The Development and Analysis of a Measure of Group Faultlines

Abstract

Lau and Murnighan introduced the "faultline" concept into the group diversity literature. Faultlines are "hypothetical dividing lines that may split a group into subgroups based on one or more attributes." Faultlines result from the pattern, not just the dispersion, of attributes among group members. Measures using dispersion indices do not capture faultline potential. This article reports the development of faultline measures and the characteristics of these measures in two samples.

Diversity faultlines, shared objectives, and top management team performance

“…faultlines based on artificial dichotomies may both be underestimated and overestimated as compared with what would be justified on the basis of the continuous variable). We therefore propose an index that allows incorporation of continuous variables, but is otherwise consistent with the principles behind Shaw's (2004) index, which differs from Thatcher et al's (2003) in that it captures within-subgroup similarity as well as between-subgroup differences consistent with the notion that within-group similarities and between-group differences drive categorization salience based on faultlines (van Knippenberg et al, 2004a).…”

“…Second, our measure maintains the continuous nature of an interval variable like tenure. Various methods for measuring faultlines have been proposed and used, notably Thatcher et al's (2003) index (Fau), and Shaw's (2004) index based on a Chi-squared statistic. All of these measures have in common that they are based on categorical measurement of variables.…”

“…Faultlines are typically operationalized by an index that reflects the extent to which positions on different dimensions of diversity are correlated, where perfect correlations represent perfect faultlines (Shaw, 2004;cf. Thatcher et al, 2003).…”

Team Diversity and Categorization Salience

“…Phase 3 requires the creation of dummy diversity variables at the dyadic level (relational diversity). Crocker, Fiske, & Taylor, 1984;Rosch, 1975;Shaw, 2004;Turner, 1987).…”

“…Phase 3 requires the creation of dummy diversity variables at the dyadic level (relational diversity). Table 1)Insert Fiske, & Taylor, 1984;Rosch, 1975;Shaw, 2004;Turner, 1987).According to Turner (1987), "category formation is relative to the frame of reference (the pool of psychologically relevant stimuli) and hence the available contrasts provided by the salient stimulus field, and depends not just on 'similarities' between stimuli, as is usually assumed, but on relative similarities, on more similarity (or less difference) between certain stimuli than between those and others" (p. 47).To apply the measure of categorization salience that we are proposing, participants should therefore be given in a survey the names of all members of their team and asked to categorize them as belonging to in-groups vs. out-groups. Tajfel and colleagues (e.g., Billig & Tajfel, 1973) noted that the variable of social categorization always coincides with the variable of similarity.…”