Exploring the utility of verbal and visuospatial working memory for identifying children with language impairment

Objectives: In the clinical setting, it is an important issue to find more efficient measures to identify language impairment equivalent to standardized language tests can be a useful tool in identification of language impairment. This study examined whether a combination of verbal and visuospatial working memory tasks can discriminate children with language impairment from those with typical language development. Method: A total of 36 children participated-18 children with language impairment (LI) and 18 typically developing (TD) children. Participants completed verbal working memory tasks (non-word repetition [NWR], sentence repetition [SR], Competing Language Processing Task [CLPT]), visuospatial working memory task (Matrix). Multivariate analysis, Pearson's product-moment correlation and discriminant analysis were used. Results: MANOVA results indicated that group differences for SR, CLPT scores were significant. Discriminant analysis conducted with all measures (NWR, SR, CLPT, Matrix) showed good sensitivity (83.3%) and specificity (94.4%) in distinguishing between TD and LI groups. Conclusion: Children with LI exhibited deficits in verbal tasks (SR, CLPT) rather than visuospatial task (Matrix) relatively. However, visuospatial ability was not correlated any standardized language test scores in LI group although their performance was not behind. Nevertheless, a combination of various working memory tasks can be a useful tool in discrimination of children with LI from their TD peers.