Phonological decoding ability, spatial attention, and event-related potentials

Evidence for a selective spatial attention deficit among children and adults with developmental dyslexia has been interpreted to reflect impairment in the posterior attentional network including the magnocellular-mediated posterior parietal cortex, and particularly the right hemisphere. According to cognitive interpretations, dyslexia stems from a core phonological deficit and it has been shown that phonological decoding is essential for normal reading development. Thus, the aim of the present research was to investigate the relationship between selective spatial attention and reading ability among adults with good and poor phonological decoding ability as assessed by nonword reading ability. Five experiments were conducted to investigate this relationship. The results of Experiment 1 indicated that poor phonological decoders are slower to search for feature conjunctions, particularly for searches defined by the features of form and motion. These findings are consistent with previous research in populations with dyslexia and suggest differences between good and poor phonological decoders in terms of the guidance of spatial attention. Experiments 2 to 5 aimed to examine differences in attentional processing in more detail by examining the specific mechanisms involved in both sustained and transient spatial attention tasks and their relationship to early event-related potential components. Research has demonstrated that the early sensory components of the event-related potential waveform (P1 and N1) index early visual processing and are modulated during tasks in which attention is manipulated. However few previous studies have investigated the electrophysiological correlates of spatial attention in good and poor readers. In Experiment 2 the mechanisms of attentional focussing and inhibition were investigated using a task in which a focussing cue preceded a target that was flanked by compatible or incompatible flankers. Poor decoders showed a greater reaction time cost for incongruent stimuli preceded by large cues, which was suggested to indicate difficulty in focussing attention and suppressing information at unattended locations. This finding was accompanied by a reduction in the modulation of N1 amplitude by both cue-size and flanker compatibility for poor decoders, and a reduction in the modulation of the frontal N2 component by flanker compatibility. Together, these findings suggested hemispheric differences in the functioning of the posterior attentional network as well as differences in inhibitory processing within the frontal attentional network. Experiment 3 aimed to examine differences between good and poor phonological decoders in the allocation of attention to global and local levels of hierarchical stimuli. Poor phonological decoders were slower than good phonological decoders when attention was directed to both the global and local processing levels. This was accompanied by a lack of task related modulation of the posterior N1 and N2 components, and an overall increase in N2 amplitude among poor decoders. Together, these findings suggested differences in the early allocation of spatial attention and compensatory processing at later perceptual stages. Dyslexia has also been associated with performance differences on covert orienting tasks involving valid and invalid spatial cues. These differences are often greater for stimuli presented in the left visual field which is suggestive of a right hemisphere parietal deficit. The aim of Experiment 4 was to investigate covert orienting in good and poor phonological decoders. Poor phonological decoders showed fewer reaction time benefits of valid spatial cues relative to good decoders, particularly for left visual field trials. This effect was greatest for male phonological decoders, who also showed a lack of N1 modulation in the right hemisphere for left visual field trials and an overall lack of attentional modulation of N1 latency. In comparison, female poor decoders showed a greater involvement of the right hemisphere which may reflect compensatory processing due to a left hemisphere deficit. The aim of Experiment 5 was to investigate the effect of valid and invalid spatial cues on the performance of orthographic and phonological decision tasks. Consistent with early selection models of attention in word recognition, good decoders showed consistent behavioural effects such that reaction time benefits were observed when words were preceded by valid spatial cues. However, poor phonological decoders showed fewer reaction time benefits for words preceded by valid spatial cues, particularly for words presented in the left visual field when processing was biased towards phonological processing. These behavioural differences were accompanied by an absence of the attentional modulation of both PI and N1 latency in poor decoders, and this was explained by differences in early perceptual and attentional processing in the posterior attentional network. Together, the findings of the present series of experiments provide evidence that the spatial attention difficulties observed in developmental dyslexia are also observed in adults who are poor phonological decoders. The poor phonological decoding group generally showed less attentional modulation of the early posterior N1 and P1 components which is consistent with differences in the functioning of the posterior attentional network. The findings of the present research are broadly consistent with the proposal that the phonological decoding deficits observed in developmental dyslexia are associated with attentional processing differences in the posterior parietal cortex. This research also provided preliminary evidence for sex differences in the lateralisat ion of ERP components which require further investigation.