팀 내 개인간 인구통계학적 비유사성과 개인성과의 관계에 관한 연구

The Curvilinear Effect of Demographic Dissimilarity(Relational Demography) On Individual Performance

본 연구의 목적은 팀 내 개인이 다른 사람과 차이나는 정도 즉, 개인간 인구통계학적 비유사성(demographic dissimilarity) 혹은 관계적 인구통계(relational demography)가 개인의 성과에 미치는 영향을 알아보기 위한 것이다. 자기범주화 관점(self-categorization perspective)과 유사성/매력 관점(similarity/attraction perspective) 및 정보/의사결정관점(information/decision-making perspective)을 바탕으로, 인구통계학적 특성 중 연령, 성별, 직급, 그리고 근속년수의 네 가지 측면에서 팀 내 구성원들 간의 관계적 인구통계를 분석하였고, 이들과 성과와의 관계를 비교분석하였다. 분석결과, 성별 비유사성 및 근속년수 비유사성과 개인성과 간의 관계는 약한 역 U자형 형태, 비선형관계를 보이는 것으로 나타났다. 이는 최근 다양성(diversity) 연구에서 성과와의 비선형관계를 제시하는 연구들과 일치되는 결과이다. 이로써 개인성과에 긍정적인 영향을 주는 적정 수준의 다양성 또는 비유사성이 존재한다는 사실을 알 수 있으며, 이는 성과관리의 책임을 지고 있는 관리자들에게도 중요한 시사점을 제시한다.

Organizations are increasingly adopting work group arrangements such as cross-functional project teams that incorporate differences in functional or educational background (van Knippenberg & Schippers, 2007). Diversity is also introduced into work groups by mergers, acquisitions and joint ventures. As a result, managing diversity forms a major challenge for organizations. In this context, considerable research has focused on the distribution of demographic characteristics within work groups, and the consequences that arise (e.g., Chatman et al., 1998). Although diversity is a group characteristic, there is a stream of research on what is called relational demography (Chattopadgyay et al., 2004; Tsui & O’Reilly, 1989), which focuses on the effects of individuals’ similarity to their work group (e.g., Chatman & Flynn, 2001; Chattopadhyay, 1999) or to their leader (e.g., Tsui et al., 2002) as predictors of individual outcomes (van Knippenberg & Schippers, 2007). A recent development in this research is based largely on two related theories: what has been termed social identity theory (Tajfel & Turner, 1986) and self-categorization theory (Turner, 1987). The key questions in diversity or relational demography research are how similarities or differences between work group members affect individual outcomes or group processes and performance (e.g., van Knippenberg & Schippers, 2007). In this vein, our study explores how the relational demography in work teams can affect individual performance. In contrast to the social categorization perspective, the information/decision-making perspective emphasizes the positive effects of work group diversity. To integrate these opposing perspectives, van Knippenberg et al.(2001) propose the Categorization-Elaboration Model (CEM). This model posits that social category diversity may have a positive effect on task performance, because idea that the effects of relational demography might be curvilinear. The potentially positive effects of the relational demography on individual performance may only apply at a certain level, beyond which the lack of a common frame of reference may obstruct or prevent a full appreciation of all group members’ contributions. Thus, the following hypotheses were made: H1 : Demographic dissimilarity (Relational demography in work group) will have an inverted U-shaped curvilinear relationship to individual performance. H1a : Age dissimilarity will have an inverted U-shaped curvilinear relationship to individual performance. H1b : Gender dissimilarity will have an inverted U-shaped curvilinear relationship to individual performance. H1c : Rank (hierarchical status) dissimilarity will have an inverted U-shaped curvilinear relationship to individual performance. H1d : Tenure dissimilarity will have an inverted U-shaped curvilinear relationship to individual performance. We collected data from a private-sector firm. Participants comprised 130 employees from an electronic company located in Seoul. Relational demography and performance data were collected from the personnel records and documents of the company. We used performance appraisal ratings from supervisors and executives. The results of this study indicate that demographic dissimilarity (relational demography in work group) will have an inverted U-shaped curvilinear relationship to individual performance. The results provided support for Hypothesis 1b and Hypothesis 1d. However, the results did not support Hypothesis 1a and Hypothesis 1c. As noted above, the current study explored the curvilinear effects of relational demography on individual performance. The lack of significant curvilinear effects for age and rank dissimilarity might be explained by the organizational context in which these relational differences are embedded. The workforce at this firm had a high proportion of men (85%). Previous research has shown that the salience of demographic attributes as social categories is lower in