Evaluation of the Efficacy and Usability of Artificial Intelligence (ChatGPT) for Health Sciences Students

AI (Artificial Intelligence), Educational Activities

Artificial intelligence, Education, Health Science Students

Crossover Randomized Control trial, in which subjects are randomly assigned to one of two groups: one (ChatGPT) receiving the intervention that is being tested, and the control group receiving usual online resources.

A two-stage randomization strategy will be adopted. The allocation of participants into event groups A and B will be determined through computer-generated randomization in an allocation ratio: 1:1. To address ethical concerns regarding power dynamics between professors and students; the investigators have opted to involve an individual who is independent of the research team. This decision aims to minimize any potential bias or influence that could arise from direct professor involvement in the intervention. This approach helps ensure that the study is conducted fairly and unbiasedly, fostering a research environment that prioritizes equal opportunities for all participants. Both groups will simultaneously work on an assignment (A and B). The investigators employed a cross-over design to ensure equity in the learning experience for all students. This involved assigning students either the ChatGPD (A) or the regular online tools (B) for accessing resources and assisting them with their assignments. During the initial phase of the intervention, participants in group A will utilize ChatGPT to complete their assignments, while participants in Group B will serve as the control group. In the subsequent period, the interventions will switch between the two groups. To assess the participants' technology usability, both groups will complete pre-and post-questionnaires. Additionally, participants will be requested to fill out a post-intervention questionnaire to evaluate their perceptions of the assistance methods. After a period of a wash-out of 13 days/week, participants who were previously allocated to Group A will be allocated to Group B and the same with participants previously allocated to Group B will be to Group A. This crossover randomized control trial will include 50 health sciences students who are registered to the course Chronic Health Conditions and Disability ( HLTH 3503), Carleton University. During the first part of the intervention, A group will use ChatGPT to complete their assignment, while B participants will act as a control group. In the second week, the interventions will cross over. Both groups' participants will complete a pre-and post-questionnaire to assess their knowledge. Participants will also complete a post-intervention questionnaire to evaluate their perceptions. The System Usability Scale (SUS) will be administered at the end of the interventions and when participants cross over to the other assistance tool. This allows for the assessment of outcomes and comparisons within and between the two intervention arms. By collecting data at multiple points in the study, the investigators can evaluate the effects of each intervention and any potential carryover effects during the crossover period. Additionally, it provides an opportunity to measure changes in outcomes within individual participants as they switch from one intervention to another. The questionnaire about socioeconomic, literacy questions, Benefits of AI and risks and negative experiences associated with AI usage.

Canada