Homework assignment and student achievement - Evidence from TIMSS 2019

Abstract: Homework is widely used in most schools across the world to complement classroom teaching, increase knowledge retention, and foster good study habits. The research on the effect of homework assignment on student achievement is, however, inconclusive with earlier research indicating a positive relationship and newer studies giving a more nuanced picture. This thesis uses data from the 2019 wave of the Trends in International Mathematics and Science Study (TIMSS) to examine how teacher-reported homework frequency and homework amount relate to student test scores in 58 countries for grade 4 and 26 countries for grade 8. The TIMSS data is treated as a panel where the effect of homework assignment is identified using the within-student between-subject variation in homework assignment between mathematics and science. Fixed student, class, school, and country characteristics can thereby be controlled for, gaining an estimator that comes significantly closer to a causal estimator than much of the previous research. Results indicate that homework assignment has a null or negative effect on 4th grade students’ test scores and a null effect on 8th grade students’ test scores. The nonlinear relationship hypothesis, with an optimal level of homework assignment above no homework, is not supported by the results, with relationships either being linear or homogeneous compared to receiving no homework. No heterogeneity is found when splitting the sample into OECD and non-OECD countries for grade 8. In grade 4, the estimated negative effect comes from the non-OECD countries, while a null effect is estimated in the OECD countries. Based on the results, education policy makers are urged to not treat increased homework assignment, without considering the quality of the assigned homework, as a way to increase student achievement.