An investigation of non-word discrimination and phonological representation in a sample of children with speech sound disorders

Background: Resolution of SSDs without therapy occurs for some children. Identification of those unlikely to resolve spontaneously would facilitate effective targeting of intervention resources. A test of non-word discrimination (NWD) may be effective in gaining a measure of phonological representation while controlling for confounding speech-production variables (Reuterkiӧld-Wagner, Sahlen & Nyman; 2005). Objectives: To find out: 1) what proportion, in a sample of children with SSDs, show deficits in NWD; 2) whether scores on NWD correlate with measures of phonological representation (Real-word Discrimination; RWD); and 3) whether scores on NWD are a useful predictor of which children with SSDs will resolve, with/without therapy. Method: 13 participants with SSDs: 8 boys, 5 girls; aged 3; 02-4;08 (M = 3;10). Participants were recruited through another study on the efficacy of a parent-administered home programme. Participants completed tests of NWD and RWD at study outset. The phonology sub-test of the Diagnostic Evaluation of Articulation and Phonology (DEAP; Dodd et al, 2006) was administered at initial assessment, and again 8 weeks later. Results: Data from 2 children were excluded due to non-completion of the test of NWD. Only 2 children had good NWD scores; both made significant improvement in PCC by Assessment Time 2, despite not receiving therapy. The only other child making significant improvement received therapy. There was no significant correlation between scores of NWD and RWD; however children with good NWD had highest scores in RWD. Conclusions: Findings from this small sample suggest the utility of NWD as a predictor of spontaneous resolution for children with SSDs continues to warrant investigation. Further investigation is needed of the relationship between good NWD and phonological representations of real words.