A psychometric evaluation of the Occupational Stress Indicator

Three samples of volunteer working adult participants, drawn from two countries (United Kingdom and New Zealand), provided data using the pencil and paper version of the Occupational Stress Indicator questionnaire (OSI; Cooper, Sloan, & Williams, 1988). In an attempt to examine the degree to which the psychometric structure of the test was reflected in the published score key, the three datasets were comprehensively analysed using item analysis and exploratory factor analysis. Little similarity existed between the published 25 subscale score keys and the meaningful, 11‐scale psychometric structure found from these analyses. It is concluded that the current norms for the test are of dubious validity, and if these findings are replicated the normative data should be re‐scored, perhaps using the revised score keys presented in this paper. The model underlying the OSI has not been tested comprehensively to date. The present results neither support nor disprove the model, but the new scales could be used to test it.