Children with dyslexia have a deficit in visual encoding of letter strings, but not in exogenous attention

Abstract

Correctly identifying letters and letter positions within a word is a fundamental process, when learning to read. To measure children’s ability to accurately encode letters within a string, we used a multi-element processing task and manipulated spatial attention via pre-cues to separate the effects of encoding and attention. The experiment was gamified, to make it interesting for children, and was administered on Pavlovia - a web-based platform. A string of 6 letters was presented for 120ms followed by a post-cue in one of the letter positions. Participants were instructed to select the letter that was at the post-cued location from a set of 12 letter choices. We manipulated exogenous attention by presenting uninformative pre-cues that flashed below the three letter positions to the right, left, or both sides of fixation, just before the target letter strings appeared. In a large and diverse sample of N=140 children (ages 6y - 16y) we observed a moderate correlation between reading ability, measured with a standardized assessment, and performance on the multi element encoding task (Pearson’s r = 0.35; p = 0.000022). Specifically, children with dyslexia (n=50) showed reduced ability to encode multiple letters simultaneously compared to typically developing children [Cohen’s d = 0.41]. Examining the effect of attentional cues, we see that the benefit conferred by a valid cue develops with age, but cue benefits show no correlation with reading ability. We find no evidence for a difference in exogenous attention between dyslexic and control participants and, instead, our findings highlight that children with dyslexia struggle to rapidly encode visual information. Our findings shed light on the controversial association between exogenous visual-spatial attention and reading ability.