OBJECTIVE: The theory behind the Job Content Questionnaire (JCQ) presumes that the "objective" social environment is measurable via self-report inventories such as the JCQ. Hence, it is expected that workers in identical work will respond highly similar. However, since no studies have evaluated this basic assumption, we decided to investigate whether workers performing highly similar work also responded similarly to the JCQ. METHODS: JCQ data from a rubber-manufacturing (RM: n = 95) and a mechanical assembly company (MA: n = 119) were examined. On each worksite, men and women performed identical machine-paced job tasks. A population sample (n = 8,542) served as a reference group. RESULTS: In both the RM and MA groups, the job support questions were rated most similar. Yet, there was a substantial variation as regards choosing to agree or disagree with single JCQ items. The variation was also reflected in the scale scores. In the RM and MA groups, the variance of job demand and job control scores was 64-87% of that of the population sample. For job support scores, the corresponding variation was 42-87%. CONCLUSION: Conducting highly similar work does not lead to highly similar reports in the JCQ. In view of the large response variation, it seems that the attempt to avoid personal influence by minimizing the self-reflexive component in the questions asked, and using response alternative that indicates degree of agreement, does not seem to work as intended.
OBJECTIVE: The theory behind the Job Content Questionnaire (JCQ) presumes that the "objective" social environment is measurable via self-report inventories such as the JCQ. Hence, it is expected that workers in identical work will respond highly similar. However, since no studies have evaluated this basic assumption, we decided to investigate whether workers performing highly similar work also responded similarly to the JCQ. METHODS: JCQ data from a rubber-manufacturing (RM: n = 95) and a mechanical assembly company (MA: n = 119) were examined. On each worksite, men and women performed identical machine-paced job tasks. A population sample (n = 8,542) served as a reference group. RESULTS: In both the RM and MA groups, the job support questions were rated most similar. Yet, there was a substantial variation as regards choosing to agree or disagree with single JCQ items. The variation was also reflected in the scale scores. In the RM and MA groups, the variance of job demand and job control scores was 64-87% of that of the population sample. For job support scores, the corresponding variation was 42-87%. CONCLUSION: Conducting highly similar work does not lead to highly similar reports in the JCQ. In view of the large response variation, it seems that the attempt to avoid personal influence by minimizing the self-reflexive component in the questions asked, and using response alternative that indicates degree of agreement, does not seem to work as intended.