A re-analysis of the occupational stress indicator

This paper investigates the reliability of the Occupational Stress Indicator (OSI). Data from a sample of university staff, drawn from all areas of an urban university, are used to reassess the apparently low reliabilities of many of the OSI subscales reported by Cooper et al. (1988). In addition, factor analysis results are reported for the first time for the sources of pressure data. The reliability data reported here, while higher reliabilities than originally obtained, remain unacceptably low. A lack of stability in the device as it is currently formulated seems apparent. The paper also presents for the first time a detailed analysis of the sources of pressure scale, indicating a solution different from that proposed by Cooper et al. (1988). Certain areas of the OSI clearly need refinement; the inclusion of locus of control and type A behaviour as personality variables in particular is called into question. Observations regarding the particular strengths and weaknesses of this device, and suggestions for future refinements, are offered.