Measuring Health Status: An Examination of Self-, Interviewer- and Physician-Assessments of Overall Health

Kimberly V. Smith, Princeton University

Self-assessed health status (SAH) is one of the most frequently-used summary health measures in empirical research. Recent studies, however, highlight the problem of reporting heterogeneity in SAH, calling into question the comparability of SAH ratings across countries and population sub-groups. This paper extends previous work to better understand self-evaluations of health and to correct for possible biases arising from the evaluation process. We examine SAH in relation to two alternative measures of respondent health in survey data: interviewer- and physician- assessments of overall health. We use data from a nationally-representative survey of older adults in Taiwan, which includes extensive self-reported, clinical and biomarker data. We find similarities and differences in factors affecting health ratings across evaluators, but low levels of inter-evaluator ratings’ agreement. Our results suggest that SAH tends to be worse than interviewers’ and physicians’ health ratings, and that differences exist across evaluators in the thresholds used when rating health.