Abstract: University Student Attitudes Towards and Intentions to Use Generative AI Use in the Classroom: A Pre- and Post-Semester Comparative Study (Society for Social Work and Research 30th Annual Conference Anniversary)

Schedule:

Sunday, January 18, 2026

Marquis BR 7, ML 2 (Marriott Marquis Washington DC)

* noted as presenting author

Yanfeng Xu, PhD, Assistant Professor, University of South Carolina, Columbia, SC

Mercy Namateng, MSW, Phd Student, University of South Carolina, Columbia, SC

Stella Radke, MSW, Research Assistant, University of South Carolina, Columbia, SC

Gareth Rees-White, PhD, Instructor, University of South Carolina, SC

Hengtao Tang, PhD, Associate Professor, University of South Carolina, Columbia, SC

Oz Ince, PhD, Clinical Associate Professor, University of South Carolina, SC

Allison Byxbe, PhD, Lecturer, University of South Carolina, SC

Sicheng Wang, PhD, Assistant Professor, University of South Carolina

Michael Stoelzner, PhD, Professor, University of South Carolina, SC

Shan Qiao, PhD, Associate Professor, University of South Carolina, Columbia, SC

Omar Roy, PhD, Assistant Professor, University of South Carolina, SC

Gloria Washington, Ed.D, Instructional Designer, University of South Carolina, SC

Background and Purpose: In response to the growing integration of generative AI (GenAI) in higher education, a large R1 university implemented the applications of GenAI tools across multiple of classes taught by ten instructors in Fall 2024. This study aimed to explore and compare factors associated with both student support for integrating GenAI into classroom settings along with their intentions to use GenAI tools in their education.

Methods: Surveys were administered at the start and end of the semester. In the present study, we analyzed two outcomes: (1) support for integrating GenAI in classrooms, and (2) intention to use GenAI in future education. Predictors included students’ demographics, interest in GenAI, perceived importance of AI for future jobs, perceived usefulness, and perceived ease of use. Descriptive statistics and ordered logistic regression models were conducted using STATA 17.0.

Results: Pre-implementation results: 168 students responded (76% White, 60% female, 80% undergraduates). Among them, 9.6% did not support the application of GenAI, 18% slightly supported it, 43% moderately supported it, 24% strongly supported it, and 5% fully supported it. Regarding their intention to use GenAI tools, 15% reported no intention, 20% were neutral, 21% somewhat agreed, 29% agreed, and 14% strongly agreed. Significant predictors of their support included: interest in GenAI (b = 1.27, p < 0.001), perceived importance of GenAI for their future jobs (b = 0.49, p = 0.046), perceived usefulness (b = 0.81, p < 0.001), and perceived ease of use (b = 0.42, p = 0.041). Regarding their intention to use GenAI in future education, significant predictors included being White (b = -1.07, p = 0.033), interest in GenAI (b = 0.65, p = 0.022), perceived usefulness (b = 1.34, p < 0.001), and a supportive attitude toward GenAI (b = 0.78, p = 0.009). Post-implementation results: 205 students responded to the post-implementation survey, with similar demographic characteristics. After the implementation of the GenAI initiative, only 3.48% of students reported no support at all, while 16.92% slightly supported, 40.30% moderately supported, 28.36% strongly supported, and 10.95% fully supported the use of GenAI. Regarding their intention to use GenAI, 5.18% disagreed, 15.54% were neutral, 26.94% somewhat agreed, 36.27% agreed, and 16.06% strongly agreed. Significant predictors of students’ support for GenAI included: interest in GenAI (b = 1.23, p < 0.001), perceived importance of GenAI for future jobs (b = 0.46, p = 0.038), perceived usefulness (b = 0.70, p < 0.001), and being an engineering student (b = -2.62, p = 0.023). For their intention to use GenAI tools, significant predictors were international student status (b = 4.17, p = 0.002), perceived usefulness (b = 1.11, p < 0.001), and a supportive attitude toward GenAI (b = 0.98, p < 0.001).

Conclusions and Implications: Students showed increased support for and intention to use GenAI after the implementation. Consistent predictors were interest in GenAI and perceived usefulness to future careers. Integrating GenAI into higher education is feasible but needs to emphasize the importance and usefulness to students’ current study and future careers.