Evaluating an AI-guided clinical reasoning tutor to support National Council Licensure Examination preparation among nursing students: a convergent parallel mixed-methods study

Abstract There has been growing interest in AI tools to develop nursing students’ clinical reasoning skills, given the strong emphasis on clinical judgment in the Next Generation National Council Licensure Examination (NGN). Although the benefits of AI tools for developing clinical reasoning skills are promising, there is a lack of high-quality evidence due to methodological constraints and, more importantly, the lack of a structured learning design. Therefore, this convergent parallel mixed-methods study examined the performance and perceptions of pre-licensure nursing students after using an AI tutor designed using the Learner Journey Framework to support preparations for pathophysiology-focused NGN-style questions. Eighty-nine matched cases were included in the paired pretest-posttest analysis, and 88 poststudy survey responses were analyzed qualitatively. Posttest scores were significantly higher across matched cases (Mean ± SD = 11.71 ± 2.00 vs. 8.01 ± 2.14; p < .001). Thematic analysis of the qualitative items yielded four themes. Students’ responses across the first three themes indicated that the AI tutor helped slow down premature answer selection, identify cues, link findings to pathophysiology, and justify answers, and served as both cognitive and affective scaffolding, allowing them to work through uncertainty with less anxiety. In the fourth theme, students’ responses indicated that they also valued the tutor’s calm tone and step-by-step pacing, while still asking for brief summaries and quicker confirmation at selected points. Taken together, the findings suggest that the value of the tutor was not simply more practice, but a clearer and more repeatable way to work through clinical reasoning tasks.