An Alternative Explanation for the Asymmetrical Effects in Relational Demography Research

Abstract

Several relational demography studies have demonstrated asymmetrical effects of dissimilarity across different groups of employees. Some groups of employees appear to be more influenced by differences from fellow employees, whereas other groups of employees appear indifferent. Although numerous theoretical explanations have been developed to account for these asymmetrical effects, this paper argues that such effects are actually methodological artifacts resulting from an imbalance in the proportion of group members and deficiencies in the most commonly used measure of dissimilarity, Euclidean distance. This paper illustrates how such asymmetrical effects can be observed even when none exist in the population. Suggestions for methodological improvements in future diversity research are discussed along with recommendations for managing diversity in organizations.

A Socioecological Approach to Relational Demography: How Relative Representation and Respectful Coworkers Affect Job Attitudes

“…The effect on turnover is particularly compelling given that employees who encounter low levels of relative representation might leave organizations and seek refuge in more balanced environments. That these effects were pronounced among ethnic minority employees is also meaningful in light of previous work on asymmetry which suggested that restriction of range might underestimate the effects of representation among ethnic minority group members (Tonidandel, Avery, Bucholtz, & McKay, 2008).…”

“…As Tonidandel and colleagues argued, ÒBecause threats to oneÕs identity are more commonplace for those in the numerical minority, identity affirmation concerns tend to be greater among women and minorities than among White menÓ (Tonidandel, Avery, Bucholtz, & McKay, 2008;p. 619).…”

The moderating role of leader and follower sex in dyads on the leadership behavior–leader effectiveness relationships

“…Some researchers have argued that the effects of similarity in demographic variables such as sex may be asymmetric across different demographic groups (Chattopadhyay, 1999;Tsui et al, 1992). Other researchers have claimed that such results represent a methodological artifact resulting from an imbalance in proportions of different demographic combinations (Tonidandel, Avery, Bucholtz, & McKay, 2008); however, it is a fact of organizational life that men and women are unequally distributed across leader and subordinate roles in most organizations. Applying the notion of asymmetric effects of sex similarity to the present study, we suggest that female leaders may experience different consequences of being similar or dissimilar in sex to their subordinates than male leaders as a result of gender-based expectations of leaders (Eagly & Karau, 2002).…”

Managing diversity in organizations: An integrative model and agenda for future research

“…The first proposition suggested that identity concerns moderate the relationship between employee dissimilarity and work group identification. Existing measures in the tradition of the relational demography approach, which are frequently used to capture employee dissimilarity, have been criticized on multiple grounds, such as their inability to compensate for missing data or account for unequal group and subgroup sizes, and their leading to ambiguity regarding their conceptual interpretation (Allen, Stanley, Williams, & Ross, 2007;Riordan & Wayne, 2008;Tonidandel, Avery, Bucholtz, & McKay, 2008). Adding to this debate, these existing measures cannot capture differences on categorical variables or variables with an ordinal scale, nor are they able to capture different types of dissimilarity other than separation (e.g., categorical differences), such as for instance variety (e.g., differences in knowledge and information) and disparity (e.g., status differences) (cf.…”