Download Now Free registration required
Job satisfaction may affect the propensity to respond to job satisfaction surveys, so that estimates of average satisfaction and the effects of determinants of satisfaction may be biased. The authors examine response bias using data from a postal job satisfaction survey of family doctors. They link all the sampled doctors to an administrative database and so have information on the characteristics of responders and non-responders. Allowing for selection increases the estimate of mean job satisfaction in 2005 and the estimated change in mean job satisfaction between 2004 and 2005. Estimates of the determinants of job satisfaction are generally insensitive to response bias.
- Format: PDF
- Size: 544.3 KB