BACKGROUND: Employers are legally obligated to ensure the safety and health of employees, including the organizational and psychosocial working environment. The Copenhagen Psychosocial Questionnaire (COPSOQ III) covers multiple dimensions of the work environment. COPSOQ III has three parts: a) work environment b) conflicts and offensive behaviours and c) health and welfare. We translated all three parts into Norwegian and evaluated the statistical properties of the 28 work environment dimensions in part a), using a sample of registered nurses.

METHODS: The original English version was translated into Norwegian and back translated into English; the two versions were compared, and adjustments made. In total, 86 of 99 items from the translated version were included in a survey to which 8804 registered nurses responded. Item response theory models designed for ordinal manifest variables were used to evaluate construct validity and identify potential redundant items. A standard confirmatory factor analysis was performed to verify the latent dimensionality established in the original version, and a more exploratory factor analysis without restrictions is included to determine dependency between items and to identify separable dimensions.

RESULTS: The measure of sampling adequacy shows that the data are well suited for factor analyses. The latent dimensionality in the original version is confirmed in the Norwegian translated version and the scale reliability is high for all dimensions except 'Demands for Hiding Emotions'. In this homogenous sample, eight of the 28 dimensions are found not to be separate dimensions as items covering these dimensions loaded onto the same factor. Moreover, little information is provided at the low and high ends of exposure for some dimensions in this sample. Of the 86 items included, 14 are found to be potential candidates for removal to obtain a shorter Norwegian version.

CONCLUSION: The established Norwegian translation of COPSOQ III can be used in further research about working environment factors and health and wellbeing in Norway. The extended use of the instrument internationally enables comparative studies, which can increase the knowledge and understanding of similarities and differences between labour markets in different countries. This first validation study shows that the Norwegian version has strong statistical properties like the original, and can be used to assess work environment factors, including relational and emotional risk factors and resources available at the workplace.