| Jamie Shkolnik
University of California, San Diego
Ability grouping in secondary schools; The tradeoff between efficiency and peer group effects
FINAL REPORT: Past studies of education production find no conclusive evidence that school inputs such as class size, teacher experience, and teacher education affect student achievement. This dissertation examines allocation of resources and students within schools and then examines how schools could increase output via more efficient allocation. The first chapter shows that tracking, where students are placed into classes based on some measure of initial ability, has little effect on average math achievement growth after controlling for selectivity. This result is explained by the finding that the allocation of students and resources among classes is remarkably similar between tracking and non-tracking schools; tracking appears to occur in virtually all schools. If most schools track, it is important to determine what would happen if schools were to discontinue this practice. Chapter two shows that several school inputs have different marginal productivities for students of varying ability levels, indicating potential efficiency gains for all students if classes are tailored to suit group propensities. It then finds evidence for the existence of peer group externalities, where an individual's achievement is positively related to the ability level of his class. Ability grouping increases achievement for top students and decreases achievement for bottom students. The gains to students from efficient resource allocation among classrooms, however, are small compared to the effects of being in a high ability class. The choice between tracking and mixing will depend on the school's social welfare function, as demonstrated by a theoretical model of the school's optimization problem. The third chapter studies how a teacher's allocation of time affects student learning and how class size affects time allocation and student achievement. After controlling for the confounding effects of class ability, class size has a significant negative effect on all but the below average students. In addition, increases in class size affect achievement indirectly because teachers compensate by allocating more time to group instruction in place of the more expensive and less productive individual instruction. A decrease in class size would have an even larger effect on student achievement if teachers did not react by decreasing time spent on group instruction.
Back to Funded Dissertation Grants Page |