The Construct and Verification of Group Faultlines Metrics: an Integrated Perspective
HAN Li-Feng,WANG Zhong-Ming
DOI: https://doi.org/10.3724/sp.j.1041.2010.01082
2011-01-01
Acta Psychologica Sinica
Abstract:As a notable advance compared with traditional diversity research, Lau and Murnighan (1998) proposed the concept to capture aligned effects of multi-dimensional characteristics and their interrelationships. Much progress has been gained in terms of theoretical development and empirical analysis about faultline research, but an examination of previous literature has shown little further work on how to quantitatively measure faultline strength. To the best of our knowledge, only Thatcher, Jehn and Zanutto (2003) Fau-Index, Shaw (2004) FLS-Index along with Trezzini (2008) PMD-Index have provided systematic and rigor measurement of faultline strength. Each of these indices has its robustness. In nature, Fau can be interpreted as amount of total variance accounted for by intergroup heterogeneity, while FLS is more like a mixed product of subgroup internal alignment and cross-subgroup alignment. PMD, if we scrutinize, will find it a special case of class of polarization measures introduced by Esteban and Ray (1994). Be that as it may, all the indices mentioned above have ignored the faultline width, which is a critical dimension concerning that an identical array of attributes will have different dynamics if distances that exist between subgroups diverge. The fact that faultline width may play a unique role on behaviour was not reflected in any of the three indices. This study aims to resolve this problem by constructing a new integrated metric based on simultaneously measuring subgroup internal alignment (IA), the degree of dispersion between subgroups (Fau) and subgroup distance (D). The resulting metric, namely Integrated-Group-Faultline-Strength (IGFS) was represented in the following form: IGFS=IA×Fau×D. A total of 21 well-designed groups were selected to test the validity of our new metric. These groups all came from published articles. We did this because we wanted to present all sorts of teams that could be typically imagined. The benefit of doing so is apparent, with groups of all kinds be sufficiently examined, the validity of IGFS could be justified. Ten evaluation criteria were also adapted from previous research, which were theoretically derivative as well as intuitional oriented. We relied on these ten criteria to see if they could be coincided with IGFS metric and the others. Results indicated that IGFS fit better than FLS and PMD. Moreover, IGFS fit all the ten criteria well. Actually there also existed one problem related to the subjective perception of group members. In many cases, people tend to cognize and behave according to what they perceive rather than what really is. So questionnaire designed to explore people’s psychological dimensions necessitates the use of one brand new indicator to measure the distances of members’ ratings. We proposed to adopt regression coefficients as elements for distance calculation, which could replace the last term of PMD to form a modified metric appropriate for subjective perception faultlines measurement. In this regard, this study contributed substantially to faultline indices development and provided fundamental base for future research. Furthermore, cluster analysis or cohesive subgroup algorithms seem to be two prominent methods for group faultlines calculation.