Effects of respondent incentives in Health Interview Surveys. Differences according to survey modes, incentive strategies and incentive values 1
|Convenor||Dr Elena Von Der Lippe (Robert Koch Institute )|
|Coordinator 1||Mr Patrick Schmich (Robert Koch Insitute)|
|Coordinator 2||Mr Matthias Wetzstein (Robert Koch Institute)|
Background: The acceptance of participants into a survey may be the result of an individual cost-benefit consideration. The use of incentive leads to an increased personal gain and therefore, has a positive effect on the survey response.
Methods: Participants were randomly assigned to experimental or control groups. Different types of rewards and varying procedures of delivering to the participants were tested.
Results: The highest response (63%) was achieved with a priori incentives. Non-monetary incentives differed hardly from the control group.
Conclusions: Incentives may lead to an overall improvement of surveys. Preferences of the target group should be considered.
Offering an incentive to the respondents has been shown to increase the response rates in surveys. In this study different methods were tested in search for most cost-efficient procedure. Three different incentive groups were compared to a group without incentive. Lowest response rate had the non-incentive group (18,2%) followed by the group who received post stamps (20,3%). The lottery incentive reached 21,0% response rate and the 10 € voucher incentive 22,7%. Participation varied strongly across all age groups within each incentive group. It is worth to offer different incentives according to the age of respondents.
For our presentation we analyse data from an experiment conducted in the German General Social Survey (ALLBUS) 2014. Target persons were randomly assigned to one of three groups: a) 10Euro conditional on participation, b) 10Euro prepaid incentive, independent of participation, and c) No incentive was provided. The response rate in the prepaid condition was considerably higher than in the promised condition and the no incentive group. In addition to reporting response rate differences we analyse the effect of the different incentives on sample composition and on fieldwork efforts. Finally we discuss some practical issues and the generalizability of the results.
The “CINDI Health Monitoring Survey” is a periodic postal survey amongst adults in Slovenia. Approximately 15.000 individuals were included in nationally representative simple random sample in each of the four waves of the survey. The survey studies behavioural risk factors and chronic non-communicable diseases. From 2001 when response rate was 62.9 %, response rate has fallen down, and was only 49 % in 2008. Therefore, in 2012 different techniques were used to increase response rate (respondent incentives, etc.). In this article the effects of respondent incentives and additional survey mode on response rate in CINDI Survey will be presented.