Simpson's D index is insensitive to small groups, so contexts with vastly different racial/ethnic compositions can receive nearly equivalent index values¶

Claim · All claims
Evidence · 1 study · review r3 · q2
Currency · newest study 2023 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? Because the index weighs large groups more via the square of each group's proportion, two school contexts with different distributions can both yield a Simpson's D of .50. → Chan 2023

Evidence¶

Chan 2023¶

Chan, M.-k., & Benner, A. D. (2023). A Consideration of Racial/Ethnic Diversity Conceptualization and Measurement: Clarifying Ambiguities and Advancing Scholarship. Educational Psychologist (accepted manuscript). https://orcid.org/0000-0001-6451-2417

q2 · i? · review · r3

Analytical demonstration (Figure 3) comparing School A (50% Latinx, 50% White) and School B (68% Latinx, 21% White, 7% Asian, 3% Other, 1% Black): "These two contexts have nearly equivalent Simpson’s D index values despite their vastly different racial/ethnic compositions."

"These two contexts have nearly equivalent Simpson’s D index values despite their vastly different racial/ethnic compositions because the index is insensitive to the presence of small groups owing to its probabilistic origin."

Discussion¶

Related Claims¶

• The range and value of diversity indices are constrained by the number of racial/ethnic groups included, complicating comparisons across studies — related
• In school racial/ethnic diversity research, mathematical diversity indices operationalizing dual-concept diversity are the dominant measure, with Simpson's D the most widely applied, and most studies assess static diversity at a single timepoint — related