OBJECTIVE: The Attitudes to Psychiatry Scale (APS) is a tool used to assess medical students' attitudes toward psychiatry. This study sought to examine the internal validity of the APS in order to identify dimensions within the questionnaire. METHOD: Using data collected from 549 medical students from India and Ghana, the authors analyzed 28 questions of the APS independently of the original five dimensions defined by Balon et al. in the original APS questionnaire, using principal-components analysis to test whether questions correlated to form dimensions within the questionnaire. The data were further tested for redundancy by Cronbach alpha. RESULTS: The investigation yielded five dimensions, but, on filtering the information, there was good independent correlation in four of the five dimensions emerging from the analysis. These were (1) inspiration from medical school; (2) the stigma of psychiatry; (3) the merits of psychiatry as scientific medicine; and (4) the effectiveness of treatment, from which a suggested edited scale emerged. CONCLUSION: On examining the APS scale using principal-components analysis, the authors found clusters of questions around four themes that do seem intuitively relevant for attitudes toward psychiatry. They are not the same as Balon et al. 's original subscales, although there are similarities. Using the questions in the emerging themes, the authors have suggested a modified questionnaire that appears to have good internal validity.
OBJECTIVE: The Attitudes to Psychiatry Scale (APS) is a tool used to assess medical students' attitudes toward psychiatry. This study sought to examine the internal validity of the APS in order to identify dimensions within the questionnaire. METHOD: Using data collected from 549 medical students from India and Ghana, the authors analyzed 28 questions of the APS independently of the original five dimensions defined by Balon et al. in the original APS questionnaire, using principal-components analysis to test whether questions correlated to form dimensions within the questionnaire. The data were further tested for redundancy by Cronbach alpha. RESULTS: The investigation yielded five dimensions, but, on filtering the information, there was good independent correlation in four of the five dimensions emerging from the analysis. These were (1) inspiration from medical school; (2) the stigma of psychiatry; (3) the merits of psychiatry as scientific medicine; and (4) the effectiveness of treatment, from which a suggested edited scale emerged. CONCLUSION: On examining the APS scale using principal-components analysis, the authors found clusters of questions around four themes that do seem intuitively relevant for attitudes toward psychiatry. They are not the same as Balon et al. 's original subscales, although there are similarities. Using the questions in the emerging themes, the authors have suggested a modified questionnaire that appears to have good internal validity.