A Quasi-Experimental study on the use intention of TAM generative AI and the heterogeneity of college students' major types

Abstract

This study is based on the Technology Acceptance Model (TAM) and explores the impact mechanism of generative AI usage intention on the heterogeneity of college students' major types through a Quasi-Experimental design. By distinguishing between the two groups of humanities and social sciences and natural sciences, based on a Quasi-Experimental design, the Propensity Score Matching (PSM) method was applied for 1:1 matching. The paired sample t-test estimated the Average Treatment Effect on the Processed (ATT) to be 0.231. In addition, a classification model was constructed using the RandomForest Classifier model, and the predictive ability of TAM variables on major types was verified. Research has found that: (1) there are significant differences in the willingness to use generative AI among different groups (p<0.01); (2) Machine learning models can efficiently identify group characteristics: the classification accuracy based on TAM variables reaches 0.75; (3) Gender and treatment are key features that distinguish natural science students from humanities and social science students (with importance scores of 0.19 and 0.03, respectively). The research conclusion provides a theoretical basis for the differentiated education promotion of generative AI, and suggests optimizing AI curriculum design for majors heterogeneity.

Ähnliche Arbeiten