OBJECTIVES:

To examine the effect of job strain on survey response.

METHODS:

1,613 health care workers received a self-administered questionnaire. Thirty percent of them completed the survey on personal time without any personal monetary compensation. Working conditions were extracted by job title from the national database O*NET 6.0. Job strain was defined as the ratio of job demands to job control. Two complementary models (multi-level logistic and binomial pseudo Poisson regressions) were used to model individual survey response as a function of individual level demographic variables (age and gender), job-level socioeconomic status (SES) and job strain, and facility type (third level).

RESULTS:

Survey response was associated with higher SES and with less job strain. The association of SES and survey response was mediated by job strain.

CONCLUSION:

Employees' exposure to job strain may be an important influence on survey response, at least for workers who are not compensated for their time in completing a survey.

(c) 2008 Wiley-Liss, Inc.