Correction for measurement errors in survey research: necessary and possible

Abstract

Survey research is the most frequently used data collection method in Sociology, Political Science, Communication, Opinion research and Marketing (Saris and Gallhofer 2007). Nearly everybody agrees that such data contains serious measurement errors. However, only very few researchers try to correct for these errors. If the measurement errors in the different variables are not the same, the comparison of the sizes of effects of variables on each other will be wrong. If the sizes of the measurement errors are different across countries, cross national comparisons of relationships between variables cannot be made. There is ample evidence for these differences in measurements errors across variables, methods and countries (Alwin 2007, Saris and Gallhofer 2007). Therefore, correction for measurement errors is essential. This correction can be done in a simple way, but it requires that the sizes of the error variances are known for all observed variables. Within the context of the European Social Survey (ESS), an approach has been developed to solve this problem. In each ESS round 4-6 experiments are done in many different European countries and languages. For the questions involved in these experiments the quality has been estimated and the characteristics of the questions have been coded. This allows the study of the relationship between these characteristics and the quality of the questions. Because this relationship is rather good one can also use it to predict the quality of new questions. This makes the necessary correction for measurement error in survey research possible and easy.