Dimensional structure of the demand control support questionnaire: a Brazilian context

Abstract

According to Karasek, job strain results from an interaction between high demands and low decision latitude.

Purpose: To reassess the dimensional structure and evaluate the internal consistency of demand control support questionnaire (DCSQ), a shortened version of job content questionnaire that was not sufficiently evaluated in validation studies.

Methods: The study investigated 825 workers who completed the DCSQ in Rio de Janeiro, Brazil; to 399 workers, the questionnaire was self-administered at a hospital (2004-2005), and 426 workers were interviewed at nine restaurants (2006-2007). Confirmatory factor analysis using structural equation models was used to test theoretical structure of dimensionality. Internal consistency was evaluated by composite reliability and convergent validity by average variance extracted.

Results: Confirmatory factor analysis supported the instrument in three dimensions: demands, skill discretion and decision authority. The best fit model was achieved by removing social support at work and the item repetitive work (skill discretion). A cross-loading from learning new things on demands and an error measurement correlation between work fast and work intense were confirmed. Composite reliability was acceptable for all dimensions, except for demands (0.58), which also showed inadequate average variance extracted (0.32). This final model was confirmed in separate analyses according to work setting, but the loadings of demands were lower for restaurant workers.

Conclusion: Our results indicated that skill discretion and decision authority formed two distinct dimensions. Additionally, the item repetitive work should be removed, as well as one of the items work fast or work intense (demands). Future research is still required to confirm these findings.