The Cognitive and Psycholinguistic Predictors of Reading Comprehension: a case of British children

• In Event: Poster Session 02
  In Poster Session: PS 02 Topic - Language, Communication

Abstract

Background. Reading comprehension is an essential skill in literacy acquisition as reading enables one to understand written text and to gain meaning out of it. Extensive literature has examined how different factors predicted reading comprehension ability in English (e.g. Blything et al., 2015; Cain & Oakhill, 1996), but to the best of the author’s knowledge, only a few studies had considered both the cognitive and psycholinguistic aspects of reading comprehension. To bridge the gap, this study aimed to investigate the interplay of psycholinguistic and cognitive factors in predicting reading comprehension. Psycholinguistic factors were considered at word-, sentence-, and paragraph-level, while cognitive factors were operationalised as executive functions and meta-memory. These variables have been found to predict individual differences in reading comprehension in the literature (e.g. Arrington, Kulesz, Francis, Fletcher, and Barnes, 2014; Cain, 1999; Cain, Oakhill, & Bryant, 2004; Muter et al., 2004; Nation, Cocksey, Taylor, & Bishop, 2010; Ricketts et al., 2007), but the pattern and strength of their contribution were less clear.

Method. Primary school children in Britain were recruited to participate in the study, and a total of 243 participants aged between 9.51 and 11.51 (M = 10.37, SD = 0.51) took part. The participants completed tasks measuring their reading comprehension, vocabulary, syntactic awareness, discourse coherence, executive functions (cognitive flexibility, inhibition, planning, working memory), meta-memory (prospective bias, retrospective bias, prospective absolute accuracy, retrospective absolute accuracy), short-term memory, and non-verbal reasoning. Missing data was imputed using multivariate imputation by chained equation.

Hypotheses. Three hypotheses were proposed: 1) Both the cognitive and psycholinguistic predictors will predict individual differences in reading comprehension directly; 2) There will be no moderation effect of the predictors, but 3) there will be a mediation effect, with proficient executive functions and metacognition being necessary for skilful vocabulary, syntactic awareness and inference-making ability, which lead to better performances on reading comprehension.

Data Analysis Plan. To test the first hypothesis, a hierarchical regression will be conducted. The variables will be entered in the following steps: Age, Control Variables (non-verbal reasoning and short-term memory), Cognitive Predictors, and Psycholinguistic Predictors. The second hypothesis will be investigated by adding the interaction of Cognitive and Psycholinguistic predictors as the last step to the aforementioned model. Lastly, the third hypothesis will be examined using Structural Equation Modelling. The factor structure of executive functions and meta-memory will first be validated using exploratory and confirmatory factor analysis, then a full structural equation model including the other variables will be run.

Implications. The results might help initiate progress in the understanding of proficient comprehension and motivate the restructuring of existing teaching materials to address those different factors which contribute to the comprehension abilities in different languages. Furthermore, the results might also provide insight into the intervention programmes of reading difficulties.

Feasibility of Completion. Data collection has been completed. The data has been entered, screened, and processed. Hierarchical regression model has been created for the first hypothesis, and the data analyses for the remaining two hypotheses should be completed within a month.

Authors

• Claudia Pik-Ki Chu, University of Cambridge
  Presenting Author

• Michelle Ellefson, University of Cambridge
  Non-Presenting Author