The Problem of Self‐Selection

Abstract

This chapter focuses on the estimation problems caused by self-selection. After an introduction, some theory is described. The chapter explores whether weighting adjustment techniques can help to solve the problem. Practical implications are shown using simulated samples from a fictitious population. Web surveys can be based on probability sampling. What happens if a self-selection sample is selected from the Internet population? The chapter shows that estimators can be substantially biased and also that this bias can be larger than the bias caused by nonresponse in surveys based on probability samples. Two applications of self selection are discussed in this chapter. The first one is an experiment in which self-selection is simulated. Since the target population is completely known, the effects of self-selection can be shown. The second application describes an experiment that was carried out in the town of Alphen a/d Rijn in the Netherlands.

Authors 2

1. Silvia Biffignandi

   University of Bergamo

   Affiliation as printed

   University of Bergamo, Italy

2. Jelke G. Bethlehem Aachen Faculty of Social and Behavioral Sciences Institute of Political Science

   Leiden University

   Affiliation as printed

   Faculty of Social and Behavioral Sciences, Institute of Political Science, Leiden University, The Netherlands