Phonological processing skills of children who have learning disabilities in reading decoding with and without language impairments and reading- and age-matches

This investigation explored the relationship of phonological processing, reading disabilities, and semantic and syntactic language impairments. Four groups of children were compared on nine phonological processing tasks measuring phonological awareness, working memory, and the retrieval of phonological codes from memory (naming), as well as reading comprehension. Fourth- and fifth-graders with learning disabilities in reading decoding were compared to a group matched for grade and age, who had learning disabilities in reading decoding as well as semantic and syntactic language impairments. These two groups were compared to reading-matched and age-matched groups.;Children who had both learning disabilities in reading and semantic and syntactic language difficulties demonstrated quantitative and qualitative differences in working memory and naming from the children with learning disabilities without language impairments and from reading-matches. The children with learning disabilities in reading without language impairments scored similarly to younger, reading-matches. However, this learning disabled group showed relatively lower performance than the reading-matches (not significant) on tasks that required processing of nonwords or partial words.;Children having learning disabilities with and without language impairments and reading-matches all scored similarly on the phonological awareness tasks. Therefore, higher-level language difficulties occurred despite relative strengths in phonemic segmentation at lower-levels of language functioning which does not support a hierarchical model of language processing. These results were interpreted as consistent with interactive reading models and connectionist theory.;Hierarchical regressions revealed that phonological processing tasks all shared variance in language, reading decoding, and reading comprehension performance. Unique variance in reading decoding was accounted for by the Rosner Auditory Analysis Test and naming tasks. Working memory accounted for unique variance only when the Rosner was removed from the regression equations. It was concluded that some tasks in this study assessed overlapping phonological processes. Phonological processing skills appear complex and may best be considered separately rather than grouped as single construct.;The results suggest that the language basis of reading includes various types of phonological processing as well as semantics, and syntax. This has implications for increased collaboration of speech-language, reading, and learning disability professionals towards assessment and treatment involving multiple levels of oral and written language.