First grade children were tested at the beginning of the school year with tasks tapping the phonological loop and central executive components of the working memory system as well as for general processing speed. Four months later, they were administered a single-digit addition task. Addition skills were not predicted by previous measures of processing speed. Addition accuracy was correlated with performance on the central executive task 4 months earlier. But, the best predictor of children's behaviour in the addition task were measures of their phonological loop capacities. In particular, children with high skills in tasks testing the phonological loop were more accurate in addition and used more mature strategies (such as retrieval from long-term memory) than children with low phonological loop skills. Conversely, children with working memory limitations tended to use more frequently the counting-all algorithm and still used their fingers to solve single-digit additions. We interpret these data by arguing that limited working memory resources may prevent the child from using more mature strategies for solving additions.
Noël, M.-P., Seron, X., & Trovarelli, F. (2004). Working memory as a predictor of addition skills and addition strategies in children. Cahiers de Psychologie Cognitive, 22(1), 3-25. https://hdl.handle.net/2078.5/28938 (Original work published 2004)