Thalamus, Summer 2000, 18(2), pp. 1-18

Seventy-four children who had been identified as exhibiting the characteristic profile of phonological dyslexia were subdivided according to their relative skill status on a checklist of skills developed to operationalize the phase theory of reading (decoding) acquisition (after Frith, 1985; 1986; and Ehri, 1994). Performance on measures of phonological processing and word recognition were compared for students who skills clustered at the alphabetic phase (AP) with those whose skills clustered at the early orthographic phase (OP). Both groups were also compared to a sample of average readers (AR) in grades 3-7. Group mean differences and discriminant analysis confirmed significant differences among the groups. Regression analysis was used to examine the relative contribution of the decoding variables and spelling (as another index of phonological processing) to reading comprehension performance within each group. Students in the AR group presented with a pattern of decoding strategies that more closely approximated that of the AR group, suggesting a delay in the development of both phonological and orthographic/morphemic abilities. In contrast, the OP group presented with a pattern that was more suggestive of disordered development, perhaps at the level of cognitive interface of phonological processes with orthographic/morphemic representations. Compensatory mechanisms developed among students in the OP group who were, on average, older and more advanced in school might explain this finding. Differences among the two subgroups of phonological dyslexics did not comfortably fit the current operational distinction for the surface and deep dyslexia subtypes. Differences in criteria used to segregrate the subgroups is suggested as a possible explanation for the unexpected outcomes.