- Kees van Berkel
- Barry Schouten
Adaptive survey designs aim to get a better balanced response by putting different effort in different groups of the population. They are effective in improving survey results and reducing survey costs. Over the last decade, a lot of methodological research has been done on this topic. In this webinar, an overview of research on adaptive survey design will be presented. After that, the implementation of adaptive survey design is demonstrated at the hand of the Dutch Health survey. In this survey a sequential mixed mode strategy CAWI followed by CAPI is applied. The feature to adapt is the CAPI follow-up. The design is developed in such a way that through stratified selection of internet nonrespondents for face to face follow-up, nonresponse bias can potentially be reduced. Mode-specific measurement errors which likely exist in many surveys, especially in the Health survey, have not received much attention in adaptive survey design. In order to separate and quantify mode-specific measurement effects and selection effects, an experiment with re-interviews is planned for the Health survey. The sampling and questionnaire design including analysis possibilities are discussed in the webinar.