[PS-1.21] An examination of working memory processing across modalities in developmental dyslexia

Ahmad, J. , Ferguson, H. & Bowman, H.

The University of Kent

The presentation includes 4 experiments aiming to explore working memory processing in developmental dyslexia, across modalities. The current set of studies utilise the N-back task. The first two studies adopted visual presentation, and letters (experiment 1) or objects (experiment 2) were presented with a non-target: target ratio of 2:1. The first experiment demonstrated a main effect of group upon WM processing, whereby dyslexic participants had significantly fewer hits at each level of N. However, their correct reject rate was comparable to controls. This was reflected in signal detection theory, whereby dyslexic participants adopted a significantly higher criterion than controls. Dyslexic participants used strategic responses in order to maximise overall accuracy during the task. In experiment 2, which did not involve phonological processing, these between group effects were not present. In the second set of experiments, auditory working memory was assessed using the N back task, for letters (experiment 3), and words (experiment 4), which were manipulated by Age of Acquisition (AoA). Critically, in experiments 3 and 4, the non-target: target ratio was 1:1. Experiment 1 revealed a main effect of load upon performance, alongside a main effect of group, and an interaction between N*group. Thus, as the demand upon WM increase, group effects emerged. Unlike experiments 1 and 2, this group differences was driven by non-target trials, with dyslexic individuals displaying increased false alarms. Experiment 4 demonstrated an effect of WM load upon accuracy, however there was no interaction between N * Group. Furthermore we replicate experiment 3, with an increased false alarm rate amongst dyslexics. AoA played a crucial role in reaction times (RTs), with shorter RTs for earlier learned words. With a 1:1 target ratio, it was not possible for dyslexic participants to rely strategic responses. Results will be presented in terms of accuracy, signal-detection and EEG analysis.