Differential validity of reversal errors as predictors of first-grade reading achievement for blacks and whites

The aim was to explore race and sex differences on a measure of reversal and sequencing errors for a group of 401 first-grade children. Differences were examined in two ways: comparisons of mean scores on the Horst Reversals Test, and comparisons of the predictive validity coefficients of this reversals measure. White children scored significantly higher than black children on the reversals test, although predictive validity coefficients were substantial (.50s-.60s) for both races. Sex differences tended to be minor and of little consequence.