This study explored the impact of different cut-off points used to identify children with reading difficulties on the distribution of these children across the three poor reader groups predicted by the simple view of reading (dyslexia, specific comprehension difficulty and mixed difficulty). Additionally, the study investigated whether the cognitive profiles of these poor reader groups remained consistent across varying levels of reading impairment.
This study included 209 primary school children from New Zealand, in Years 4–6 and aged 8–10 years, who experienced reading comprehension difficulties. Using a two-step cluster analysis, participants were assigned to one of three poor reader groups: dyslexia, specific comprehension difficulty (SCD) and mixed difficulty. We examined the distribution of children across these groups at eight levels of reading comprehension difficulty and conducted strengths and weaknesses profiles at four levels of reading ability across 14 tests that assessed a range of reading-related skills. Additionally, we compared the performance of children in these groups who performed above and below the 10th percentile on a reading comprehension assessment across these tests.
The results showed that the proportion of children assigned to each group varied across the reading levels, with children at the lower end of the reading comprehension continuum more likely to exhibit the mixed difficulty profile than the SCD and dyslexia profiles. Overall, the research found that dyslexia, SCD and mixed difficulty profiles could be identified at all levels of reading comprehension difficulty, though weaknesses in their reading-related skills increased as reading comprehension difficulties also increased.
Our findings suggest that struggling readers demonstrate distinct cognitive profiles across all levels of reading ability. These results have important implications for identifying and supporting struggling readers, as well as enhancing our understanding of the underlying mechanisms of reading difficulties. The possible relationship between the results obtained in this study and tiered models of reading support is discussed.