Accurately understanding treatment effectiveness across heterogeneous populations requires equivalent measurement across the population. Measurement bias refers to the possibility that individuals with identical health respond dissimilarly to questions about their health as a function of their ethnicity or another variable. Without establishing equivalent measurement, the field cannot comparatively evaluate what works best for whom, draw strong conclusions about disparate outcomes, and support evidence-based practice and policy.
Using data from a representative sample of the 2001–2002 US, the application of multiple-group multiple-indicator, multiple-cause, models to evaluate and correct for measurement bias was described. Analyses investigated whether 10 items operationalizing alcohol abuse provided equivalent measurement across different education and income levels for white (n = 16,480), black/African-American (n = 4139), and Hispanic (n = 4893) individuals.
Analyses uncovered a complex pattern of measurement bias across educational attainment, poverty status, and minority status. Ignoring bias, black/African-Americans and Hispanics appeared to have significantly more alcohol abuse-related behavior than whites. After adjusting for bias, whites and Hispanics demonstrated comparable levels of alcohol abuse and black/African-Americans had significantly lower levels of alcohol abuse than whites.
Measurement bias can lead to erroneous conclusions about alcohol abuse across race and ethnicity. This would subsequently bias research that comparatively examines the correlates of alcohol abuse and research that investigates the comparative effectiveness of alcohol abuse treatments across these groups. Using model-based estimates can mitigate errors like these and lead to more accurate conclusions across heterogeneous populations.
From the Division of Health Policy and Clinical Effectiveness, Cincinnati Children's Hospital Medical Center, University of Cincinnati Medical School, Cincinnati, OH.
Supported by National Institute of Nursing Research grant R15NR10631.
Reprints: Adam C. Carle, PhD, Division of Health Policy and Clinical Effectiveness, Cincinnati Children's Hospital Medical Center, University of Cincinnati, 3333 Burnett Ave., Cincinnati, OH 45226. E-mail: email@example.com.