人工知能生成研究コンセプトペーパーに基づく評価に関する医用画像および放射線科学学生の経験

Artificial intelligence (AI) can complete tasks that once required human cognitive effort. As a result, assessments that traditionally measured a student's capacity to research, synthesise, write, or problem-solve now risk assessing AI performance rather than student learning. This necessitates rethinking and a pedagogical shift in assessments. Therefore, an assessment was developed based on students generating a research concept paper using AI. This study aimed to understand students' experiences of the assessment.

A qualitative study with an explorative and descriptive research design used an open-ended questionnaire to collect data from second-year medical imaging and radiation sciences research students. The data were analysed using Tesch's open coding.

The study response rate was 79. 3 % (n=115, N=145). Data analysis identified two themes: 1. mixed experiences reflecting the novelty and complexity of integrating AI into research-based tasks as unfamiliar and unique, but also an exciting challenge; 2. the reflective process contributed to a deeper understanding of the evolving role of AI in academic practice.

Although the assessment felt unfamiliar and complex, it also provided an engaging and valuable learning opportunity. The assessment deepened understanding of research principles while providing insight into the benefits, limitations, and ethical responsibilities associated with the use of AI. AI should be approached critically and intentionally rather than accepted passively.

The assessment demonstrated that integrating AI into learning can effectively develop students' research competence, AI literacy, and capacity for responsible, informed use of emerging technologies. Therefore, educators should consider integrating AI for authentic assessments.