This guide provides methodological advice for researchers examining diversity in organisational settings, with particular focus on boards of directors. Despite substantial academic interest in the relationship between board diversity and firm outcomes, empirical findings remain inconsistent, partly due to measurement strategies that fail to capture diversity's multifaceted nature and complex effects on group dynamics and performance. Drawing on an established theoretical typology, I distinguish between diversity-as-variety (the breadth of categories represented) and diversity-as-separation (the degree of distance between categories). Using board nationality as the primary illustration, I demonstrate how these two aspects can be measured for statistical analysis. This distinction matters because variety and separation may have countervailing effects on outcomes, which remain unobservable when researchers model only net effects using simpler measures. For diversity-as-variety, I explain the Blau Index and Normalized Diversity By number of links (NDBL). For diversity-as-separation, I describe cultural distance measures between group members. These distinctions have been shown to matter in a recent study that reports nationality diversity-as-variety positively affects firm value, but this relationship is negatively moderated by cultural separation between board members. Such nuanced findings would be impossible to detect using simple dummy, count, or proportional measures of diversity. The guide provides directions for practical implementation, including worked examples and a comprehensive reference table showing how to measure variety and separation for different board attributes. While focused on boards of directors, these measurement principles apply more broadly to research on diversity in small decision-making groups across organisational settings.
Building similarity graph...
Analyzing shared references across papers
Loading...
ELISABETH; id_orcid 0000-0001-8877-7487 DEDMAN
Building similarity graph...
Analyzing shared references across papers
Loading...
ELISABETH; id_orcid 0000-0001-8877-7487 DEDMAN (Fri,) studied this question.