Want to increase mathematics achievement for all first grade students? Focus on teacher-directed instruction.

That takeaway is from new research by Paul Morgan and Steve Maczuga of Penn State University and George Farkas of the University of California that quantifies the impact of different instructional activities on student performance. While teacher-directed instruction was beneficial for students of all ability levels, the authors found that student-centered activities (group work, solving real-life math, using manipulatives) were only helpful to students already performing well in mathematics, even though those activities were more often used in classrooms with greater numbers of struggling students.

As intriguing as these findings are, they are not without caveats. As the authors note, the classroom activities were self-reported by teachers, leaving room for error in how activities were coded. Also, the study only measured how frequently teachers employed different types of activities, with no measure of how well they were implemented. The data were collected in 1998-1999, which predates the common use of most computers in classrooms; how this technology might affect student-centered learning is unknown. Nevertheless, the findings are by no means irrelevant.