Abstract
Subgroups of children with reading disabilities were identified by using the regression method introduced by Castles and Coltheart (1993). Children who were poor in phonological, compared to orthographic, word decoding were identified as phonological-type participants, and children who were poor in orthographic, compared to phonological, decoding were identified as surface-type participants. The results replicated previous findings reported that if categorizations are based on comparisons with younger reading-level-matched controls instead of age-matched controls, the number of surface-type children is significantly reduced. Surface-type children performed below the other groups on most cognitive measures and reported that there were fewer books in their homes, and phonological-type children showed a specific deficit in phonological word decoding. The results provided additional support for the hypothesis that the surface type of reading disability can be characterized as a general developmental delay.