Abstract
Previous studies have found that self-respondents tend to report a greater number of health problems than do household proxies. It has not been clear, however, if such results indicate reporting bias or reflect underlying health differences in the self-respondent and household proxy populations. Verification data from a survey of medical providers were examined in conjunction with the results of a national household survey. Our analysis indicates that the use of proxies does not increase misreporting. Overall reporting error between self-respondents and proxy-respondents is comparable on reports of mental illness. When physical stigmatizing conditions are examined, the use of proxies actually appears to be preferable to interviewing respondents directly.
Get full access to this article
View all access options for this article.