Purpose: This study explored associations between working memory and language in children aged 2-4 years. Method: Seventy-seven children aged 24-30 months were assessed on tests measuring language, visual cognition, verbal working memory (VWM), phonological short-term memory (PSTM), and processing speed. A standardized test of receptive and expressive language was used as the outcomes measure 18 months later. Results: There were moderate-to-strong longitudinal bivariate relationships between the 3 processing measures and language outcomes. Early VWM showed the strongest bivariate relationship with both later expressive (r = .71) and receptive language (r = .72). In a hierarchical multiple regression analysis, adding early VWM, PSTM, and processing speed improved prediction of receptive and expressive language outcomes (12%-13% additional variance) compared with models consisting only of early receptive or expressive language, parent education, and age. Conclusions: Unique associations in hierarchical regression analyses were demonstrated between VWM at age two years and receptive and expressive language skills at age four, and between early processing speed and later receptive language. However, early PSTM did not predict unique variance in language outcomes, as it shared variance with other measures.
Purpose: This study explored associations between working memory and language in children aged 2-4 years. Method: Seventy-seven children aged 24-30 months were assessed on tests measuring language, visual cognition, verbal working memory (VWM), phonological short-term memory (PSTM), and processing speed. A standardized test of receptive and expressive language was used as the outcomes measure 18 months later. Results: There were moderate-to-strong longitudinal bivariate relationships between the 3 processing measures and language outcomes. Early VWM showed the strongest bivariate relationship with both later expressive (r = .71) and receptive language (r = .72). In a hierarchical multiple regression analysis, adding early VWM, PSTM, and processing speed improved prediction of receptive and expressive language outcomes (12%-13% additional variance) compared with models consisting only of early receptive or expressive language, parent education, and age. Conclusions: Unique associations in hierarchical regression analyses were demonstrated between VWM at age two years and receptive and expressive language skills at age four, and between early processing speed and later receptive language. However, early PSTM did not predict unique variance in language outcomes, as it shared variance with other measures.