Short courses
Two short courses will be offered to conference participants before the conference. The half-day courses are scheduled for Monday afternoon (June 25th.)
Course 1: Hypothesis testing using complex survey data
Lecturer: Peter Lynn, University of Essex
The testing of simple hypotheses using survey data depends on appropriate estimation of standard errors. Estimation under the assumption of simple random sampling is rarely appropriate as samples are often stratified or clustered and may have involved variable selection probabilities. All these features of sample design can affect standard errors and the effect can be large. This short course will explain how sample design affects standard errors, will demonstrate how to take sample design into account when testing hypotheses and will present examples that illustrate why it is important to do so.
Course 2: Coping with measurement error in survey research
Lecturer: Willem Saris, University Ramon Llull, Barcelona
Several authors have shown that survey data contain random and systematic measurement errors. Several studies have also shown that these errors have a considerable effect on the results of survey research. Nevertheless, these errors are most of the time ignored in the analysis. One of the reasons is that there are so many factors that can cause errors and therefore it is very difficult to cope with this problem. In this short course we will show that during the last 25 years knowledge has been collected about causes of measurement error. This information has been summarized in a computer program (SQP) that makes it possible to predict the quality of questions before they have been used in the field. At that moment the questions can still be improved. The program can also be used to correct for measurement error in data which have been collected before. In this course we will give several illustrations of the effects of these corrections.
Additional fee of 15 € is charged for participation (please tick the course you have chosen in the registration form).