Pitfalls in research on school and teacher effectiveness

Abstract

Many pitfalls in research on school and teacher effectiveness are described. The aspects of effectiveness research considered include the relationship between correlation and causation, statistical significance, practical significance, instrumentation problems, controlling for background factors, regression effects, the measurement of change, non‐linear relationships, aptitude‐treatment interactions and levels of analysis. Despite the many pitfalls in effectiveness research, one conclusion does emerge: differential school effects are relatively modest and unlikely to be clearly related to the national assessment test averages that schools will be required to publish.