Pathways to Proficiency: Mapping the Structural Relationships of Personal Attributes in AI-Integrated Self-Directed Language Learning

• In Event: AERA Roundtable Session Thursday 4:15 pm - Gold 4
  In Roundtable Session: Motivation, Mindsets, and AI: Understanding Learner Attributes in AI-Enhanced Education

Abstract

This study explores how five personal attributes—attitude, self-efficacy, motivation, strategy use, and resource use—interact in AI-integrated self-directed language learning (SDL) with ChatGPT. Data collected from 699 ChatGPT users were analyzed using structural equation modeling. Results revealed that motivation fully mediated the effects of self-efficacy on both strategy and resource use, while partially mediating the effect of attitude on strategy use. Attitude demonstrated a strong direct effect on strategy use, but contrary to expectations, no significant direct effect on resource use. Self-efficacy showed no significant direct effect on either outcome variable, but rather exerted its influence indirectly through motivation. These findings suggest that in AI-integrated learning environments, psychological drive remains essential for transforming positive attitudes and confidence into effective learning behaviors.

Authors

• Belle Li, Purdue University

• Zhuo Zhang, Towson University

• Victoria Lynn Lowell, Purdue University