An Experiment with Data Collection Modes and Incentives in the Dutch Family and Fertility Survey for Young Moroccans and Turks
Statistics Netherlands In this article, the authors discuss the results of an experiment embedded in the Dutch Family and Fertility Survey for Young Moroccans and Turks. The effects of two factors on response rates and data quality are investigated. The first factor is a promised incentive versus no incentive. In the second factor, the part of the questionnaire with presumed sensitive questions is administered via computer-assisted self-interviewing (CASI), instead of completely administering the questionnaire via computer-assisted personal interviewing (CAPI). No significant main effect of the promised incentive on the response rates could be found. There are, however, interactions between the incentive, gender, and ethnicity, which show that different subpopulations react differently to the promised incentive. Furthermore, there is no evidence that the data collected by means of CASI results in more valid answers. Still, when the data are collected by means of CASI, a substantial increase in the rate of "don't know" responses is observed.
Key Words: ethnic minoritie mode effect nonrespons promised incentiv response bia
Field Methods, Vol. 18, No. 3,
321-334 (2006) |
||||