Teacher Characteristics and Student Performance at School. Applying the First Difference Method to TIMSS 2007 Data

Abstract

Yuliya Tyumeneva, Psy.D., senior researcher at the Center for Monitoring Educational Quality, National Research University — Higher School of Economics, Moscow, Russian Federation. Email: jtiumeneva@hse.ru 

Tatyana Khavenson, junior researcher at the International Laboratory for Education Policy Analysis, National Research University — Higher School of Economics, Moscow, Russian Federation. Email: tkhavenson@hse.ru 

The Trends in International Mathematics and Science Study (TIMSS) does have a clear advantage in that, apart from a direct cognitive appraisal of students, it enables to collect information on their teachers, including their education background, work experience and teaching practices. The first difference method was used to determine how teacher characteristics were correlated with student achievements and to overcome restrictions of TIMSS correlation design. In addition, effects of teacher characteristics were evaluated by the conventional regression analysis. The discovered correlations differed across subject areas, and the results of the first difference method differed from those of the conventional correlation analysis. In mathematics, the first difference method revealed negative relation between academic progress and rule-based tasks or collaborative learning, while tasks designed to develop meta-subject skills showed a positive effect on performance. In natural sciences, on the contrary, rule-based tasks showed positive relation, while tasks designed to develop meta-subject skills produced a negative effect or no effect at all. Unlike in mathematics, teacher experience also had a significant influence on student achievements in natural sciences.

Implementing the new methodological approach has allowed not only to determine relation between teacher characteristics and academic performance of students but also to identify characteristics typical of specific teacher categories, as, for example, those who work in elite schools or teach better-prepared students. This information may be used both in developing the educational policy and in designing future studies, especially experimental ones.