Predicting response rates of all and recruited respondents: A first attempt

Abstract

This paper reports on the link between a well-defined measure of response burden and response rates among all (uncommitted) and pre-recruited respondents. We show within the limits of our sample of 68 survey waves (including pre-tests; resulting from 35 studies) that the response burden impact is mediated by the level of the commitment of the respondents and the presence of a monetary incentive. This is the first time that a research group provides a response rate forecasting model for its own work and for others to adopt, test and adapt.

A Simultaneous Model of Residential Location, Mobility Tool Ownership and Mode Choice Using Latent Variables

“…As the survey is rather comprehensive (total average response time of about 47 minutes for respondents who had moved and 33 minutes for respondents who had not done so), the response rate was limited (between 10.9% and 6.3%, respectively) -also when compared to other studies conducted at the Institute for Transport Planning and Systems (Schmid and Axhausen, 2019b). For this reason, the people that had not replied at all were contacted by phone two weeks after the invitation letters were sent out.…”