Abstract
ABSTRACT
Artificial intelligence (AI) is rapidly transforming many aspects of society, including healthcare and education. Understanding its effectiveness is essential for designing optimal learning experiences in health science and rehabilitation education. Purpose This study compared traditional learning methods (including text-only and audio-only) with an AI-generated podcast of the same learning material to determine their effects on academic performance among Doctor of Physical Therapy (DPT) students. Results were intended to inform the effectiveness of each learning method when measuring immediate basic recall. Student performance on posttest served as the dependent variable. Methods Sixty DPT students were randomly assigned to one of three learning modalities: (a) text-only, (b) audio-only, or (c) an AI-generated podcast created using Google NotebookLM. After participants engaged in each learning modality, they completed a standardized multiple-choice post-assessment. Results Sixty participants completed the study and were assigned to the groups: audio-only (n = 19), AI-generated podcast (n = 18), or text-only (n = 23) learning condition. Median post-assessment scores were 13 (IQR = 1) for the audio-only and 14 (IQR = 1) for both the AI-generated podcast and text-only conditions. A Kruskal-Wallis test revealed no significant difference in post-assessment scores among the three learning conditions, H(2) = 1.36, p = .507, with a negligible effect size (ε² ≈ .00). Conclusion Immediate recall performance was not statistically significant between text-only, audio-only, and AI-generated podcasts learning conditions among graduate students. Although no modality demonstrated superior performance, participant feedback suggests that AI-generated podcasts may offer a more engaging alternative to traditional audio narration for students who prefer auditory learning. AI-based podcasts may therefore serve as a useful supplemental learning resource. Future research should examine long-term knowledge retention, higher-order cognitive outcomes, more diverse student populations, and exploring other learning modalities.
Recommended Citation
Lewis, Christopher J.; Lin, Megan; Shalamoff, Elana; and Manickam, Arun
(2026)
"Pedagogical Assessment of Student Learning Using AI,"
International Journal of Exercise Science: Conference Proceedings: Vol. 15:
Iss.
9, Article 13.
Available at:
https://digitalcommons.wku.edu/ijesab/vol15/iss9/13