The Effect of ChatGPT-Assisted Medication Dosage Calculation Training on Accuracy, Time, and Learning Satisfaction Among Nursing Students

Abstract

Medication dosage calculation is a vital clinical skill essential for patient safety. However, many nursing students experience difficulties mastering this competency due to mathematical anxiety and limited practice. Artificial intelligence (AI)-based tools, such as ChatGPT, offer interactive and learner-centered educational experiences by providing personalized guidance and immediate feedback. This study aimed to evaluate the effects of ChatGPT-assisted medication dosage calculation training on nursing students’ knowledge, calculation accuracy, test completion time, and learning satisfaction. A single-group quasi-experimental pretest–post-test design was implemented in the nursing department of a public university in Turkey. A total of 41 first-year nursing students voluntarily participated. A 4-session ChatGPT-assisted training program was delivered, focusing on unit conversions, dilution, pediatric dosages, and infusion rate calculations. Data were collected through a medication knowledge test, stopwatch-measured test completion time, a Visual Analog Scale for satisfaction, and open-ended feedback. Data were analyzed using descriptive statistics, paired t tests, and Cohen’s d post-training knowledge scores increased from 49.85±21.83 to 77.36±19.19 (P<.001, d=1.26). Unanswered questions decreased from 7.95±5.30 to 1.17±2.08 (P<.001, d=1.28). Time per question decreased from 2.02±2.95 to 0.90±1.95 minutes (P=.019, d=0.38). The satisfaction score was 9.02±0.99. Most students (95.1%) preferred the AI-assisted method over traditional training. ChatGPT-assisted training significantly enhanced nursing students’ knowledge, accuracy, test efficiency, and satisfaction. These results support integrating AI tools into nursing education to improve clinical competence and engagement. Copyright © 2026 Wolters Kluwer Health, Inc. All rights reserved.