Meta-analysis was used to integrate statistically the literature assessing the relationship between auditory and visual perception and reading achievement. From 267 studies conducted between 1950-1980, 2,294 correlation coefficients were collected and aggregated into homogeneous groupings across five auditory perceptual skills, eight visual perceptual skills, four reading skills, and two subject groups. Findings were interpreted through the binomial effect size display, which indicates the increase in predictive accuracy rather than the percent of variance explained (r(2)). Stepwise multiple-regression analyses were used to order perceptual skills in terms of their usefulness for predicting reading skills and to examine the role of intelligence in prediction. The findings indicated that auditory and visual perceptual skills can successfully increase the accuracy of predicting reading achievement, but the magnitude of the increases in predictive accuracy was contingent upon the combination of variables studied and was significantly reduced if an IQ score was known. It was concluded that, while there was some justification for early conceptualizations of learning disability emphasizing perception, the limitations surrounding the magnitude and nature of the relationship between perceptual skills and reading as well as recent advances showing other processes holding greater promise fbr explaining reading disability, perceptual processes no longer need to be considered primary factors in predicting reading ability.
