Effects of schooling and age on performance in mathematics and science: A between-grade regression discontinuity design applied to Swedish TIMSS 95 data

One purpose of the study is to examine the relative effects of schooling and age on performance in mathematics and science by the use of a between-grade regression discontinuity approach applied to the Swedish samples of TIMSS 95. This design relies on the assumptions that there is a sharp age-based decision rule for grade assignment, and that the regression of performance on age is linear. Another main purpose is to investigate the robustness of the results against deviations from these assumptions. The issue has theoretical, methodological and practical implications, especially on the design of international comparative studies of school performance, in which it is necessary to focus on comparability with respect to either age or to the number of years in school. Even though the design has strong advantages, there are some problems as well. The greatest problem is the selection effect caused by the non-sharp decision rule for school start. According to the Swedish TIMSS 95 data the share of under-aged students is between 0.3 % and 0.7% and of over-aged about 3%. As the Swedish data comprises three successive grade cohorts it is possible to bring together students born a particular year in the analysis, and to estimate the selection effect on the within-year regression coefficients. The main pattern of results agrees with that obtained in previous research, showing that the grade effect is about twice as strong as the age effect. The results indicate that the regression of performance on birth date is linear, and that the selection effect on the within-year regression coefficient is generally relatively small.