|
Article Excerpt
Recently, organizational scholars have identified creativity as a core employee capacity in organizations (Amabile, Conti, Coon, Lazenby, & Herron, 1996; Oldham & Cummings, 1996; Zhou & George, 2001). Most studies of creativity in organizational settings have attended to contextual factors that either facilitate (challenging task, supportive supervision; Oldham & Cummings, 1996) or inhibit (rigid procedure, lack of autonomy; Shalley, Gilson, & Blum, 2000) individual creativity. The present study investigates the role of group composition as a contextual predictor of creative behaviour of employees in a group setting (Kurtzberg & Amabile, 2001).
In response to the increasing importance of the issue of diversity in organizations, scholars have examined the implications of membership heterogeneity with respect to creativity and innovation. Most studies in this line of research, however, have examined these relationships at the group level of analysis, attending to issues such as the demographic diversity of a group and its creative performance (e.g. Ancona & Caldwell, 1992; Jehn, Northcraft, & Neale, 1999). For this reason, our understanding of how group composition affects individual creativity is quite limited. This gap is critical because, assuming that individual creativity is the basis of organizational innovations (Amabile, 1988), individual members' creative performance is a necessary predecessor of group creativity.
In a recent review, Jackson, Joshi, and Erhardt (2003, p. 803) framed the diversity issue as a phenomenon that involves dynamics at various levels including individual, dyadic relationships, teams/departments and the entire organization. They maintained that the organizational demography literature has increased our understanding of group- and organizational-level dynamics of diversity, while largely ignoring its role in shaping individual behaviour: 'feelings and behaviors of individuals can be shaped by the diversity of work teams, it is possible that dyadic interactions are shaped by both relational demography and team composition' (p. 819). In a recent meta-analytic review of 182 studies published between 1997 and 2004, Dionne, Randel, Jaussi, and Chun (2004) found that (a) only 19% of these published studies explicitly addressed the issue of level of analysis, (b) only 22% of empirical studies used proper measurement approaches 'by measuring concepts at the same level of analysis at which they were specified' (p. 202), and (c) none (0%) of the studies employed appropriate multi-level data analysis strategies (Dionne et al., 2004). Given that many studies have conceptualized diversity as a multi-level phenomenon, the observation that none of these studies utilized proper multi-level analysis strategies in dealing with apparently multi-level phenomena and variables is quite surprising (Harrison & Klein, in press).
Responding to the need for more micro-level and multi-level approaches to the issue of diversity, the present study investigates individual- and cross-level processes involving group composition variables in predicting employee creative behaviour. Specifically, at the individual level, I conceptualize membership heterogeneity in a group by utilizing the concept of relational demography, which is defined as the extent to which a particular member is different from other members within the same work unit (Tsui, Egan, & O'Reilly, 1992). I also explore the effects of various group diversity characteristics on employees' creative behaviour. This study examines a common set of demographic variables including age, gender, tenure and functional background, but it also investigates the effects of variables that have rarely been included in the existing studies (Jackson et al., 2003), such as hierarchical status and performance level of employees. This study thus significantly extends the organizational demography and creativity literature by revealing multi-level dynamics through which various demographic variables influence employee creative behaviour.
The multi-level hypotheses advanced in this study will be empirically tested using data collected from a division of a large Korean electronics company. Thus far, the organizational demography literature has been developed based on empirical data from Western countries, and cross-cultural validation of the diversity-related propositions in other cultural contexts has been very rare (see Pelled, Xin, & Weiss, 2001, for an exception). Given the collectivistic and hierarchical nature of Asian culture, including Korea (Hofstede, 2001), the effects of group composition variables on employee behaviour might be more pronounced and expressed in different directions in Asian countries. In addition to extending the literature with multi-level propositions, the present study also tests the cross-cultural generalizability of theoretical arguments of organizational demography in a new cultural context. Finally, drawing on a recent conceptual clarification by Harrison and Klein (in press), demographic diversity is defined as either separation in positions and values related to task-related issues or variety in members' categorical backgrounds, but not as disparity in terms of unequal distribution of resources or power. These 'separation'-type diversity variables were operationalized as standard deviations (instead of coefficients of variation, which are appropriate only for 'disparity'-type diversity) as recommended by Harrison and Klein.
Hypotheses
The current outcome variable, creative behaviour, is defined as 'a behavioral manifestation of individual creativity that may lead to the generation of novel and useful ideas' (see Choi, 2004). This definition focuses on the behavioural process of generating a new idea or product, regardless of how it is evaluated by others afterwards. For example, to be creative in the workplace, employees may need to be sensitive about inefficient work methods, procedures and/or policies and create more efficient and effective ways to achieve a goal (Frese, Fay, Hilburger, Leng, & Tag, 1997). Scholars have used various labels to refer to this type of behaviour, including personal initiative, innovative behaviour, voice and taking charge (Frese et al., 1997; LePine & Van Dyne, 2001; Morrison & Phelps, 1999; Scott & Bruce, 1994). This individual-level behavioural process of developing new ways of doing things may ultimately lead to creative outcomes for the individual (e.g. number of successful ideas) and the group (e.g. creativity of group decisions). This study proposes that this intermediate, behavioural outcome is shaped by group composition at multiple levels of analysis.
Relational demography
Organizational researchers have argued that demographic characteristics are a source of social context for individual behaviour (Zenger & Lawrence, 1989). At the individual level, employees compare their demographic characteristics with those of their colleagues in order to judge the extent to which they are different from or similar to others in the social unit. Based on theoretical concepts such as social categorization, social identity and the similarity-attraction paradigm, researchers have argued that relational demography (the difference or similarity of a person with respect to others) is a significant predictor of various work-related attitudes (Tsui et al., 1992). Most field studies of relational demography have reported that being different from others has negative implications in regard to individual attitudes and behaviour, such as group cohesion, commitment, communication and citizenship behaviour (Riordan & Shore, 1997; Tsui, Porter, & Egan, 2002; Zenger & Lawrence, 1989).
Owing to the lack of empirical studies, the implications of relational demography with respect to creative behaviour are not clear. In a simulation-based study using MBA students as subjects, relational demography was positively related to creativity of TQM solutions (Chatman, Polzer, Barsade, & Neale, 1998). In this study, however, the relational demography measure was a composite measure that combined dissimilarities in nationality, race and gender, which rendered the source of the positive effect ambiguous. Although being different from others may be a source of new ideas, given that creative behaviour is a social process (e.g. expressing or sharing ideas with others, introducing changes to the existing task processes of the work unit), the potential of relational demography to disrupt the social process might substantially override its benefit as a source of fresh ideas. For instance, Tsui et al. (2002) reported that demographic dissimilarity in gender, race and age decreased the frequency of employees' suggestions for improving the organization. When employees feel that they are surrounded by people with different backgrounds, it may be difficult for them to develop intimate relationships, be open-minded and feel free to suggest new ways of doing things (cf. evaluation apprehension; Paulus & Yang, 2000). Studies have shown that, under such conditions, individuals tend to be less creative (Amabile et al., 1996).
To explain interactive processes, organizational demography researchers have largely depended on a similar set of theories including social categorization or identity and similarity-attraction paradigm, all of which suggest that individuals with greater similarity in their social backgrounds or demographic categories tend to identify, trust and interact more freely with each other (Williams & O'Reilly, 1998). These theories predict that practically any type of dissimilarity in social attributes of a person in comparison with his/her colleagues will impede this person's social integration, which effectively decreases his/her willingness to share new ideas or introduce change in work procedures that might result in further isolation within the group (LePine & Van Dyne, 2001). Janssen's (2003) study reveals that innovative behaviour, although it is beneficial for the organization, tends to produce negative social outcomes for employees in that it degrades their social relationships with the colleagues who resist the suggested changes.
In the present study, however, I propose that not every social attribute disrupts creative process at the individual level. The negative implications of being different on a particular social characteristic may depend on the extent to which it creates power differentials among individuals. Diversity researchers have used power or status differential to account for the effect of diversity (Jackson et al., 2003). In fact, to interpret anomalies in empirical findings and to move beyond the rather simplified arguments based on similarity attraction, recent demography research has increasingly relied on power and status as a theoretical ground (e.g. Chatman & O'Reilly, 2004; Chattopadhyay, George, & Lawrence, 2004). The issue of power or status seems particularly relevant for understanding employees' creative behaviour if it is facilitated by psychological safety or no perceived threat of social repercussions for presenting different ideas (Anderson & West, 1998).
Scholars have proposed and reported confirming evidence that gender, race and age result in either high or low status and competence expectations of a particular individual (Chattopadhyay, 1999; Tsui et al., 2002). In the present context of a Korean organization in which...
|
