Race-related cognitive test bias in the active study: a mimic model approach

Abstract

The present study investigated evidence for race-related test bias in cognitive measures used in the baseline assessment of the ACTIVE clinical trial. Test bias against African Americans has been documented in both cognitive aging and early life span studies. Despite significant mean performance differences, Multiple Indicators Multiple Causes (MIMIC) models suggested most differences were at the construct level. There was little evidence that specific measures put either group at particular advantage or disadvantage and little evidence of cognitive test bias in this sample. Small group differences in education, cognitive status, and health suggest positive selection may have attenuated possible biases.

Figures

Figure 1. MIMIC Model: An illustration

Note. a paths are factor loadings, and express the relationship between the latent construct and each of three measures of that construct; the b path shows the relationship between the exogenous covariateand the latent construct. This path addresses whether there is an overall relationship between the covariate and the construct. The c path assesses whether, after controlling for the relationship between covariate and the construct, there is an additional unique direct residual relationship between the covariate and each indicator. If the relationship between the covariate and Measure 1 is more strongly positive than the relationship between the covariate and the construct, then one would expect an additional significant direct positive path between the covariate and Measure 1. This would suggest that Measure 1 is positively biased. In other words, it shows a significantly stronger relationship to the covariate than the overall construct. Expressed differently, this would mean that the residual variance in Measure 1 (variance not explained by the common construct factor) shows additional relationship to the covariate, above and beyond that covariate’s indirect influence on Measure 1 via the latent construct.

Figure 2. MIMIC Model in the current study

Note. MIMIC model of race effects on reasoning, memory, and speed, with years of education, age and gender as covariates. Paths shown are as follows: a, influence of ethnicity on latent variable; b, influence of covariates on latent variable; c, influence of race on specific measure; d, influence of covariates on specific measure; e, factor loading (relationship between latent construct and specific measure). Paths are shown only for one latent factor, but these paths were estimated for all latent variables and measured indicators. Dummy variables representing each of the assessment sites were also included as exogenous covariates in these models, but they are not shown in order to simplify the figure.