Physiotherapy Students Show Higher AI Literacy Scores After Curriculum Changes

Physiotherapy students at Linköping University in Sweden reported higher generative artificial intelligence (GAI) literacy scores in semester 3 compared to semester 1 students, following the implementation of structured GAI learning activities. A repeated cross-sectional programme evaluation published in BMC Medical Education assessed student use, perceptions, and literacy levels after curricular adaptations were introduced. The study utilized non-equivalent, unlinked samples to evaluate the impact of these educational changes on student behavior and knowledge within the physiotherapy program.

The research aimed to evaluate two primary objectives: the use and perceptions of GAI among students during the implementation of curricular adaptations, and GAI literacy levels before and after structured learning activities. The evaluation consisted of two parts. The first part compared student surveys conducted in 2024 and 2025 to observe shifts in usage patterns and attitudes.

The second part compared literacy scores from semester 1, prior to participation in the GAI activities, with scores from semester 3, after completion of these activities. This design allowed researchers to assess associations between the curriculum changes and student outcomes without establishing direct causality due to the cross-sectional nature of the data. Reported GAI use differed significantly between the 2024 and 2025 survey waves.

In the 2024 survey, 13% of respondents reported using GAI daily. By the 2025 survey, this figure had risen to 57%. Statistical analysis confirmed that the difference in usage frequency was significant, with a chi-square value of 62.40 and a p-value less than 0.001.

This shift indicates a marked increase in the frequency with which students engaged with generative AI tools during their studies over the one-year period. Beyond frequency, the nature of student interactions with GAI became more diverse. Respondents in the 2025 survey reported using GAI for a wider range of academic tasks, including brainstorming, question generation, and text refinement.

These specific applications were reported significantly more often than in the 2024 survey, with p-values less than 0.001 for the overall difference in use categories. This diversification suggests that students were utilizing the tools for various cognitive processes required for their physiotherapy education, rather than limiting use to a single type of task. Perceptions regarding the educational value of GAI also showed a positive shift.

Students in the 2025 survey were more likely to view GAI as contributing to their learning compared to respondents in the 2024 survey. The statistical significance of this change in perception was confirmed with a chi-square value of 24.94 and a p-value less than 0.001. However, perceptions regarding the clarity of information about permissible and impermissible uses of GAI remained similar across both survey waves, indicating that while attitudes toward learning value changed, understanding of usage guidelines did not show a significant difference between the two groups.

The study measured specific GAI literacy using the 10-item Generative Artificial Intelligence Literacy Scale (GAILS-10). When comparing semester 1 students, who were assessed before participating in the structured GAI learning activities, with semester 3 students who had completed them, significant differences were observed. Semester 3 students reported higher GAILS-10 scores across all measured dimensions.

The total score showed a large between-group effect size, with a Cohen's d value of 1.0, indicating a substantial difference in reported literacy levels between the two groups. The curricular adaptations involved the delivery of structured GAI learning activities during semesters 1 through 3. The study notes that structured GAI training has been proposed as an effective approach to strengthen healthcare students' confidence in navigating AI-related challenges and ethical considerations, though few prior studies had evaluated the impact of implemented curricular adaptations in physiotherapy education specifically.

The findings suggest a need for curricular adaptations and educational support to promote appropriate GAI use, reduce misuse, and strengthen student GAI literacy. The significant increase in daily usage and the higher literacy scores in semester 3 are associated with the implementation of the structured learning activities. However, the lack of significant change in perceptions regarding permissible use highlights an area where further educational support may be required.

The study provides evidence that students can engage with AI tools more frequently and diversely when provided with structured educational support.

Subject of Research: Evaluation of generative AI literacy and usage in physiotherapy education following curricular adaptations

Article Title: Integrating generative AI into physiotherapy education: students’ use, perceptions, and generative AI literacy following curricular adaptations: a repeated cross-sectional programme evaluation

Article References: Schröder, K., Lindbäck, Y., Engström, T., & Sonesson, S. (2026). Integrating generative AI into physiotherapy education: students’ use, perceptions, and generative AI literacy following curricular adaptations: a repeated cross-sectional programme evaluation. BMC Medical Education, 26(1), Article 1399. https://doi.org/10.1186/s12909-026-10256-0

Image Credits: AI Generated

DOI: 10.1186/s12909-026-10256-0

Keywords: physiotherapy education, generative AI, curricular adaptation, student literacy, healthcare education, cross-sectional study, AI ethics, learning outcomes, Integrating, generative, physiotherapy, education