Combination Classrooms Revisited

In their box score review, Mason and Burns (1997) conclude that while comparisons of combination and single‐grade classes show no general differences in student achievement, combination classes nevertheless have a small negative effect. Based on the results of a best‐evidence synthesis (Veenman, 1995) and a meta‐analysis of an extensive and international research base (Veenman, 1996), I argue that there is no empirical evidence at this moment showing student learning to suffer in combination classes. Mason and Burns also conclude that the negative instructional effects characteristic of combination classes are compensated for by placing better students and teachers in combination classes. On the basis of my best‐evidence synthesis and meta‐analysis, however, I can find no indication of the selection bias suggested by Mason and Burns.