A methodological note on diversity measurement

Abstract

This research addresses the question, do categorical and continuous measures capture the same construct of diversity? Using analysis of variance, cluster analysis, and discriminant analysis, we investigated whether continuous measures (entropy and product count) differentiate between diversification categories, whether continuous measures converted to categories and subjectively assigned categories classified companies similarly, and whether continuous and categorical measures predicted similar diversity–performance relationships. We concluded that the techniques were associated but did not yield the same performance predictions. For researchers investigating diversity–performance relationships, choice of measurement technique will influence research results. Our research results suggest that attempts to combine categorical and continuous techniques as a way to overcome the limitations of both methods is not appropriate.

The Performance of Private Equity Funds: Does Diversification Matter?

“…Finally, the measures of 'systematic' diversification are not capturing all aspects of diversification. The literature on the impact of diversification in big sized public companies find quantitative diversification measures to explain less of financial success than qualitative diversification measures (Hall & John 1994). However, applying qualitative diversification measures similar to the relation-ratios of Rumelt on a sample of portfolio companies appears to be an unpractical task (Rumelt 1974, Rumelt 1982.…”

Redirecting Research Efforts on the Diversification–Performance Linkage: The Search for Synergy

“…Results suggest that while using both Rumelt and entropy measures improves the accuracy of the study, using either of the two measures should still lead to acceptable results Lubatkin, Merchant & Srinivasan, 1993 Comparison between Rumelt's categorical measure of diversification and continuous approaches. Results reveal high degree of correspondence between the narrow (4-digit) and broad (2-digit) spectrum measures of diversification and Rumelt's categorical measures Hall and St. John, 1994 Comparison between Rumelt's categorical measure, continuous measures (product count and entropy) and continuous scores converted to strategy categories. Results report a close association between the three types of measures, but different performance predictions Robins and Wiersema, 2003 Comparison in terms of content validity between the related component of the entropy index and the concentric index as measures of relatedness.…”

“…Hall and St. John (1994) show that categorical and continuous measures appear to be associated, but they seem to capture different aspects of the relationship between diversification and performance. Robins and Wiersema (2003) show that the related components of both concentric and entropy measures can be sensitive to features of the corporate portfolio composition and can therefore create ambiguities.…”

“…see Chatterjee and Blocher, 1992;Hoskisson, Hitt, Johnson & Moesel, 1993;Lubatkin, Merchant & Srinivasan, 1993;Hall & St. John, 1994;Montgomery, 1982;Pitts and Hopkins, 1982;Robins & Wiersema, 2003 for extensive reviews and comparison), which overall suggest thatalthough the different measures are characterized by a good level of consistency-they tend to capture slightly different aspects of the phenomenon and should be used accordingly.…”

Diversification, Relatedness, And Firm Performance: Empirical Evidence From China

“…ROI is one of the most widely employed measures of performance (Hall, 1994;Markides, 1995;Palich et al, 2000). ROI is calculated as:…”