A study of the effects of GenAI-based pair programming strategy on middle school students'attitudes toward computational thinking and programming self-efficacy

Abstract

This study explores the integration of GenAI technology with pair programming in K-12 education to address issues like limited teaching methods and students' lack of programming confidence. It involved three seventh-grade classes in Wenzhou, divided into three groups: Experimental Group 1 (P-GenAI programming), Experimental Group 2 (C-GenAI programming), and a control group (traditional programming). After a 10-week teaching experiment, the results showed no significant differences in computational thinking attitude or programming self-efficacy among the groups. However, Experimental Group 1 generally performed better in all aspects, suggesting positive trends. The study recommends enhancing programming resilience, overcoming human-computer collaboration challenges, and building a collaborative multi-intelligence ecosystem to improve programming education.