Despite a rapid acceleration in the capacity and subsequent adoption of generative AI in higher education, there remains a dearth of empirical research inquiring into the perspectives, experiences, and efficacy of its use within speech language pathology [SLP]. Relative to the sciences and humanities, clinical SLP and audiology programs have a modest number of published empirical works delving into how students in these programs are using generative AI to learn, navigate coursework, and add the professionalizing skills they will need to thrive after graduation. To better understand generative AI’s potential impact on learning, we evaluate its perceived affordances and constraints based on student interviews and an artifact analysis of their submitted assignments. In this presentation and paper, we investigate the perceptions and experiences of undergraduate SLP students in using generative AI to support learning outcomes.

In this qualitative study, we employed ethnographic methods to collect and analyze data for the purpose of:
1) better understanding the roles generative AI played in the learning experiences of 12 SLP students in a course on technology use in clinical and educational settings;
2) gaining insight as to how we might redesign Alex-TA, a custom-designed generative AI-enabled chatbot based on GPT 4, to better serve future cohorts of SLP students.

This manuscript is part of a larger study undertaken at a Midwest US university, data included AI-related course assignments and individual semi-structured interviews for 12 undergraduate SLP students. All students in the course, including study participants, were given scaffolded access to Alex-TA. Students were encouraged to use Alex-TA when they had questions regarding the course, needed help preparing for their information session project, and to experiment with generative AI-supported learning. We analyzed this data using the constant comparative method, reading, rereading, and analyzing the data with and against itself as well as in light of the adjacent existing literature on student perceptions of generative AI in higher education, common challenges in SLP education, and strategies for providing informational counseling.

Our initial findings confirm that a significant portion of respondents used Alex-TA to ask clarifying questions about course expectations, have complex concepts explained to them in simpler terms, and receive ideas and feedback on their assignments. Specifically, we found that participants were more inclined to use Alex-TA for questions specifically related to the course rather than general inquiries about SLP. Furthermore, use of Alex-TA was especially frequent when participants felt the instructor was not available. Our findings suggest a range of attitudes and dispositions concerning the potential benefits and limitations of adapting generative AI innovations for SLP education. Undergraduate SLP students demonstrated cautious receptiveness towards using generative AI to support their learning. Many advocated for a balanced approach that integrated AI strategies with traditional pedagogical practices. The implications of this study can be used to aid educators in SLP and the related fields of audiology and communication sciences and disorders in developing effective approaches that support student learning and simultaneously promote responsible technology use.

Keywords: Speech Language Pathology, Generative AI, Learning, chatbot, higher education.