The National Reading Panel’s Meta‐Analysis of Phonics Instruction: Another Point of View

Abstract

The purpose of this article is to provide an alternative interpretation of the National Reading Panel’s meta‐analysis on phonics instruction. The NRP based its conclusion that phonics instruction was superior to other approaches on the interpretation of mean effect d statistics reported in its meta‐analysis. In addition to looking at these d’s, we computed and interpreted the mean effect correlations (r’s) and r‐squares that corresponded to the reported d’s. Our conclusions were different from those reported by the NRP. Instead of phonics approaches being superior to nonphonics approaches (as indicated when statistically significant d’s are the only criterion considered), we argue that the advantages of phonics instruction relative to nonphonics instruction have not been demonstrated clearly (as indicated when practical significance r’s and r‐squares are considered).