintain the latter erating (Tushman & Romanelli, 1985), making it
(Schlenker, 1980) rather than change it. easier to break down resistance and reducing po-
Employees who enjoy a reputation for being in- tential criticisms and image risks associated with
novative are also more likely to internalize the rock-the-boat innovative behaviors. Moreover, poor
value of innovation and more likely to believe that unit or organizational performance provides an op-
innovative behavior will benefit their work. A so- portunity for self-enhancement. When their work
cial reputation as innovative builds into one’s self- unit or organization is less effective, people are
concept. Research on the social bases of self-knowl- more likely to get credit for introducing new tech-
edge suggests that self-concepts are shaped by how nologies and suggesting new ways to achieve ob-
people believe that others perceive them (Shrauger jectives. Employees who demonstrate these behav-
& Schoeneman, 1979). More importantly, people iors are more likely to be considered as
have a need to maintain a positive view about self conscientious and competent (if not heroic), in-
because such self-esteem serves as a valuable affec- creasing the potential for image gain.
tive resource in coping with life challenges
(Steele’s [1988] self-affirmation theory is relevant Hypothesis 8a. Dissatisfaction with the status
here). From an efficiency-oriented perspective, quo is positively related to expected positive
therefore, once an individual has a reputation for performance outcomes of innovative behavior.
being innovative and views her- or himself as an Hypothesis 8b. Dissatisfaction with the status
innovative person, her/his self-esteem will rein- quo is negatively related to expected image
force the positive view of innovation, strengthening risks in innovative behavior.
the belief that innovations will make meaningful
contributions to performance and work efficiency. Hypothesis 8c. Dissatisfaction with the status
quo is positively related to expected image
Hypothesis 7a. Reputation as an innovative per- gains from innovative behavior.
son is positively related to expected positive per-
In sum, our theoretical model suggests that per-
formance outcomes of innovative behavior.
ceived organization support for innovation, super-
Hypothesis 7b. Reputation as an innovative visor relationship quality, innovativeness as a job
person is negatively related to expected image requirement, reputation as innovative, and dissat-
risks in innovative behavior. isfaction with the status quo will affect individual
innovative behavior indirectly by shaping expected
Dissatisfaction with the status quo. Dissatisfaction positive performance outcomes, expected image
is an important individual attitude that makes people risks, and expected image gains.
aware of the need to change (Farr & Ford, 1990) and
the value of introducing new ideas. Dissatisfaction Hypotheses 9a, 9b, and 9c. The relationship
with the status quo is defined here as an employee’s between perceived organization support for
innovation and individual innovative behavior complete questionnaires reduced the final usable
is mediated by (a) expected positive perfor- sample to 216.
mance outcomes, (b) expected image risks, and Our final sample included employees from a
(c) expected image gains. broad cross-section of jobs, including technicians
(21%), sales and marketing personnel (20%), pro-
Hypotheses 10a, 10b, and 10c. The relation-
duction foremen and quality control inspectors
ship between supervisor relationship quality (13%), service representatives (7%), R&D scientists
and individual innovative behavior is medi- and engineers (6%), middle managers (6%), and
ated by (a) expected positive performance out- others such as purchasing agents, human resource
comes, (b) expected image risks, and (c) ex- personnel, and shipping/stock clerks (27%). The
pected image gains. average age range of employee respondents was
Hypotheses 11a, 11b, and 11c. The relation- 40 – 49 years, and 72 percent of them were men.
ship between innovativeness as a job require- Seventy-eight percent of the respondents had at
ment and individual innovative behavior is least some college, and 38 percent had at least
mediated by (a) expected positive performance bachelor’s degrees. Their average organization ten-
outcomes, (b) expected image risks, and (c) ure was 5.55 years, and their average tenure in their
expected image gains. current jobs was 3.25 years.
Hypotheses 12a and 12b. The relationship be-

tween reputation as innovative and individual Measures
innovative behavior is mediated by (a) ex- Unless otherwise indicated, all measures used a
pected positive performance outcomes, and (b) response scale in which 1 was “strongly disagree”
expected image risks. and 5 was “strongly agree.” The Appendix gives all
the scale items.
Hypotheses 13a, 13b, and 13c. The relation-
Individual innovative behavior. This variable
ship between dissatisfaction with the status
was measured by Scott and Bruce’s (1994) six-item
quo and individual innovative behavior is me-
innovative behavior scale. Each participant’s su-
diated by (a) expected positive performance
pervisor indicated how characteristic each behav-
outcomes, (b) expected image risks, and (c)
ior was of the employee being rated on a scale
expected image gains.
ranging from 1, “not at all characteristic,” to 5,
“very characteristic” (! ! .93).1 Following Scott
and Bruce (1994), we combined the six items to
METHODS
Sample and Procedures
1
A total of 73 supervisors rated innovative behavior for
We surveyed 425 full-time employees and their 216 employees in the final sample. Fifty of the 73 supervi-
96 direct supervisors from four U.S. companies in sors provided ratings for multiple subordinates, which
several different industries (information technol- raised the issue of nonindependence of the supervisor rat-
ogy service, computer system development, furni- ings and among the respondents. Even if parameter esti-
ture design/manufacturing, chemical instruments mates are accurate, Bliese and Hanges (2004) suggested that
development/manufacturing). Questionnaires were in models involving only individual-level variables this
nonindependence may result in too many type II errors,
administered via company mail. Completed sur-
and thus a loss of power. The average intraclass correlation
veys were mailed back directly to us. We received
(ICC) values in this study were .41 for individual innovative
287 employee questionnaires (measuring all vari- behavior, .13 for outcome expectations, and .16 for distal
ables except innovative behavior), which consti- variables. These values suggest that nonindependence in
tuted a response rate of 68 percent. No significant our data would have led to an approximately 20 percent
difference was found between respondents and drop of power in our tests of the effects of outcome expec-
nonrespondents with regard to organization mem- tations on innovative behavior and a 5 percent drop in tests
bership and job position. We received 84 question- of the effects of distal variables on outcome expectations
naires from supervisors (with their ratings on em- (Bliese & Hanges, 2004). In our results, the effects of all
three outcome expectations were found to be strong and
ployee innovative behavior and perceptions of the
significant, even after the 20 percent drop in power. For
quality of their relationships with subordinates), tests of distal variable effects on outcome expectations, the
constituting a response rate of 88 percent. The over- 5 percent drop in power was relatively modest, and most of
all response rate (employees and supervisors com- these effects were also found to be significant. Taken to-
bined) was 71 percent. Matching employee and gether, this analysis suggests that nonindependence did not
supervisor questionnaires resulted in 238 pairs. In- significantly affect our results.
create an overall scale of innovative behavior. Objec- .001), indicating the data were of good quality. Since
tive measures of innovative behavior were not ob- what directly affects outcome expectations is an em-
tained because these indicators (e.g., number of re- ployee’s perception of her/his relationship with su-
search reports and patents) were largely unavailable pervisor, we used employee-reported LMX scores to
for the diverse employee sample used in this study. measure this variable. Innovativeness as a job re-
Outcome expectations. Expected positive perfor- quirement, reputation as innovative, and dissatisfac-
mance outcomes (! ! .77) was measured by three tion with the status quo were measured by items
items modified from House and Dessler’s (1974) out- developed for this study (!’s ! .85, .78, .75, respec-
come expectancy scale. Expected image risks (! ! tively; please refer to the Appendix for scale items).
.77) was measured by three items adapted from Ash- Control variables. We controlled for employee
ford’s (1986) measure of image risk. Expected image intrinsic interest in innovative activities using five
gains (! ! .86) was measured by four items adapted items from Tierney, Farmer, and Graen (1999) rated
from Ashford et al.’s (1998) measures of image out- on a scale ranging from 1, “strongly disagree,” to 6,
comes associated with issue selling. “strongly agree” (! ! .85). We measured education
Distal antecedents. Perceived organization sup- level, hierarchical distance, and organization tenure
port for innovation was measured by 13 items (! ! to control for the knowledge an employee can draw
.92) measuring support for creativity and tolerance on to innovate (Mumford & Gustafson, 1988; Scott &
of differences in Scott and Bruce’s (1994) support Bruce, 1994) and the employee’s access to organiza-
for innovation scale.2 We conducted a confirmatory tion resources for conducting innovative behavior
factor analysis (CFA) using LISREL 8.72 (Jöreskog & (Aiken, Bacharach, & French, 1980; Daft, 1978; Kim-
Sörbom, 1996a) on the 13 items to compare a one- berly & Evanisko, 1981). Table 1 shows the coding for
factor model in which all items loaded on a single education level. For hierarchical distance, each re-
factor (as found in Scott and Bruce [1994]) with a spondent reported the number of levels his/her posi-
two-factor model in which items loaded on their tion was below the president of his/her company, and
corresponding support for creativity and tolerance we divided the reported number by the total number
of differences dimensions. The two-factor model fit of hierarchical levels in the particular organization to
the data significantly better than the one-factor create our measure. The higher the score, the lower
model ("diff2 [1, n ! 216] ! 77.32, p " .001). To the respondent’s position in his or her organization.
equalize the measurement weighting for the two Job positions and organization membership were ex-
dimensions, we used the two scores for support for plored as controls, but they did not change the results
creativity and tolerance of differences as manifest of our hypothesis testing. Intrinsic interest had a sig-
indicators of this latent factor in the following CFA nificant effect on innovative behavior in our study,
and in hypothesis testing. Supervisor relationship and the effects are reported below.
quality was measured by the seven-item LMX scale
developed by Graen and colleagues (1982). Em-
Confirmatory Factor Analysis
ployees completed the member’s version of the
scale (! ! .90), and their corresponding supervisors We conducted a CFA to further validate our mea-
completed the leader’s version (! ! .82). The cor- sures. The hypothesized 13-factor measurement
relation between employee and supervisor reports model (including dependent variable, independent
in this study was positive and significant (r ! .42, p " variables, and control variables) was tested by sub-
mitting raw data to LISREL 8.72 and requesting an
analysis based on the covariance matrix (Jöreskog &
2
Scott and Bruce’s (1994) 16-item measure for support Sörbom, 1996a). Fit indexes indicated a good fit for
for innovation, which loaded on one factor in their study, the hypothesized model ("2[785, n ! 216] ! 1,219.49,
was intended to measure three dimensions: support for " .01, "2/df ! 1.55; RMSEA ! .05, SRMR ! .06,
creativity, tolerance of differences, and perceptions of CFI ! .96, NNFI ! .95, IFI ! .96). All items loaded
reward-innovation dependency. The 13 items measuring significantly on the latent constructs they were de-
support for creativity and tolerance of differences were signed to measure. Factor loadings are listed in pa-
drawn from the support for creativity and tolerance of rentheses after the relevant items in the Appendix. To
differences subscales in Siegel and Kaemmerer’s (1978)
further validate our measures for the three outcome
original innovative climate measure. In this study, these
13 items were used to measure perceived organization
expectations, we compared the hypothesized mea-
support for innovation. We did not use the reward-inno- surement model with three alternative measurement
vation dependency items because the support for creativ- models to test the following possibilities: (1) the three
ity and tolerance of differences items are more estab- outcome expectations loading on one general out-
lished measures for support for innovation and they come expectation factor rather than three factors, (2)
seem to represent the construct well. expected image gains and expected image risks load-
TABLE 1
Means, Standard Deviations, and Correlationsa
Variable Mean s.d. 1 2 3 4 5 6 7 8 9 10 11 12
1. Individual innovative 3.17 0.93 (.93)

behavior
2. Expected positive 3.92 0.65 .29** (.77)
performance outcomes
3. Expected image risks 2.17 0.77 #.17* #.24** (.77)
4. Expected image gains 3.65 0.79 .05 .45** #.31** (.86)
5. Perceived organization 3.12 0.73 .03 .18** #.41** .31** (.92)
support for innovation
6. Supervisor relationship 3.76 0.80 .22** .27** #.31** .37** .45** (.90)
quality
7. Innovativeness as a job 3.31 0.90 .18* .51** #.36** .48** .34** .39** (.85)
requirement
8. Reputation as innovative 3.35 0.83 .32** .47** #.08 .30** .08 .13 .47** (.78)
9. Dissatisfaction with the 3.30 0.78 #.07 .00 .30** #.07 #.41** #.39** #.19** .02 (.75)
status quo
10. Intrinsic interest 4.91 0.74 .30** .40** #.03 .12 #.06 .03 .31** .42** .15* (.85)
11. Educationb 3.64 1.34 .13 #.01 .10 .06 #.11 #.04 #.01 .09 .12 .19**
12. Hierarchical distance 1.0 0.46 #.10 .05 #.01 .05 #.08 .07 #.08 #.03 #.02 #.03 #.14*
13. Organization tenure (in 5.55 5.44 .05 #.02 #.01 .07 .09 .06 .14* .02 .04 #.12 #.11 #.16*
years)
a
n ! 216. Where relevant, Cronbach’s coefficient alphas are given on the diagonal in parentheses.
b
Education was coded as follows: “some high school,” 1; “high school diploma,” 2; “some college,” 3; “associate degree,” 4; “bachelor’s
degree,” 5; “master’s degree,” 6; Ph.D., 7.
* p " .05
** p " .01
ing on one general factor of image outcome expecta- Burton (1990) demonstrated statistically that, if
tion, and (3) expected image gains and expected pos- measurement error is incorporated into a model,
itive performance outcomes loading on one general this single-indicator procedure results in path esti-
factor of positive outcome expectations. The results mates virtually identical to the estimates generated
indicated that the hypothesized measurement model by using multiple indicator variables. Specifically,
fit the data significantly better than did any of the we used its scale mean as the single manifest indi-
2
alternative models (!diff [23, n ! 216] ! 444.07, p " cator of each unidimensional latent variable. Be-
2 2
.001; !diff [12, n ! 216] ! 222.52, p " .001; !diff [12, cause of its multidimensionality, perceived organ-
n ! 216] ! 229.75, p " .001, respectively), providing ization support for innovation was indicated by the
support for the distinctiveness of the three outcome two support for creativity and tolerance of differ-
expectation factors. ence scores to equalize the measurement weighting
across the two dimensions. Following Kenny
(1979), James, Mulaik, and Brett (1982), and Wil-
RESULTS
liams and Hazer (1986), to account for measure-
Table 1 reports the means, standard deviations, ment error we set the loading from each single
and correlations for all variables. We tested the indicator to its latent construct as the square root of
hypothesized paths in our theoretical model with the reliability coefficient alpha for each scale mul-
structural equation modeling by submitting raw tiplied by the standard deviation of the indicator,
data to LISREL 8.72 and requesting an analysis and the error variance for each indicator was set
based on the covariance matrix (Jöreskog & Sörbom, equal to one minus alpha times the variance of the
1996a). Because our model contained relatively indicator. As with other variables, we did not as-
large number of variables and measurement items sume that education, hierarchical distance, and or-
and because the study’s major purpose was testing ganization tenure were measured without error. In
structural paths rather than validating the measure- the past, the reliability of a single measure has been
ment model, we followed the single-indicator ap- set at conservative values of either .90 (cf. Ander-
proach in testing the structural model (Allen, son & Gerbing, 1988; Sörbom & Jöreskog, 1982) or
Mahto, & Otondo, 2007; Scott & Bruce, 1994; Wil- .85 (Jöreskog & Sörbom, 1996b). An analysis based
liams & Hazer, 1986). Netemeyer, Johnston, and on either value resulted in the same results we
obtained in this study. We report findings below Perceived organization support for innovation
based on a reliability value of .90 for these three had a significant negative effect on expected image
variables. The exogenous variables in our model risks and a marginally significant positive effect on
were allowed to covary to account for the relation- expected image gains, providing strong support for
ships existing among distal antecedent variables. Hypothesis 4b and some support for Hypothesis 4c.
The hypothesized model provided an adequate No support was found for Hypothesis 4a (the path
fit to the data (!2[32, n ! 216] ! 73.55, p " .01, from perceived support for innovation to expected
!2/df ! 2.30; RMSEA ! .08, SRMR ! .04, CFI ! .96, positive performance outcomes). Supervisor rela-
NFI ! .93, IFI ! .96). Figure 2 presents the results. tionship quality had significant positive effects on
The model accounted for 25 percent of the variance expected positive performance outcomes and ex-
in individual innovative behavior. The distal vari- pected image gains, but a nonsignificant effect on
ables accounted for 55, 37, and 41 percent of the
expected image risks. Thus, support was found for
variance in expected positive performance out-
Hypotheses 5a and 5c, but not for Hypothesis 5b.
comes, expected image risks, and expected image
Hypotheses 6a, 6b, and 6c predict positive effects of
gains, respectively. Eleven of the 17 hypothesized
innovativeness as a job requirement on expected
paths were significant at the .05 level and in the
predicted directions. All three outcome expectations positive performance outcomes and expected im-
related significantly to innovative behavior. Consis- age gains, and a negative effect on expected image
tently with Hypotheses 1 and 2, expected positive risks. The results strongly supported these three
performance outcomes associated positively with in- hypotheses. Reputation as innovative had a signif-
novative behavior (b ! .33, p " .01), and expected icant positive effect on expected positive perfor-
image risks related negatively to innovative behavior mance outcomes, supporting Hypothesis 7a. No
(b ! #.19, p " .05). Contrary to the prediction of support was found for Hypothesis 7b (the path from
Hypothesis 3, expected image gains had a significant reputation to expected image risks). Dissatisfaction
yet negative relationship with innovative behavior with status quo had significant effects on expected
(b ! #.22, p " .01). Figure 2 shows overall results for positive performance outcomes and expected im-
the tests of the hypothesized paths. age gains, but not on expected image risks. There-
FIGURE 2
Results for the Hypothesized Paths (Hypotheses 1– 8c)a
Perceived organization
–.30*
.03
.17 †
–.19*
Expected image
Supervisor relationship –.03
risks
quality .25*
.21*
–.22*
–.36* Expected image
Innovativeness as a job gains
requirement .46*
.37* Individual
.12 innovative
Reputation as behavior
innovative
.38*
Expected positive .33*
.14 .24* performance
Dissatisfaction with outcomes
.20*
the status quo .19*
Intrinsic interest
a
Fit: !2 (32, n ! 216) ! 73.55, p " .01, !2/df ! 2.30; RMSEA ! .08, SRMR ! .04, CFI ! .96, NFI ! .93, IFI ! .96. Parameter estimates
are from the completely standardized solution.
†
p " .10
* p " .05
fore, support was found for Hypotheses 8a and 8c, Kinnon, Lockwood, and Hoffman (1998) distribution
but not for Hypothesis 8b. of products (P ! z!z") method was used, as recom-
In their review of 14 methods for testing mediation, mended by MacKinnon and colleagues (2002).
MacKinnon, Lockwood, Hoffman, West, and Sheets Twelve indirect effects significant at the .05 level
(2002) suggested that the test for the joint significance were found. Table 2 presents the results of the prod-
of alpha and beta has the best balance of type I error uct of coefficients test. Overall, 9 of the 14 hypothe-
and statistical power across all cases. In this ap- sized mediation relationships (Hypotheses 9b, 10a,
proach, mediation is supported when both the ante- 10c, 11a, 11b, 11c, 12a, 13a, and 13c) were consis-
cedent-mediator path alpha and mediator-outcome tently found to be significant by both tests (high-
path beta are jointly significant in a model in which lighted in bold in Table 2) and were interpreted as
the direct path from antecedent to outcome is also significant results in this study.
modeled. Following this approach, we reran our
model with the direct paths from distal variables to
innovative behavior added. The significance of all DISCUSSION
hypothesized paths remained the same. All three beta
Theoretical Implications
paths were significant, and nine alpha paths were
significant, therefore supporting 9 of the 14 media- This study is the first attempt to examine how the
tion hypotheses. Figure 3 depicts these results of me- expected outcomes of innovative behavior affect
diation testing. Moreover, to provide a direct test of employee innovation at work. We found that the ex-
the statistical significance of the mediated or indirect pectations for potential performance and image con-
effects (i.e., the significance of the alpha-beta prod- sequences significantly affected employee innovative-
ucts) we used the product of coefficients approach ness after we controlled for individual intrinsic
(MacKinnon et al., 2002) to supplement the test for interest and capabilities. In particular, we brought
the joint significance of alpha and beta. The Mac- together two major theoretical perspectives (the effi-
FIGURE 3
Results of Mediation Testing Using the Joint Significance of ! and " Test (Hypotheses 9a–13c)a
–.31*
.27*
–.16
Perceived organization
–.30*
.04
.18†
–.25*
–.02
Expected image
Supervisor relationship risks
quality .25*
.18*
–.20*
–.38* Expected image
Innovativeness as a job gains
.46*
requirement
.40* Individual
.14 innovative
Reputation as behavior
innovative .34*
Expected positive .25*
.13 .24* performance
outcomes .17 †
Dissatisfaction with
.20* Intrinsic interest
the status quo
–.09
.32*
a 2 2
Fit: # (27, n ! 216) ! 56.01, p " .01, # /df ! 2.07; RMSEA ! .07, SRMR ! .03, CFI ! .97, NFI ! .95, IFI ! .97. Parameter estimates
are from the completely standardized solution.
†
p " .10
* p " .05
TABLE 2
Results of the Product of Coefficients Test on Indirect Effects Mediated through Performance and Image
Outcome Expectationsa
Mediated Indirect
Distal Variable Mediator Effectb P
Perceived organization support for innovation Expected positive performance outcomes .01 0.91
Expected image risks .08*c 7.43
Expected image gains #.03* #3.99
Supervisor relationship quality Expected positive performance outcomes .05*c 3.96

Expected image risks .00 0.44
Expected image gains !.05*c #5.99
Innovativeness as a job requirement Expected positive performance outcomes .10*c 7.26

Expected image risks .10*c 8.20
Reputation as innovative Expected positive performance outcomes .09*c 6.77

Expected image risks #.04* #3.29
Dissatisfaction with the status quo Expected positive performance outcomes .05*c 4.20
Expected image risks #.03* #3.20
a
The MacKinnon et al. (1998) distribution of products P ! z!z" method was used to test the significance of mediated or indirect effects
as recommended by MacKinnon and colleagues (2002); z! equals the path coefficient for path ! divided by its standard error; z" equals the
path coefficient for path " divided by its standard error. The distribution of P follows the distribution of the product of two normal random
variables from Craig (1936). The critical value is 2.18 for the .05 significance level.
b
The !" product.
c
Mediation effects found to be significant by both the joint significance of ! and " test (see Figure 3) and the product of coefficients test.
* p " .05
ciency-oriented and the social-political perspectives) cated that (1) expected image gains had a near-zero
to examine these major outcome expectations, and correlation with innovative behavior (r ! .05, n.s.); (2)
the results of our study suggest that using a single it correlated positively with expected positive perfor-
perspective to understand the psychological consid- mance outcomes (r ! .45, p " .01); and (3) the inclu-
erations behind innovative behavior does not capture sion of expected image gains enhanced the path co-
the full picture. Providing direct support for the effi- efficient of expected positive performance (b ! .33)
ciency-oriented perspective, we found that employ- beyond its bivariate correlation with innovative be-
ees were more innovative when they anticipated such havior (r ! .29). This result pattern is consistent with
behavior would benefit their work. At the same time, classic suppression as discussed by various scholars
the concern for potential image risks and unfavorable (e.g., Cohen & Cohen, 1983; Horst, 1941; Maassen &
social impressions negatively affected innovative- Bakker, 2001; Pedhazur, 1997). As a suppressor, ex-
ness, suggesting a significant impact of social-politi- pected image gains had no significant relationship
cal considerations on employee innovation. with innovative behavior but shared some variance
Contrary to expectations, we found a significant yet with expected positive performance outcomes that
negative effect of expected image gains on innovative was irrelevant to innovative behavior. Expected im-
behavior. One possible explanation is that employees age gains may share some common variance with
who intended to use innovative behavior for the pur- expected positive performance outcomes because an
poses of showing off or pleasing other people were employee is more likely to achieve image gains when
indeed less innovative and were perceived and rated his or her innovative behavior brings performance
by their supervisors as such. This possible explana- gains. “Partialing out” this image-related reason to
tion would suggest that managers are able to see the improve performance purified and enhanced the re-
political motive and tend to give negative evaluations lationship between a true efficiency-oriented consid-
to employees who manipulate “innovativeness” to eration and innovative behavior. Under this situation,
pursue personal agendas. Another possibility is that the effect of the suppressor variable (expected image
expected image gains were not related to innovative gains) could become negative, even though it was not
behavior but served as a suppressor variable for ex- significantly related to the criterion variable (innova-
pected positive performance gains. Our results indi- tive behavior) (Horst, 1941; Pedhazur, 1997).
By examining performance and image outcome helpful for their work. Moreover, even when there is
expectations as intermediate psychological pro- a good idea that will benefit their work, these employ-
cesses, our findings also help to explain how ees may hesitate to give it a try because other people
contextual and individual difference antecedents may consider their ideas invalid and/or they simply
influence individual innovation indirectly by shap- do not want to rock the boat.
ing these efficiency-oriented and social-political As we hypothesized, employees who enjoyed a
considerations. Scott and Bruce (1994) suggested reputation of being innovative tended to internalize
organization support for innovation affects innova- the belief that innovative behaviors would benefit
tive behavior by signaling potential outcomes of their work, which motivated them to innovate
such behavior. Our findings support this argument. more. An innovative reputation, however, did not
When an organization supported innovation and significantly reduce employees’ concern about image
tolerated differences, employees felt psychologi- risk. It is possible that reputation has two contradic-
cally safe and had much less concern for image tory effects that cancelled each other out in this study.
risks. They might also have been more likely to On the one hand, as hypothesized, when innovative
perceive innovativeness as a desirable social image people do innovative things, they will not surprise
to achieve. Perceived organization support for in- others. Doing so fits their social image and results in
novation, however, did not affect their expectation less concern about image risk. Yet at the same time,
that being innovative would benefit their work. It is employees with an innovative reputation may also be
possible that expectation of positive performance in the risky position of being seen as troublemakers
outcomes is a more specific judgment than a gen- who always want to turn things around. Their inno-
eral proinnovation attitude. Whether innovative at- vative attempts may actually be less tolerated by co-
tempts will benefit a particular individual’s work is workers than the attempts of those lacking a reputa-
also influenced by other factors, such as the spe- tion for innovativeness and may instead be more
cific nature of her or his job. likely to irritate others.
As we hypothesized, employees who had good Employees who were less satisfied with the cur-
relationships with their supervisors were more rent performance of their department and organiza-
confident that their innovative behavior would re- tion had higher appreciation for the potential perfor-
sult in performance gains. Also, high-quality rela- mance and image benefits brought by new ideas,
tionships with their supervisors increased employ- technologies, and processes. This sense of dissatisfac-
ees’ confidence that their innovative attempts tion, though, did not reduce their concerns about
would receive favorable social evaluations and potential image risks associated with innovative be-
therefore enhance their images in their organiza- havior. This finding was contrary to our expectation
tion. Contrary to expectations, mutually trusting that a less effective situation could justify innovation
and respectful relationships with supervisors were and make innovative behaviors more legitimate. One
not related to reduced concerns for image risk. One possible explanation is that our construct of dissatis-
possible explanation is that image risk concerns a faction with the status quo is more a subjective per-
larger social context that includes not just an em- ception than a social consensus. Personal dissatisfac-
ployee’s supervisor but also others, such as work tion with the status quo does not necessarily relieve
group members and coworkers. the concern that other people (who might be satisfied
Previous studies have tended to explain the effects with the current condition) will frown upon the focal
of job characteristics on employee innovation and person’s innovative attempts.
creativity on the basis of the different capabilities of Taken together, our findings contribute to the
job incumbents and the different levels of freedom innovation literature in three major ways. First, our
available to them (e.g., Munton & West, 1995), or on results suggest the importance of expected perfor-
the basis of these factors’ implications for intrinsic mance and image outcomes in explaining why em-
work motivation (e.g., Oldham & Cummings, 1996). ployees innovate in the workplace, a critical ques-
Findings from this study suggest that the perceived tion that has been largely ignored in the literature.
nature of an employee’s job may also affect the em- In keeping with Janssen and colleagues’ (2004)
ployee’s innovativeness by influencing his or her ex- view of workplace innovation as a costly and risky
pectations for potential performance and image out- endeavor, the significant effects of outcome expec-
comes. Certain social stigmas are associated with tations on innovative behavior found in this study
different jobs. Our results suggest that employees suggest that these potential costs and benefits are
working in positions in which innovativeness is not important extrinsic motivational considerations
required (e.g., non-R&D-related jobs) may be less mo- that need to be assessed in addition to employees’
tivated to apply new ideas for two reasons. For one intrinsic interest. Second, our study examined how
thing, they do not consider new ideas or processes as outcome expectations serve as the psychological
mechanism to explain how and why organization- context (e.g., supervisor-subordinate relationships)
al, supervisor relational, job, social, and individual would seem to play an important role here. Also, it is
factors affect individual innovation. Our results important to provide positive social recognition for
add to the innovation literature by revealing how innovative employees and increase the extent of em-
the effects of these contextual and individual ante- ployees’ self-views as innovative. Companies with
cedents can be understood in light of their impacts histories of successful performance need to take steps
on these efficiency-oriented and social-political to break psychological comfort with the status quo
considerations. Third, an important theoretical and sensitize employees to opportunities for further
contribution of this research is the articulation of a improvement.
view of individual innovation that does not rely on a
single principle or kernel of truth as its core. This
Limitations and Suggestions for Future Research
study answers calls in the literature for utilizing mul-
tiple perspectives to study innovation (Abrahamson, The contributions discussed above should be inter-
1991; Wolfe, 1994). Our results suggest that both ef- preted in light of this study’s limitations. First, the
ficiency-oriented and social-political processes si- links in our theoretical model follow the hypothe-
multaneously influence individual innovation and sized causal order, with contextual and individual
that both perspectives need to be applied to under- difference factors affecting outcome expectations,
stand the dynamics of workplace innovation. In par- which lead to innovative behavior. The cross-sec-
ticular, the findings of this study suggest the need to tional research design we used, however, limited our
go beyond the dominant efficiency-oriented perspec- ability to determine causality. It is possible, for exam-
tive and to assess both the technical and symbolic ple, that the relationships between some distal vari-
functions of innovative behavior. ables and innovative behavior are reciprocal. For in-
stance, reputation and dissatisfaction affect an
employee’s innovative behavior as theorized in our
Practical Implications
model. Yet at the same time, a person’s innovative-
One major reason employees do not innovate is ness might also influence her or his reputation and
their fear of being perceived negatively by others. dissatisfaction level in the future, which in turn will
Findings from this study suggest organization sup- affect future outcome expectations and innovative
port and job requirements as two areas to focus on to behavior from this employee. We strongly recom-
reduce the image risks associated with innovative mend future studies that use longitudinal designs to
behavior. Although the importance of building a cul- explore the relationships posited in our model and
ture supportive of innovation (e.g., by establishing these possible reciprocal relationships.
special rewards for innovation and establishing fo- To alleviate potential common methods bias, we
rums for diverse ideas) is widely accepted, the rele- collected data on innovative behavior from super-
vance of job requirements has been less emphasized. visors, as an alternative to employee self-reports.
Most previous innovation studies have focused on Despite the appropriateness of measuring percep-
R&D departments, where innovative behaviors are tual and attitudinal variables using self-reports, it is
part of employees’ job descriptions. For employees a limitation that our distal variables and outcome
whose jobs are not, by definition, technology or in- expectation variables were collected from the same
novation related, their company’s mission of “inno- source using the same method (although how much
vation” could appear rather remote or irrelevant, pre- of a limitation is open to debate [cf. Spector, 2006]),
venting them from contributing valuable ideas. It is which may have biased the relationships observed
therefore important for managers to break job position among these variables. To address this potential
stereotypes and to demystify innovation. Communi- concern, we undertook the procedural remedies
cating with those employees to let them know that (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003) of
they too are expected to contribute new ideas is one protecting respondent confidentiality and reducing
way. Explicitly incorporating innovativeness into item ambiguity (by piloting the survey with MBA
their job descriptions is another possibility. students and discussing the questionnaire with
Another reason why employees do not innovate is managers before administering it to employees).
that they don’t believe doing so will benefit their Statistically, we investigated the potential impact
work. Our results suggest four areas management can of common method variance by controlling for an
amend to establish a strong association between in- unmeasured latent method factor and allowing all
novative behavior and job performance: employee- self-reported items to load both on their theoretical
supervisor relationships, job requirements, employee constructs and on the method factor (Podsakoff et
social reputation, and employee dissatisfaction with al., 2003). The results for all structural path param-
the status quo. In addition to job requirements, social eters remained the same after our controlling for
the method factor, suggesting that common method examine whether the effects of performance and im-
variance did not bias our findings. age outcome expectations also vary across cultures.
Using supervisor ratings to measure employee This study is the first attempt to directly theorize
innovative behavior also has its limitations. First, and test major outcome expectations associated with
in some situations one supervisor provided ratings employee innovative behavior. Drawing from both
for multiple subordinates, raising the issue that the efficiency-oriented and social-political perspec-
nonindependence may have reduced statistical tives on innovation, the model tested here provides a
power in hypothesis testing. Overall, our analysis theoretical framework for understanding why em-
suggested that nonindependence did not signifi- ployees engage in innovative behavior in the work-
cantly affect our results. However, in further studies place and how different contextual and individual
researchers should strive to avoid the nonindepen- difference factors affect employee innovation indi-
dence issue when measuring individual-level vari- rectly by shaping these intermediate psychological
ables to avoid its potential negative impact on statis- considerations. We hope that this study will stimu-
tical power. Supervisor ratings can also be biased by late more theory building and testing to investigate
the individual response characteristics of different the processes leading to individual innovation.
supervisors (cf. Landy & Farr, 1980; Mount, Judge,
Scullen, Sytsma, & Hezlett, 1998). Future studies
should include attempts to collect a second measure REFERENCES
of innovative behavior (e.g., a peer rating or objective Abrahamson, E. 1991. Managerial fads and fashions: The
indexes) to cross-validate supervisor ratings. diffusion and rejection of innovations. Academy of
The symbolic function of innovative behavior is Management Review, 16: 586 – 612.
relatively underaddressed in the existing literature. Aiken, M., Bacharach, S. B., & French, J. L. 1980. Orga-
More future research needs to be done to examine nizational structure, work process, and proposal
the role of image outcome expectations in the inno- making in administrative bureaucracies. Academy
vation process. For example, the negative relation- of Management Journal, 23: 631– 652.
ship between expected image gains and innovative Allen, D. G., Mahto, R. V., & Otondo, R. F. 2007. Web-
behavior found in this study is both surprising and based recruitment: Effects of information, organiza-
interesting. Future research needs to examine the vi- tional brand, and attitudes toward a web site on
ability of different interpretations. Does an image- applicant attraction. Journal of Applied Psychol-
enhancing motive have a significant impact on inno- ogy, 92: 1696 –1708.
vative behavior, in addition to performance and Amabile, T. M. 1988. A model of creativity and innova-
image risk considerations? Are supervisors truly able tion in organizations. In B. M. Staw & L. L. Cum-
to see a selfish motive behind employee innovative mings (Eds.), Research in organizational behavior,
behaviors? Is it possible that employees can engage in vol. 10: 123–167. Greenwich, CT: JAI Press.
artificial innovative acts to enhance image only in Amabile, T. M. 1996. Creativity in context: Update to
certain situations, such as when the results of the “The social psychology of creativity.” Boulder, CO:
innovation are difficult to objectively evaluate? Westview Press.
Future studies are also recommended to extend Anderson, N., De Dreu, C. K. W., & Nijstad, B. A. 2004.
our research by examining the impact of perfor- The routinization of innovation research: A con-
mance and image outcome expectations in different structively critical review of the state-of-the-science.
situations and contexts. For example, individual Journal of Organizational Behavior, 25: 147–173.
innovation can be viewed in terms of different Anderson, J. C., & Gerbing, D. W. 1988. Structural equa-
stages, such as idea generation, promotion, and im- tion modeling in practice: A review and recom-
plementation (cf. Kanter, 1988); different levels, mended two-step approach. Psychological Bulletin,
including incremental versus radical; and different 103: 411– 423.
types, such as technical versus administrative Arkin, R. M. 1981. Self-presentation styles. In J. T. Tede-
(Daft, 1978; Van de Ven, 1986). Specific anteced- schi (Ed.), Impression management theory and so-
ents of these subcategories of innovative behavior cial psychological research: 311–333. New York:
can be assessed. It is possible that the effects of Academic Press.
performance and image outcome expectations Arndt, M., & Bigelow, B. 2000. Presenting structural in-
might be more important for certain types, levels, or novation in an institutional environment: Hospitals’
stages of innovative behavior. Also, several recent use of impression management. Administrative Sci-
reviews have suggested the necessity of taking a ence Quarterly, 45: 494 –522.
cross-cultural perspective for understanding innova- Ashford, S. J. 1986. Feedback seeking in individual ad-
tion and creativity (Anderson, De Dreu, & Nijstad, aptation: A resource perspective. Academy of Man-
2004; Shalley et al., 2004). Future studies need to agement Journal, 29: 465– 487.
Ashford, S. J., & Northcraft, G. B. 1992. Conveying more (or 25 years: Applying multi-level multi-domain per-
less) than we realize: The role of impression-manage- spective. Leadership Quarterly, 6: 219 –247.
ment in feedback-seeking. Organizational Behavior Green, S. G., & Mitchell, T. R. 1979. Attributional process of
and Human Decision Processes, 53: 310 –334. leaders in leader-member interactions. Organizational
Ashford, S. J., Rothbard, N. P., Piderit, S. K., & Dutton, Behavior and Human Performance, 23: 429 – 458.
J. E. 1998. Out on a limb: The role of context and Harrison, J. R., & Carroll, G. R. 1991. Keeping the faith: A
impression management in selling gender-equity is- model of cultural transmission in formal organizations.
sues. Administrative Science Quarterly, 43: 23–57. Administrative Science Quarterly, 36: 552–582.
Bliese, P. D., & Hanges, P. J. 2004. Being both too liberal Horst, P. 1941. The role of prediction variables which are
and too conservative: The perils of treating grouped independent of the criterion. In P. Horst (Ed.), The
data as though they were independent. Organiza- prediction of personal adjustment. Social Research
tional Research Methods, 7: 400 – 417. Bulletin, 48: 431– 436.
Bunce, D., & West, M. A. 1995. Self perceptions and House, R. J., & Dessler, G. 1974. The path goal theory of
perceptions of group climate as predictors of indi- leadership: Some post hoc and a priori tests. In J.
vidual innovation at work. Applied Psychology: An Hunt & L. Larson (Eds.), Contingency approaches to
International Review, 44: 199 –215. leadership: 29 –55. Carbondale: Southern Illinois
Chatman, J. A. 1991. Matching people and organizations: University Press.
Selection and socialization in public accounting firms. James, L., Hartman, E., Stebbins, M., & Jones, A. 1977. An
Administrative Science Quarterly, 36: 459 – 484. examination of the relationship between psycholog-
Cohen, J., & Cohen, P. 1983. Applied multiple regres- ical climate and a VIE model for work motivation.
sion/correlation analysis for the behavioral sci- Personnel Psychology, 30: 229 –254.
ences (2nd ed.). Hillsdale, NJ: Erlbaum. James, L., Mulaik, S., & Brett, J. 1982. Causal analysis:
Craig, C. C. 1936. On the frequency function of xy. An- Assumptions, models, and data. Beverly Hills, CA:
nals of Mathematical Statistics, 7: 1–15. Sage.
Daft, R. L. 1978. A dual-core model of organizational Janssen, O. 2000. Job demands, perceptions of effort-
innovation. Academy of Management Journal, 21: reward fairness and innovation work behavior. Jour-
193–210. nal of Occupational and Organizational Psychol-
ogy, 73: 287–302.
Dean, J. W., Jr. 1987. Building the future: The justification
process for new technology. In J. M. Pennings & A. Janssen, O. 2003. Innovative behavior and job involve-
Buitendam (Eds.), New technology as organizational ment at the price of conflict and less satisfactory
innovation: The development and diffusion of micro- relations with co-workers. Journal of Organiza-
electronics: 35–58. Cambridge, MA: Ballinger. tional and Occupational Psychology, 76: 347–364.
Dyer, W. G., & Page, R. A., Jr. 1988. The politics of Janssen, O., Van de Vliert, E., & West, M. 2004. The bright
innovation. Knowledge in the Society: An Interna- and dark sides of individual and group innovation:
tional Journal of Knowledge Transfer, 1(2): 23– 41. A special issue introduction. Journal of Organiza-
tional Behavior, 25: 129 –145.
Farr, J. L., & Ford, C. M. 1990. Individual innovation. In
M. A. West & J. L. Farr (Eds.), Innovation and cre- Janssen, O., & Van Yperen, N. W. 2004. Employees’ goal
ativity at work: Psychological and organizational orientations, the quality of leader-member exchange,
strategies: 63– 80. Chichester, U.K.: Wiley. and the outcomes of job performance and job satis-
faction. Academy of Management Journal, 47:
Gilbert, D. T. 1998. Ordinary personology. In D. T. Gil- 368 –384.
bert, S. T. Fiske, & G. Lindzey (Eds.), The handbook
of social psychology (4th ed.) vol. 2: 89 –150. Bos- Jöreskog, K. G., & Sörbom, D. 1996a. LISREL 8: User’s
ton: McGraw-Hill. reference guide. Lincolnwood, IL: Scientific Soft-
ware International.
Graen, G. B. 1976. Role making processes within com-
plex organizations. In M. D. Dunnette (Ed.), Hand- Jöreskog, K. G., & Sörbom, D. 1996b. LISREL 8: Structural
book of industrial and organizational psychology: equation modeling with the SIMPLIS command lan-
1201–1245. Chicago: Rand-McNally. guage. Lincolnwood, IL: Scientific Software Interna-
tional.
Graen, G. B., Novak, M. A., & Sommerkamp, P. 1982. The
effects of leader-member exchange and job design on Judge, T. A., & Ferris, G. R. 1993. Social context of per-
productivity and satisfaction: Testing a dual attach- formance evaluation decisions. Academy of Man-
ment model. Organizational Behavior and Human agement Journal, 36: 80 –105.
Performance, 30: 109 –131. Kanter, R. M. 1983. The change masters. New York:
Graen, G. B., & Uhl-Bien, M. 1995. Relationship-based Simon & Schuster.
approach to leadership: Development of leader- Kanter, R. M. 1988. When a thousand flowers bloom:
member exchange (LMX) theory of leadership over Structural, collective, and social conditions for in-
novation in organization. In B. M. Staw & L. L. Cum- izational effectiveness and efficiency. Academy of
mings (Eds.), Research in organizational behavior, Management Journal, 36: 1345–1361.
vol. 10: 169 –211. Greenwich, CT: JAI Press. Pedhazur, E. J. 1997. Multiple regression in behavioral
Kenny, D. 1979. Correlation and causality. New York: research: Explanation and prediction. Fort Worth,
Wiley-Interscience. TX: Harcourt.
Kimberly, J. R. 1981. Managerial innovation. In P. C. Pennings, P. S., & Goodman, J. M. 1977. Toward a work-
Nystrom & W. H. Starbuck (Eds.), Handbook of or- able framework. In P. S. Pennings & J. M. Goodman
ganizational design, vol. 1: 84 –104. New York: Ox- (Eds.), New perspectives on organizational effec-
ford University Press. tiveness: 146 –184. San Francisco: Jossey-Bass.
Kimberly, J. R., & Evanisko. M. 1981. Organizational in- Podsakoff, P. M., MacKenzie, S. B., & Lee, J., & Podsakoff,
novation: The influence of individual, organization- N. P. 2003. Common method biases in behavioral
al, and contextual factors on hospital adoption of research: A critical review of the literature and rec-
technological and administrative innovations. Acad- ommended remedies. Journal of Applied Psychol-
emy of Management Journal, 24: 689 –713. ogy, 88: 879 –903.
Landy, F. L., & Farr, J. 1980. Performance rating. Psycho- Regan, D. T., & Totten, J. 1975. Empathy and attribution:
logical Bulletin, 87: 72–102. Turning observers into actors. Journal of Personal-
Lant, T. K., & Mezias, S. J. 1992. An organizational learn- ity and Social Psychology, 32: 850 – 856.
ing model of convergence and reorientation. Orga- Rioux, S. M., & Penner, L. A. 2001. The causes of organiza-
nization Science, 3: 47–71. tional citizenship behavior: A motivational analysis.
Leary, M. R., & Kowalski, R. M. 1990. Impression man- Journal of Applied Psychology, 86: 1306 –1314.
agement: A literature review and two-component Rogers, E. M. 1983. Diffusion of innovation (2nd ed.).
model. Psychological Bulletin, 107: 34 – 47. New York: Free Press.
Maassen, G. H., & Bakker, A. B. 2001. Suppressor variables Schlenker, B. R. 1980. Impression management: The
in path models: Definitions and interpretations. Socio- self-concept, social identity, and interpersonal re-
logical Methods & Research, 30: 241–270. lations. Monterey, CA: Brooks/Cole.
MacKinnon, D. P., Lockwood, C., & Hoffman, J. 1998. A Scott, S. G., & Bruce, R. A. 1994. Determinants of inno-
new method to test for mediation. Paper presented vative behavior: A path model of individual innova-
at the annual meeting of the Society for Prevention tion in the workplace. Academy of Management
Research, Park City, UT. Journal, 37: 580 – 607.
MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., Shalley, C. E., Zhou, J., & Oldham, G. R. 2004. The effects
West, S. G., & Sheets, V. 2002. A comparison of of personal and contextual characteristics on creativ-
methods to test mediation and other intervening ity: Where should we go from here? Journal of Man-
variable effects. Psychological Methods, 7: 83–104. agement, 30: 933–958.
Markus, H., & Zajonc, R. B. 1985. The cognitive perspec- Shrauger, J. S., & Shoeneman, T. J. 1979. Symbolic inter-
tive in social psychology. In G. Lindzey & E. Aronson actionist view of self-concept: Through the looking
(Eds.), The handbook of social psychology (3rd ed.), glass darkly. Psychological Bulletin, 86: 549 –573.
vol. 1: 137–230. New York: Random House. Siegel, S. M., & Kaemmerer, W. F. 1978. Measuring the
Mount, M. K., Judge, T. A., Scullen, S. E., Sytsma, M. R., perceived support for innovation in organizations.
& Hezlett, S. A. 1998. Trait, rater and level effects in Journal of Applied Psychology, 63: 553–562.
360-degree performance ratings. Personnel Psychol- Sörbom, D., & Jöreskog, K. G. 1982. The use of structural
ogy, 51: 557–576. equation models in evaluation research. In C. Fornell
Mumford, M., & Gustafson, S. 1988. Creativity syndrome: (Ed.), A second generation of multivariate analy-
Integration, application, and innovation. Psycholog- sis, vol. 2: 381– 418. New York: Praeger.
ical Bulletin, 130: 27– 43. Spector, P. E. 2006. Method variance in organizational
Munton, A. G., & West, M. A. 1995. Innovations and per- research: Truth or urban legend? Organizational Re-
sonal change: Patterns of adjustment to relocation. search Methods, 9: 221–232.
Journal of Organizational Behavior, 16: 363–375. Steele, C. M. 1988. The psychology of self-affirmation:
Netemeyer, R., Johnston, M., & Burton, S. 1990. Analysis Sustaining the integrity of self. In L. Berkowitz (Ed.),
of role conflict and role ambiguity in a structural Advances in experimental social psychology, vol.
modeling framework. Journal of Applied Psychol- 21: 261–302. New York: Academic Press.
ogy, 75: 148 –157. Sutton, R. I., & Hargadon, A. 1996. Brainstorming groups
Oldham, G. R., & Cummings, A. 1996. Employee creativ- in context: Effectiveness in a product design firm.
ity: Personal and contextual factors at work. Acad- Administrative Science Quarterly, 41: 685–718.
emy of Management Journal, 39: 607– 634. Tedeschi, J. T., & Riess, M. 1981. Identities, the phenomenal
Ostroff, C., & Schmitt, N. 1993. Configurations of organ- self, and laboratory research. In J. T. Tedeschi (Ed.),
Impression management theory and social psycho- Wolfe, R., Lennox, R., & Cutler, B. L. 1986. Getting along
logical research: 3–22. New York: Academic Press. and getting ahead: Empirical support for a theory of
Tetlock, P. E., & Manstead, A. S. R. 1985. Impression protective and acquisitive self-presentation. Journal
management versus intrapsychic explanations in so- of Personality and Social Psychology, 50: 356 –361.
cial psychology: A useful dichotomy? Psychological Woodman, R. W., Sawyer, J. E., & Griffin, R. W. 1993.
Review, 92: 59 –77. Toward a theory of organizational creativity. Acad-
emy of Management Review, 18: 293–321.
Tierney, P., Farmer, S. M., & Graen, G. B. 1999. An
examination of leadership and employee creativity: Zelditch, M., Jr. 2001. Theories of legitimacy. In J. T. Jost
The relevance of traits and relationships. Personnel & B. Major (Eds.), The psychology of legitimacy:
Psychology, 52: 591– 620. Emerging perspectives on ideology, justice, and
intergroup relations: 33–53. Cambridge, U.K.: Cam-
Tolber, P. S., & Zucker, L. 1983. Institutional sources of
bridge University Press.
change in the formal structure of organizations: The
diffusion of civil service reform, 1880 –1935. Admin- Zhou, J. 2003. When the presence of creative coworkers is
istrative Science Quarterly, 28: 22–39. related to creativity: Role of supervisor close monitor-
ing, developmental feedback, and creative personality.
Tushman M. L., & Romanelli, E. 1985. Organizational
Journal of Applied Psychology, 88: 413– 422.
evolution: A metamorphosis model of convergence
and reorientation. In L. L. Cummings & B. M. Staw Zhou, J., & Woodman, R. W. 2003. Manager’s recognition
(Eds.), Research in organizational behavior, vol. 7: of employees’ creative ideas: A social-cognitive
171–222. Greenwich, CT: JAI Press. model. In L. V. Shavinina (Ed.), The international
handbook on innovation: 631– 640. Oxford, U.K.:
Van de Ven, A. H. 1986. Central problems in the man- Elsevier Science.
agement of innovation. Management Science, 32:
590 – 607.
APPENDIX
Vroom, V. H. 1964. Work and motivation. New York:
Wiley. Scale Itemsa
Wayne, S. J., & Liden, R. C. 1995. Effects of impression Individual Innovative Behavior (Scott & Bruce, 1994;
management on performance ratings: A longitudinal ! ! .93)
study. Academy of Management Journal, 38: 232– Supervisor indicated how characteristic each of the
260. following behaviors was of a particular employee:
West, M. A. 1989. Innovation amongst health care pro- 1. Searches out new technologies, processes, techniques,
fessionals. Social Behavior, 4: 173–184. and/or product ideas. (.92)
West, M. A., & Farr, J. L. 1989. Innovation at work: Psycho- 2. Generates creative ideas. (.91)
logical perspectives. Social Behavior, 4: 15–30. 3. Promotes and champions ideas to others. (.83)
4. Investigates and secures funds needed to implement
West, M. A., & Farr, J. L. (Eds.). 1990a. Innovation and new ideas. (.67)
creativity at work: Psychological and organization- 5. Develops adequate plans and schedules for the imple-
al strategies. Chichester, U.K.: Wiley. mentation of new ideas. (.72)
West, M. A., & Farr, J. L. 1990b. Innovation at work. In 6. Is innovative. (.88)
M. A. West & J. L. Farr (Eds.), Innovation and cre-
Expected Positive Performance Outcomes (adapted from
ativity at work: Psychological and organizational
House & Dessler, 1974; ! ! .77)
strategies: 3–13. Chichester, U.K.: Wiley.
1. The more innovative I am, the better my job perfor-
West, M. A., & Wallace, M. 1991. Innovation in health
mance. (.80)
care teams. European Journal of Social Psychology,
2. Coming up with creative ideas helps me do well on
21: 303–315.
my job. (.81)
Westphal, J. D., Gulati, R., & Shortell, S. M. 1997. Cus- 3. My work unit will perform better if I often suggest
tomization or conformity? An institutional and net- new ways to achieve objectives. (.59)
work perspective on the content and consequences
of TQM adoption. Administrative Science Quar- Expected Image Risks (adapted from Ashford, 1986; ! !
terly, 42: 366 –394. .77)
Williams, L., & Hazer, J. 1986. Antecedents and conse- 1. My co-workers will think worse of me if I often try out
quences of satisfaction and commitment in turnover new approaches on my job. (.78)
models: A reanalysis using latent variable structural 2. People will think I am crazy if I come up with new
equation methods. Journal of Applied Psychology, ways of doing my job. (.68)
71: 219 –231. 3. Other people will think worse of me if I try to change
the way things operate within the organization. (.75)
Wolfe, R. A. 1994. Organizational innovation: Review,
critique and suggested research directions. Journal Expected Image Gains (Adapted from Ashford et al.,
of Management Studies, 31: 405– 431. 1998; ! ! .86)
1. If I were to do something innovative, my image in the 7. How would you characterize your working relation-
organization would be enhanced. (.85) ship with your supervisor? (.78)
2. Searching out new technologies or techniques for the
Innovativeness as a Job Requirement (developed for this
organization will make me look good. (.80)
study; ! ! .85)
3. Participating in the implementation of new ideas will
improve my images in the organization. (.81) 1. My job duties include searching for new technologies
4. Suggesting new ways to achieve goals will improve and techniques. (.67)
my supervisor’s evaluation of me. (.68) 2. Introducing new ideas into the organization is part of
my job. (.82)
Perceived Organization Support for Innovation (Scott & 3. I don’t have to be innovative to fulfill my job require-
Bruce, 1994; ! ! .92) ments. (reverse-coded) (.66)
Standardized factor loadings for the two indicators of 4. My job requires me to try out new approaches to
perceived organization support for innovation were sup- problems. (.73)
port for creativity, .83; tolerance of difference, .88. 5. Suggesting new ideas is part of my job duties. (.80)
1. Creativity is encouraged here. Reputation as Innovative (developed for this study; ! !
2. Our ability to function creatively is respected by the .78)
leadership.
3. Around here, people are allowed to try to solve the 1. People come to me when they want new ideas. (.81)
same problems in different ways. 2. Others in the organization often expect me to contrib-
4. The main function of members in this organization is ute innovative ideas. (.79)
to follow orders which come down through chan- Dissatisfaction with the Status Quo (developed for this
nels. (reverse-coded) study; ! ! .75)
5. Around here, a person can get in a lot of trouble being
different. (reverse-coded) 1. Many things in my department need improvement.
6. This organization can be described as flexible and (.65)
continually adapting to change. 2. The performance of my organization needs to be im-
7. A person can’t do things that are too different around proved. (.85)
here without provoking anger. (reverse-coded) 3. The performance of my work unit needs to be im-
8. The best way to get along in this organization is to think proved. (.66)
the way the rest of the group does. (reverse-coded) Intrinsic Interest (Tierney et al., 1999; ! ! .85)
9. People around here are expected to deal with prob-
lems in the same way. (reverse-coded) 1. I enjoy finding solutions to complex problems. (.75)
10. This organization is open and responsive to change. 2. I enjoy coming up with new ideas for products. (.75)
11. The people in charge around here usually get credit 3. I enjoy engaging in analytical thinking. (.76)
for others’ ideas. (reverse-coded) 4. I enjoy creating new procedures for work tasks. (.64)
12. In this organization, we tend to stick to tried and true 5. I enjoy improving existing processes or products. (.80)
ways. (reverse-coded) a
Standardized factor loadings from the measurement
13. This place seems to be more concerned with the model are listed in parentheses after the relevant item.
status quo than with changes. (reverse-coded) All factor loadings were significant at .05.
Leader-Member Exchange (Graen et al., 1982; ! ! .90)
1. Do you know where you stand with your super-
visor . . . do you usually know how satisfied your
Feirong Yuan (fyuan@ku.edu) is an assistant professor of
supervisor is with what you do? (.64)
management at the School of Business, The University of
2. How well does your supervisor understand your job
Kansas. She received her Ph.D. in management from
problems and needs? (.77)
Texas A&M University. Her research interests include
3. How well does your supervisor recognize your poten-
employee creativity and innovation, organizational
tial? (.73)
change, and cross-cultural studies.
4. Regardless of how much formal authority he/she has
built into his/her position, what are the chances that Richard W. Woodman (dwoodman@mays.tamu.edu) is
your supervisor would use his/her power to help you the Fouraker Professor of Business and Professor of Man-
solve problems in your work? (.82) agement at Texas A&M University. He received his Ph.D.
5. Again, regardless of the amount of formal authority from Purdue University. Dr. Woodman is currently the
your supervisor has, what are the chances that he/she editor of the Journal of Applied Behavioral Science. His
would “bail you out,” at his/her expense? (.74) research and writing focuses on organizational change
6. I have enough confidence in my supervisor that I and organizational creativity.
would defend and justify his/her decision if he/she
were not present to do so. (.77)
Copyright of Academy of Management Journal is the property of Academy of Management and its content may
not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.

Yuan & Woodman (2010)

Uploaded by

Copyright:

Available Formats

Yuan & Woodman (2010)

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

Yuan & Woodman (2010)

Uploaded by

Copyright:

Available Formats

!

Academy of Management Journal

INNOVATIVE BEHAVIOR IN THE WORKPLACE: THE ROLE OF

Why do employees engage in innovative behavior at their workplaces? We examine

Perceived organization The social-political

Reputation as Expected positive

Dissatisfaction with the

Hypotheses 12a and 12b. The relationship be-

Variable Mean s.d. 1 2 3 4 5 6 7 8 9 10 11 12

1. Individual innovative 3.17 0.93 (.93)

Supervisor relationship quality Expected positive performance outcomes .05*c 3.96

Innovativeness as a job requirement Expected positive performance outcomes .10*c 7.26

Reputation as innovative Expected positive performance outcomes .09*c 6.77

You might also like