Cross-modal working memory binding and learning of visual-phonological associations in children with reading difficulties.

Reading decoding requires the association of a visual input (the printed word) with a verbal output (the spoken word), and the learning of visual-phonological associations via repeated exposure occurs both explicitly (by instruction and training) and implicitly (by exposure to written material). However, a general ability to implicitly learn visual-phonological associations (or cross-modal bindings) in children with reading difficulties (RD) has not been deeply explored. The present paper examined this issue in two studies comparing groups of children with RD with matched groups of control children in a working memory binding task involving sequences of variable combinations of nonsense shapes and nonwords (odd trials), as well as fixed combinations (even trials). Stimuli presentation was followed by a recognition test in which the nonwords were given one at a time and the children were required to mouse click on their respective shapes. In Study 1, fixed bindings were presented in random sequences across even trials, and the recognition test presented the nonwords at random. In Study 2, fixed bindings were presented in the same order across even trials, and the recognition test presented the nonwords following the same sequence order. In both studies, a consistent and strong learning effect of fixed bindings was observed for the control groups, with significantly better performance than the groups with RD. Such results suggest that poor cross-modal binding learning is associated with reading difficulties. Implications for the study and treatment of dyslexia are discussed.