Distorting judgments of mental health.

Abstract

Labeling bias refers to subjects judging normal individuals as mentally ill as a function of diagnostically irrelevant stimuli (e.g., suggestions of mental illness made by a prestige figure). Three studies investigating the generality of the labeling bias effect are reported. Experiment 1 failed to find the effect with undergraduates, but demonstrated that these subjects can discriminate an audio tape recording of a psychiatric patient from a tape of a normal person. Experiment 2 demonstrated labeling bias in mental health service workers. Experiment 3 produced labeling bias in undergraduates by manipulating the "relevance" of the prestige figure. The data from the three studies were interpreted as indicating that, contrary to previous interpretation, professional identity is not a necessary condition for producing labeling bias. Reanalysis of the data of previous research supported the hypothesis that labeling bias is extremely general, and probably related to attractiveness of the communicator who gives the suggestion of mental illness, setting, ambiguity of desired response, and other similar variables well documented in the social psychology literature.