Identifying Influential Contextual Factors of Online Reading Literacy Through a Machine Learning Approach

• In Event: AERA Poster Session 3
  In Poster Session: Concepts and Considerations in Literacy Education

Abstract

Few of previous reading studies examined systematically the contributing factors of students’ online reading literacy. To fill this gap, based upon the ecological perspective and the opportunity-propensity model, this study aims to investigate student, home and school contextual factors that are influential for predicting students’ online reading literacy. The data of the Progress in International Reading Literacy Study 2016 with 73,962 Grade 4 students from 14 countries and economies was analyzed with a machine learning approach, support vector machine with recursive feature elimination. Results showed that reading self-efficacy, parents and family supports for learning, school instruction, computer access in school and school climate are the strongest predictors of top-performing online readers.

Authors

• Fu Chen, University of Macau

• Alfred Sakyi, Government of Alberta

• Ying Cui, University of Alberta