Prediction of the quality of survey questions in cross cultural research

Daniel Oberski

Project:  NWO project at Tilburg University / University Ramon Llul, Barcelona

Project running from:  1 October 2006 – 1 October 2011

Supervisors:  Prof. J.A.P. Hagenaars (Tilburg University), Prof. W. Saris (Univ. Ramon Llull, Barcelona)

Summary:
Without correction for measurement error, comparative survey research is not possible. During the last 15 years research was done to develop a scientific approach to predict and improve the quality of survey questions. This led to the development of an application programme, SQP, that predicts the quality of survey questions for 3 languages (Dutch, English and German). Now in the European Social Survey (ESS), 20 multitrait-multimethod experiments have been done in more than 20 languages. This huge database offers a tremendous and incomparable opportunity to further investigate the quality of survey questions and the development of the SQP programme for 20 languages. With this programme measurement errors in survey questions can then be investigated and predicted which is essential for comparative survey research in Europe.

Date of defence: 28 January 2011

Title of thesis: Measurement error in comparative surveys