Abstract

We propose to conduct a nonresponse follow up survey with nonrespondents to the Filterfragen. This study will determine which method most successfully collects important data from these cases and provide guidance to other surveys at IAB and elsewhere how to better understand their nonrespondent population.

All surveys struggle with nonresponse, and often just a small bit of information about the nonresponding cases would be helpful in building adjustment models. One method used to understand nonrespondents are follow-up surveys with the nonresponding cases. These are often abbreviated questionnaires, conducted in a different mode than the initial survey, which attempt to collect a few key data points about the cases which did not take part in the initial study. Methodological studies show that switching modes tends to increase response rates among initial nonrespondents (Millar and Dillman, 2011). While these follow-up studies themselves suffer from nonresponse, they seem to offer useful data about the nonrespondents (Raessler and Schnell, 2003; Singh et al, 2003).

However, the quality of the data collected via these techniques has never been evaluated against an external source. Taking advantage of the BeH, LeH and LHG datasets available to researchers at the IAB, our follow-up study with the Filterfragen nonrespondents will compare two different follow-up study methodologies and evaluate which one provides the most accurate and useful data about the nonresponding cases.