Utilizing parametric and nonparametric techniques, we assess the role of a heretofore relatively unexplored 'input' in the educational process, homework, on academic achievement. Our results indicate that homework is an important determinant of student test scores. Relative to more standard spending related measures, extra homework has a larger and more significant impact on test scores. However, the effects are not uniform across different subpopulations. Specifically, we find additional homework to be most effective for high and low achievers, which is further confirmed by stochastic dominance analysis. Moreover, the parametric estimates of the educational production function overstate the impact of schooling related inputs. In all estimates, the homework coefficient from the parametric model maps to the upper deciles of the nonparametric coefficient distribution and as a by-product the parametric model understates the percentage of students with negative responses to additional homework.
Change
text size
