Wang, Tianyi; (2025) Exploration of the role of visual attention span in reading and spelling in Mandarin-speaking children and adults. Doctoral thesis (Ph.D), UCL (University College London).

Preview

Text Wang_10212729_Thesis.pdf Download (13MB) | Preview

Abstract

This thesis investigates the role of Visual Attention Span (VAS)—the number of visual elements that can be processed simultaneously—in the development of reading and spelling skills among Mandarin-speaking children and adults. The research is grounded in the Multitrace Memory Model, which distinguishes between global and analytical reading modes, and explores how VAS contributes to literacy acquisition in a logographic language like Chinese. The existing research in VAS has been focused on alphabetic languages including Arabic, Dutch, English, French, German, Greek and Spanish, while few research has explored Chinese. The studies in Chinese also failed to provide a comprehensive picture of the relationship between VAS and Chinese literacy. Thus, this research is to fill in these research gaps. In the first study oral and silent reading fluency were the outcome measures, involving global and partial report tasks and visual 1-back task as VAS measures. In Study 1a participants were 56 Mandarin-speaking children, and in Study 1b participants were 58 Mandarin-speaking adults. Some literacy-related variables were included with age, such as nonverbal ability, receptive vocabulary, phonological awareness, rapid automatised naming, morphological awareness, verbal and visual short-term memory and single character identification skills. Results of hierarchical regression and structural equation modelings revealed that, after controlling for literacy-related variables, VAS significantly predicted oral sentence reading fluency in children and silent sentence reading fluency in children and adults. Examination of the patterns of performance in the VAS assessments in terms of array positions, as an index of visual attention allocation, revealed similar patterns to those in studies carried out in alphabetic writing systems. The position pattern in the global report task showed a left-right asymmetry; the position pattern in the partial report task showed a ‘w-shape’; the patterns in visual 1-back accuracy and d prime showed a reverse ‘V’ shape or an inverted ‘U’ shape; the position pattern in visual 1-back correct reaction times showed a ‘v’ shape. In Study 2 under both the quantitative and qualitative methods, 60-word spelling-to-dictation task was the outcome measure, involving the same VAS tasks and literacy-related assessments as Study 1. Participants were the same as Study 1. Results revealed that VAS (especially global report performance) was a unique predictor of spelling accuracy in both children and adults. Children’s spelling was more influenced by word length (analytical strategy), while adults’ spelling was more influenced by word frequency (global strategy). Compared with poor adult spellers, better adult spellers relied more on global and whole-word processing route. Poor adult spellers showed VAS deficits but not phonological deficits, suggesting a distinct cognitive profile. Previous studies with poor and good adult spellers in alphabetic writing systems revealed that they could be distinguished by letter report performance. Analyses conducted with the data from Study 2b revealed a comparable result. Child and adult participants who were identified as having poor literacy ability (n=9, six children and three adults) took part in Study 3, in which the effectiveness of three types of training was assessed. One was whole-word practice (WWT) involving global orthographic processing, a second was the retention of symbol arrays (V1BT) involving pure visual global processing and attention, and a third was phonologically-focused (NVT) involving pure phonological processing. The outcome variables were reading and VAS performance, plus spelling-to dictation. Case series analyses and group analyses showed that WWT and V1BT programmes produced significantly better post-test outcomes than NVT in improving reading fluency and VAS performances. Following WWT generalisation of spelling improvement to untrained words was found. All participants improved in at least one of the reading measures at post-test. VAS-based interventions were thus effective across age groups and reading profiles. The results of the studies are discussed in terms of theories of reading, and their potential implications for education are considered. In terms of theoretical implications, these studies support the independence of VAS from phonological processing, highlight the development of VAS and its increasing role in literacy with age growth, and suggest VAS may function similarly to Orthographic Working Memory (OWM) in spelling. In terms of educational and clinical implications, VAS assessments can aid in identifying dyslexia and dysgraphia in Chinese speakers. VAS-based interventions offer a promising route for remediation of literacy difficulties. Finally, these findings are applicable across languages, especially those with opaque orthographies. As for conclusions, this thesis provides robust evidence that VAS is a critical cognitive mechanism in reading and spelling development in Mandarin. It challenges the traditional phonological deficit model of dyslexia and offers new directions for assessment and intervention, particularly in non-alphabetic languages. The research bridges cognitive theory and educational practice, emphasising the need for language-specific literacy models and targeted support strategies.

Download activity - last month

Export as JSONXMLCSV

Download activity - last 12 months

Export as JSONCSVXML

Downloads by country - last 12 months

Export as XMLCSVJSON

Archive Staff Only

View Item