| The present study examined the relative importance of general and specific abilities from the WISC-III in predicting reading and mathematics achievement through a series of structural equation models. The study was based on 1,107 children from the WISC-III and WIAT linking sample. Results across analyses for the various structural equation models in predicting reading and mathematics achievement showed that mixed models (i.e., Model 6a for reading achievement and Model 7 for mathematics achievement) were best.; Model 6a was specified to evaluate mixed effects by simultaneously examining the effects of general ability on general reading achievement (i.e., g to Reading) and specific ability on specific reading achievements (i.e., Verbal Comprehension to Basic Reading and Reading Comprehension). Model 7 examined the simultaneous effects of general ability and some of the specific abilities on general mathematics achievement (i.e., g, Verbal Comprehension, Perceptual Organization, and Freedom from Distractibility to Math). Results confirmed that general ability plays an important role in predicting both reading and mathematics achievement. Once the effect of general ability was accounted for, the Verbal Comprehension and Freedom from Distractibility factors made important contributions to the prediction of specific reading achievements and general mathematics achievement, respectively. However, it should be considered whether these factors do make a meaningful contribution in diagnosing the learning disabilities above and beyond the general ability. Multicollinearity problems were discussed in depth to explain “Heywood case” and “empirical underidentification” presented in the current study. |
