Providing Cross-National Comparability of Test Results for International Assessment in Higher Engineering Education (Evidence from Russia and China)

Despite the rapid expansion of higher engineering education all over the world in past few years, still relatively little is known about its quality. This trend becomes especially notable in developing countries (like BRIC countries) where higher education in general is extremely popular among population. This paper is aimed to investigate cross-national comparability of mathematics test results obtained in Study for Undergraduate Performance – international project studying quality of higher engineering education and factors affecting it cross-nationally and on a large scale. Since cross-national comparability issues are crucial methodological challenges for any international research, providing it requires from researchers deep understanding of all reasons for possible bias in the data. Moreover, methodology of test development process is sophisticated even more due to the specific field of study (higher education), which has been regarded as an extraordinarily difficult area for assessment. While all reasons for incomparability of test results are eliminated by the project methodology, this study focuses on post hoc procedures – Differential Item Functioning analysis across countries. As a result of study, even although few items exhibit DIF, sufficient number of DIF-free items allows us to construct a common scale and through this provide a comparability of results. Key words: Cross-National Comparability, Differential Item Functioning, Higher Engineering Education, Educational Assessment, Study for Undergraduate Performance.