Performance on Geometric Figure Copying Tests as Predictors of Types of Errors in Decoding

Abstract

MANY DIAGNOSTICIANS ASSUME a relationship between visual perception as measured by geometric figure copying tests and types of errors made by poor readers in decoding. The purpose of this study was to determine whether poor readers with high scores on such tests make different types of errors than poor readers with low scores on these tests. If certain types of reading errors, such as confusing b and d occurred more frequently among poor readers with low visual scores than among poor readers with high visual scores, then that finding would lend support to this relationship. Eighty-four second grade poor readers were given 3 experimental tasks based on the letters b, d,p, m ,and n. A multivariate analysis of covariance showed that the low group made types of decoding errors similar to those made by the high group. An exception was that the low group made significantly more errors of letter sequence on a sentence task. The low visual performers contained more of the lowest readers and there was a quantitative but not qualitative difference in errors made.