Investigating the Drivers of Generative AI Help-Seeking: Structural Equation Model of Attitudes, Norms, and Control

• 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

We investigate the predictors of college students’ intention to seek instrumental and executive help from GenAI tools. We analyzed survey data from 1,015 students enrolled in U.S. higher education institutions using structural equation modeling. Our analyses showed support for trust in GenAI content and worry about GenAI reliance being positively associated with attitudes. Additionally, being taught to use GenAI within one’s major and professor encouragement were positively associated with norms. Last, help-seeking self-efficacy toward GenAI and average GenAI use were positively associated with perceived behavioral control. The findings did not support any significant path between latent factors and instrumental and executive GenAI help-seeking. This study enables us to begin understanding the factors that motivate different GenAI help-seeking.

Authors

• Andrea Macias, University of Southern California

• Stephen J. Aguilar, University of Southern California