The rapid increase of ethnically diverse late onset dementia-affected persons is bringing a special urgency to Alzheimer disease and associated disorders research. By this decade's end, non-European heritage ethnically diverse persons will account for 25% of the dementia-affected cohort and 33% by mid-century. These populations have received some attention in both dementia-focused basic biomedical and social-behavioral research. However, studies differ greatly with regard to the constructs of culture used, the methodologies used, and representativeness of the populations sampled. Moreover, social status factors are often confounded with cultural variables. Therefore, several clarifying research approaches are suggested. In biomedical research, categorically ascribed (US census type) definitions, or assumed (self-reported) ethnic group designators will not suffice where actual biophysiological sampling precision is needed. A strategy for the phenotypic sampling of cohorts is therefore suggested for this research domain. In social-behavioral research, the acceptance of a common operational definition of culture is urged. And, with reference to a specific social status confound, namely literacy, a neuroimaging research strategy is proposed to determine whether non-literates might not be misclassified relative to the determination of their actual cognitive functioning status. Additionally, two conceptual models for addressing and are briefly presented.