During the POLPAN Seminar on October 24th Professor Zbigniew Sawinski (Institute of Philosophy and Sociology of the Polish Academy of Sciences) presented a talk entitled “Harmonization without a harmonized variable: The case of education.”
When countries differ in information loss during the harmonization process, the results of harmonization may lead to invalid conclusions about importance of the harmonized variable within countries. The European Social Survey has introduced a three-step approach to the harmonization of education. It begins with the collection of data on the specific types of educational institutions in each country. In the second step, these data are encoded into the specific categories of the International Standard Classification of Education (ISCED), and in the last step the detailed ISCED categories are grouped into one common classification that can be considered relevant to all countries participating in the project. In this paper, using the ESS-7 data it was demonstrated that countries significantly differ in information loss in each step of this harmonization approach when values and attitudes are taken as outputs of education. The question arises whether more efforts should be made towards creating a common variable with more equal information loss, or whether further efforts should focus on creating country-specific ‘harmonization’ variables which would better reflect the outputs of different types of educational institutions.