Web Survey Bibliography
Title Correcting for bias in a self-selected sample using a random sample as auxiliary information: the Dutch Mobility Study
Author Oberski, D.
Source SMABS-EAM Conference 2006
Access date 24.09.2006
Abstract Compared to traditional survey modes, web surveys - as they are usually done - are different in two important respects: the data collection mode on the one hand, and the sampling mechanism on the other. That is, the use of web pages as an interview method, and samples from self-selected access panels as a selection mechanism. Most serious web surveys combine these two aspects, so that in practice it is often impossible to study the effects these features have on the accuracy of the results separately. This study focuses on the bias caused by self-selection in web surveys. It investigates whether the propensity score weighting method can correct for this bias using two datasets from the Dutch Mobility Study (MON). Both datasets measured the same variables, but one was collected from a highly selective panel, while the other is based on a probability sample. This design makes it possible to study sampling effects separately from mode effects, and evaluate whether propensity score weighting can work in a practical application.
Access/Direct link Conference homepage (abstract)
Year of publication2006
Bibliographic typeConferences, workshops, tutorials, presentations
Full text availabilityNon-existant