Computational Thinking in Reading Literacy: How to Develop and How to Assess Using LLMs (Poster 10)

• In Event: AERA Poster Session Wednesday 10:50 a.m.
  In Poster Session: Literacy Innovations Across the PK-20 Spectrum

Abstract

Computational thinking (CT) as one of the core 21st century digital skills, is not limited to Science, Technology, Engineering, and Mathematics (STEM) areas. In social and humanities disciplines, we developed a three-dimensional CT reading literacy framework based on Marzano’s new taxonomy and integrated it into Chinese curriculum design for two types of text reading—argumentative essay and narrative novel. We also explored the potential of combining Retrieval-Augmented Generation and few-shot learning in large language models (LLMs) for automated scoring and personalized feedback on reading comprehension questions following the proposed CT reading literacy framework. Rating students’ performance by LLMs showed advantages and benefits for students in facilitating autonomous knowledge construction and enhancing self-directed learning capabilities.

Authors

• Chenxuan He, Central China Normal University

• Zhen Li, Central China Normal University

• Luyao Ye, Central China Normal University

• Henan Zhang, Texas A&M University