Brief Report: An Evaluation of Item Bias on the Functional Activities Questionnaire

Abstract Objective To evaluate whether the Functional Activities Questionnaire (FAQ), a commonly used measure of functional status in neurodegenerative disease research, performs equivalently across demographically diverse subgroups of participants. Method The FAQs from 30,613 National Alzheimer’s Coordinating Center participants were evaluated with a hybrid graded response model-logistic ordinal regression approach to determine the presence of differential item functioning (DIF) within five demographic groups: education, ethnicity, race, language, and sex. Results Measurable DIF was observed for FAQ items in all groups; however, measures of effect size, latent trait distributions, and item characteristic curves suggested that the impact was minimal for research and practice. Conclusions The FAQ is able to provide minimally biased assessments of daily functioning across diverse participants, suggesting potential value for offsetting disparities in diagnosis and treatment.