Blame the data or blame the theory? : On(in-)comparability in international survey research

This dissertation project contains three articles dealing with particular examples of cross-national (in)comparability in international survey research, addressing different elements of the Total Survey Error framework. The central focus of the articles is on measurement error and construct validity. By examining the relationship between socioeconomic status and attitudes toward progressive taxation, the first article reveals theory generalization limitations across countries with different welfare state designs. Dealing with the latent construct of religiosity, the second article discusses the cross-national comparability of religious profiles (typologies) and the role denominations play for the profile formation process. The third article evaluates an indicator of religious self-assessment by comparing its predictive potential with two multi-variate survey instruments for religiosity across countries. All articles are based on data taken from the International Social Survey Programme (ISSP)