Notice: the WebSM website has not been updated since the beginning of 2018.

Web Survey Bibliography

Title “I think I know what you did last summer” Improving data quality in panel surveys
Year 2012
Access date 29.01.2013

In the following five chapters, several methodological innovations in panel surveys are evaluated. In each chapter, one of the methods discussed above to study and correct for measurement errors will be used to study how these methodological innovations affect survey errors and/or substantive conclusions derived from these survey data. The techniques discussed in the different chapters all build on one or more of the basic methods, but describe and explore the techniques in far more detail. In Chapter 3, the technique of propensity score matching is used to study the effects a mixed‐mode respondent recruitment strategy for a survey. It shows how matching can be used to separate nonresponse error from measurement error in a mixed telephone and Internet survey. Separating the two enables us to study how differences between the samples that remain after correcting for nonresponse error persist: the mode effect. In Chapter 4, we turn to the technique of Dependent Interviewing (DI). Different versions of DI are experimentally compared and evaluated using a quasi‐simplex model. This chapter shows how DI and the extent of measurement error present in a survey question on income affects the reliability coefficient. Chapter 5 further explores the use of Dependent Interviewing in panel surveys. This chapter focuses on the effect DI has on substantive estimates that use income questions. Apart from this, details of a validation study using the same income questions shed light on how DI works to affect survey estimates. Chapter 6 focuses on the topic of change in attitude question in a population that experiences a period of life changes. A mixed‐method study that combines longitudinal survey data with qualitative interviews shows how attitudes change over time. Not only do levels of attitudes towards their study change among a group of first year psychology students, the concept of interest itself also changes. The chapter shows how the meaning of study motivation for students itself changes over time. The final chapter focuses on panel attrition. Recent advances in mixture Structural Equation Modeling are used to describe the process of attrition in a panel study with monthly measurements. The chapter shows how different archetypes of respondents drop out of a study in different ways and for different reasons. This chapter concludes by showing how every group of attriters affects longitudinal nonresponse error in a different way.

Access/Direct link

Igitur Homepage (abstract) / (full text)

Year of publication2012
Bibliographic typeThesis, diplomas