Effects of generative AI virtual humans on nursing students' self-reported communication skills, self-directed learning, and self-efficacy: A quasi-experimental study.

AIM: This study aimed to examine the effects of generative artificial intelligence (AI) virtual humans (VHs) on nursing students' self-reported communication skills, self-directed learning, and self-efficacy.

DESIGN

A quasi-experimental design, with pretest, posttest, and follow-up test measures, was adopted and included an experimental group and a control group.

METHODS

Eighty-six second-year nursing students from a junior college in Taiwan were assigned to an experimental group (n = 45) receiving generative AI VHs or a control group (n = 41) receiving traditional lectures and role-play.

RESULTS

The experimental group showed significantly greater improvements in students' self-reported communication skills (both p < 0.001), self-directed learning (both p < 0.001), and self-efficacy (both p < 0.001) compared to the control group from pretest to posttest and follow-up test. The intervention effects were large, with Cohen's d ranging from 1.19 to 2.24.

CONCLUSIONS

Generative AI VHs can enhance nursing students' self-reported communication skills, self-directed learning, and self-efficacy through realistic and feedback-oriented practice. This approach is practical and scalable, making it suitable for resource-limited educational settings.