Effect of AI Chatbot-Assisted Versus Traditional Case-Based Learning on Clinical Reasoning in Occupational Therapy Students: A Study on Parkinson's Disease

Overview

This study aims to examine whether using an artificial intelligence (AI) chatbot can enhance occupational therapy students' learning during a case-based activity focused on Parkinson's disease. The research compares two groups of students: one using traditional learning materials, and another using both traditional resources and a conversational AI chatbot. Students in both groups work in teams to analyze the same clinical case and propose assessment and treatment strategies for a hypothetical patient. The main purpose of the study is to evaluate whether the AI chatbot helps improve students' performance in three learning domains: cognitive (knowledge and understanding), affective (empathy and attitudes), and psychomotor (planning and action skills). Students' performance is assessed through a structured written examination. The hypothesis is that students who use the AI chatbot will achieve higher scores, especially in the cognitive and psychomotor domains, compared to those who rely on traditional methods only. The study also examines how students interact with the chatbot and whether they use it to support deeper clinical reasoning. By exploring the role of AI in occupational therapy education, this research seeks to inform future teaching strategies and support the thoughtful integration of digital tools in health professions training.

This study investigates the impact of an artificial intelligence (AI) chatbot on occupational therapy (OT) students' clinical reasoning skills within the context of neurological rehabilitation, specifically focusing on Parkinson's disease. The study uses a post-test only control group design and adopts a mixed-methods approach to assess educational outcomes across three learning domains: cognitive, affective, and psychomotor. A total of 25 OT undergraduate students enrolled in a Neurological Rehabilitation course are randomly assigned to one of two groups: (1) Chatbot Group and (2) Classic Group. Both groups receive the same didactic instruction and engage in a small-group, case-based learning session featuring a hypothetical Parkinson's disease case. The Classic Group uses traditional learning resources such as lecture notes and textbooks, while the Chatbot Group additionally interacts with an AI language model simulating a patient. The chatbot allows students to ask open-ended questions to gather occupational history, explore symptoms, and plan interventions. After the learning task, all students complete a six-item written exam assessing their performance in the cognitive, affective, and psychomotor domains. Two independent raters, blinded to group assignment, evaluate student responses using a predefined rubric. Inter-rater reliability is calculated. To account for possible differences in baseline academic performance, Grade Point Average (GPA) is used as a covariate in the statistical analysis. The study also includes a qualitative component in which students in the Chatbot Group submit the queries they posed during the interaction. These queries are analyzed inductively using a content analysis approach to explore how students engage with AI support-whether for conceptual clarification, procedural guidance, or ethical reasoning. This research addresses a growing interest in how AI-based tools may enhance, supplement, or potentially limit professional training in occupational therapy. Although AI chatbots may offer convenient access to information and support student creativity, concerns remain about their effectiveness in fostering reflective and ethical clinical reasoning. By analyzing both performance data and interaction patterns, this study aims to offer evidence-based insights into the role of digital tools in OT education.