Assessment of explanatory models of mental illness: effects of patient and interviewer characteristics

Explanatory models (EMs) refer to patients' causal attributions of illness and have been shown to affect treatment preference and outcome. Reliable and valid assessment of EMs may be hindered by interviewer and respondent disparities on certain demographic characteristics, such as ethnicity. The present study examined (a) whether ethnic minority patients reported different EMs to ethnically similar interviewers in comparison with those with a different ethnicity, and (b) whether this effect was related to respondents' social desirability, the perceived rapport with the interviewer and level of uncertainty toward their EMs. A total of 55 patients of Turkish and Moroccan origins with mood and anxiety disorders were randomly assigned to ethnically similar or dissimilar interviewers. EMs were assessed, using a semi-structured interview, across 11 different categories of causes. Participants who were interviewed by an ethnically similar interviewer perceived interpersonal, victimization and religious/mystical causes as more important, whereas interviews by ethnically dissimilar interviewers generated higher scores on medical causes. These effects were not mediated by the perceived rapport with the interviewer, and social desirability had a modest impact on the results. Higher uncertainty among participants toward medical and religious/mystical causes seemed to be associated with greater adjustment in the report of these EMs. The findings have significant implications for interviewer selection in epidemiological research and clinical practice.