Mitigating Systematic Measurement Error in Comparative Effectiveness Research for Children

US minority children and children from low socioeconomic status (SES) and non-English speaking homes suffer health disparities. Accurately evaluating children's health requires sound measurement. Yet, little research explores the psychometric properties of children's health estimates derived from the primary surveys and measures used to understand children's health. Less research addresses the possibility that systematic measurement error influences research. Measurement bias refers to the possibility that individuals with identical health respond dissimilarly to questions about their health as a function of their race, ethnicity, SES, or language. Bias can obscure differences, decrease reliability and validity, and render comparisons impossible. Without establishing equivalent measurement across heterogeneous backgrounds, the field cannot comparatively evaluate what works best for whom, support evidence-based care, and address disparities.

I will apply latent variable models to examine the measurement of health conditions among children (e.g., depression, generalized anxiety disorder, attention-deficit hyperactivity disorder) using data from the primary surveys used to understand children's health (e.g., National Health and Nutrition Examination Survey, National Survey of Children's Health). I will establish the measures' psychometric properties; examine bias' role across diverse populations (e.g., race, ethnicity, SES, language), and assess bias' influence on comparative effectiveness research. I will develop a model-based correction to mitigate bias' influence.

Findings will address bias as an impediment to comparative effectiveness research and clarify whether children's outcomes differ across racial, ethnic, and SES backgrounds. This will promote efforts: to conduct comparative effectiveness research, apply and advance evidence-based practice and policy, improve and understand children's health, and eradicate disparities.