Rothwell, Charlotte and Hartley, Calum and Westermann, Gert (2023) Investigating the role of attentional biases on language development in autistic children and neurotypical children. PhD thesis, Lancaster University.
![[thumbnail of 2023RothwellPhD]](https://eprints.lancs.ac.uk/style/images/fileicons/text.png)
Abstract
From their first year of life, neurotypical children learn words rapidly and with apparent ease. However, many autistic children experience delays in language development and struggle to acquire vocabulary. To learn a word, children must identify the intended referent (referent selection) and then store the word-referent association in memory for later retrieval (retention). In light of recent evidence that referent selection and retention mechanisms are unimpaired in autism, difficulties acquiring vocabulary may be attributed to differences in attention to stimuli. Autistic children often focus intensely on preferred stimulus categories, potentially disregarding stimuli that do not align with their preferential interests. Therefore, the purpose of this thesis was to examine how attentional salience of stimuli influences autistic and neurotypical children’s novel word learning. The three studies that comprise this thesis investigated how children’s stimulus preferences and interests influence their referent selection and retention after delays of 5 minutes and 24 hours. Participants were neurotypical and autistic children with delayed language development matched on receptive vocabulary (age equivalent ~5 years). All studies utilised touch-screen technology to capture children’s response times and accuracy. Studies 1 and 2 examined how children’s word learning is impacted by the presence of target and distractor stimuli related to a preferred stimulus category. ‘High interest’ stimuli in these studies were novel animals, as they are often favoured by both neurotypical and autistic children. Participants’ interest in animals was confirmed via a questionnaire. Study 3 investigated how children’s individual preferences for novel objects (i.e. generic experimental stimuli) affect novel word learning. Studies 2 and 3 employed a novel combination of accuracy and looking measures, exploring children’s in-trial visual attention and how this predicted children’s response accuracy. In Study 1, autistic and neurotypical children were taught novel words associated with unfamiliar animals (high interest stimuli) and objects (neutral interest stimuli) in the presence of high interest animal distractors during referent selection. This manipulation enabled us to investigate how salient non-target stimuli affect novel word learning. The presence of interesting distractor stimuli enhanced autistic children’s encoding of novel word-referent associations by increasing attention across the whole visual scene. Study 2 studied how children’s categorical interests influenced their word learning. Children learnt names for novel animals and novel objects, but the distractor stimuli presented at referent selection were neutral interest familiar objects, allowing us to identify how perceptual distinctions between stimulus categories influences novel word-referent encoding. Categorically interesting novel stimuli, that stood out from less interesting distractors, enhanced autistic children’s retention of novel words after 5 minutes. However, autistic children demonstrated a substantial increase in their retention of novel object labels between 5 minutes and 24 hours, outperforming neurotypical peers in both conditions following sleep consolidation. Study 3 examined how children’s individual preferences for novel objects impacts word learning (labels were taught for liked and disliked objects, identified by children in a preceding task). The findings showed that idiosyncratic preferences for novel experimental stimuli, an object category that is unlikely to align with children’s pre-existing interests, do not affect autistic or neurotypical children’s attention allocation to an extent that is sufficient to impact retention after 5 minutes or 24 hours. Our looking data in Studies 2 and 3 indicated that greater visual attention towards stimuli predicted novel word retention. Moreover, observed population differences in visual attention did not directly translate to diminished behavioural accuracy. This thesis advances understanding of relationships between word learning processes and attentional mechanisms in autism and neurotypical development. Our data demonstrate that categorically interesting stimuli aid novel word learning in autistic children. This finding, coupled with observed relationships between visual input at referent selection and novel word retention, confirm that increased attentional allocation to stimuli during encoding leads to more robust novel word-referent representations. Contrasts between population differences in visual attention and response accuracy outcomes suggest that autistic children potentially achieve similar results through a different route than neurotypical peers. Our novel combination of visual attention and accuracy measures highlight the importance of choosing dependent variables that appropriately target mechanisms of interest and the risk of drawing inaccurate conclusions based on looking behaviour alone. Given that stimulus categories influenced autistic and neurotypical children’s learning to different extents, researchers should carefully select experimental stimuli that appeal to different populations comparably. Overall, our findings have the potential to inform practice in clinical and educational settings, providing insight into how learning contexts can be optimised for maximally efficient language acquisition.