What does the MAIHDA method explain?

Area of Science:

• Quantitative intersectional modeling
• Statistical methodology
• Health disparities research

Background:

• Multilevel analysis of individual heterogeneity and discriminatory accuracy (MAIHDA) is a novel quantitative intersectional modeling approach.
• MAIHDA utilizes demographic variables and their combined constructs (strata) to analyze outcomes.
• The method aims to uniquely identify additive and intersectional effects.

Purpose of the Study:

• To critically evaluate the methodological claims and operationalization of MAIHDA.
• To demonstrate the mis-specification of MAIHDA in quantitative intersectional analyses.
• To propose alternative, more accurate modeling strategies.

Main Methods:

• Analysis of the MAIHDA framework, specifically its treatment of strata as a level 2 context.
• Comparison of MAIHDA's multilevel specification with single-level models.
• Identification of collinearity and redundancy issues arising from variable duplication across model levels.

Main Results:

• MAIHDA incorrectly operationalizes strata as a level 2 context; strata are individual-level composite variables.
• The inclusion of demographic variables at both level 1 and level 2 leads to mis-specified, collinear, and redundant models.
• MAIHDA fails to uniquely identify additive and intersectional effects as promised.

Conclusions:

• MAIHDA is methodologically flawed for quantitative intersectional analysis.
• Single-level models incorporating demographic variables with interactions or strata as fixed effects are more accurate.
• The findings necessitate a re-evaluation of multilevel approaches for intersectional research.