Validation and benchmarks for the Copenhagen Psychosocial Questionnaire (COPSOQ III) in an Australian working population sample

Abstract

Background

This study presents an analysis of the reliability and validity of the long version of the Copenhagen Psychosocial Questionnaire (COPSOQ III) in Australia and establishes benchmarks for key dimensions of psychosocial work environment.

Methods

A random sample was sourced from a panel of 2,446 working Australians aged 18–79. The study establishes population benchmarks for key measures of the psychosocial work environment and employee mental health, and assesses variation by gender, employment status, work sector, professional status, and occupational classification. Validation entailed an assessment of internal consistency reliability of the measures, followed by a preliminary Exploratory Factor Analysis (EFA) to initially examine construct validity of the instrument, and continued with a Confirmatory Factor Analysis (CFA) to validate the scales. Working population mean scores for the key dimensions of COPSOQ-III were calculated and presented as benchmarks for use in workplace risk assessment and management.

Results

Except for one measure, the analysis revealed strong reliability coefficients for 31 measures with more than two items (> 0.7). Only one measure for ‘demand for hiding emotions’ with more than two items had a reliability coefficient of 0.663. The EFA unveiled a four-factor structure in psychosocial working conditions, and a two-factor model in factors related to the effects. The CFA involved item-level analysis, leading to the removal of seven items to improve model fit.

Conclusion

The present study provides a high-level validation of the long version of COPSOQ, and thus supports its use as an instrument for research as well as psychosocial risk assessment and management in Australia.