Artificial intelligence (AI) performance on pharmacy skills laboratory course assignments

IF 1.4 Q3 EDUCATION, SCIENTIFIC DISCIPLINES Currents in Pharmacy Teaching and Learning Pub Date : 2025-07-01 Epub Date: 2025-04-23 DOI:10.1016/j.cptl.2025.102367

Vivian Do , Krista L. Donohoe , Apryl N. Peddi , Eleanor Carr , Christina Kim , Virginia Mele , Dhruv Patel , Alexis N. Crawford

{"title":"Artificial intelligence (AI) performance on pharmacy skills laboratory course assignments","authors":"Vivian Do , Krista L. Donohoe , Apryl N. Peddi , Eleanor Carr , Christina Kim , Virginia Mele , Dhruv Patel , Alexis N. Crawford","doi":"10.1016/j.cptl.2025.102367","DOIUrl":null,"url":null,"abstract":"<div><h3>Objective</h3><div>To compare pharmacy student scores to scores of artificial intelligence (AI)-generated results of three common platforms on pharmacy skills laboratory assignments.</div></div><div><h3>Methods</h3><div>Pharmacy skills laboratory course assignments were completed by four fourth-year pharmacy student investigators with three free AI platforms: ChatGPT, Copilot, and Gemini. Assignments evaluated were calculations, patient case vignettes, in-depth patient cases, drug information questions, and a reflection activity. Course coordinators graded the AI-generated submissions. Descriptive statistics were utilized to summarize AI scores and compare averages to recent pharmacy student cohorts. Interrater reliability for the four student investigators completing the assignments was assessed.</div></div><div><h3>Results</h3><div>Fourteen skills laboratory assignments were completed utilizing three different AI platforms (ChatGPT, Copilot, and Gemini) by four fourth-year student investigators (<em>n</em> = 168 AI-generated submissions). Copilot was unable to complete 12; therefore, 156 AI-generated submissions were graded by the faculty course coordinators for accuracy and scored from 0 to 100 %. Pharmacy student cohort scores were higher than the average AI scores for all of the skills laboratory assignments except for two in-depth patient cases completed with ChatGPT. Conclusion. Pharmacy students on average performed better on most skills laboratory assignments than three commonly used artificial intelligence platforms. Teaching students the strengths and weaknesses of utilizing AI in the classroom is essential.</div></div>","PeriodicalId":47501,"journal":{"name":"Currents in Pharmacy Teaching and Learning","volume":"17 7","pages":"Article 102367"},"PeriodicalIF":1.4000,"publicationDate":"2025-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Currents in Pharmacy Teaching and Learning","FirstCategoryId":"1085","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S1877129725000887","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2025/4/23 0:00:00","PubModel":"Epub","JCR":"Q3","JCRName":"EDUCATION, SCIENTIFIC DISCIPLINES","Score":null,"Total":0}

引用次数: 0

Abstract

Objective

To compare pharmacy student scores to scores of artificial intelligence (AI)-generated results of three common platforms on pharmacy skills laboratory assignments.

Methods

Pharmacy skills laboratory course assignments were completed by four fourth-year pharmacy student investigators with three free AI platforms: ChatGPT, Copilot, and Gemini. Assignments evaluated were calculations, patient case vignettes, in-depth patient cases, drug information questions, and a reflection activity. Course coordinators graded the AI-generated submissions. Descriptive statistics were utilized to summarize AI scores and compare averages to recent pharmacy student cohorts. Interrater reliability for the four student investigators completing the assignments was assessed.

Results

Fourteen skills laboratory assignments were completed utilizing three different AI platforms (ChatGPT, Copilot, and Gemini) by four fourth-year student investigators (n = 168 AI-generated submissions). Copilot was unable to complete 12; therefore, 156 AI-generated submissions were graded by the faculty course coordinators for accuracy and scored from 0 to 100 %. Pharmacy student cohort scores were higher than the average AI scores for all of the skills laboratory assignments except for two in-depth patient cases completed with ChatGPT. Conclusion. Pharmacy students on average performed better on most skills laboratory assignments than three commonly used artificial intelligence platforms. Teaching students the strengths and weaknesses of utilizing AI in the classroom is essential.

查看原文

微信好友朋友圈 QQ好友复制链接

本刊更多论文

人工智能（AI）在药学技能实验课程作业中的表现

目的比较药学专业学生与人工智能（AI）生成的三个常用平台在药学技能实验室作业中的得分。方法由4名药学四年级学生在ChatGPT、Copilot和Gemini三个免费人工智能平台上完成药学技能实验课程作业。评估的作业包括计算、患者病例小品、深入患者病例、药物信息问题和反思活动。课程协调员对人工智能生成的提交进行评分。使用描述性统计来总结AI分数，并将平均值与最近的药学学生队列进行比较。对完成作业的四名学生调查者进行了评估。结果四名四年级学生利用三种不同的人工智能平台（ChatGPT、Copilot和Gemini）完成了14项技能实验室作业（n = 168份人工智能生成的提交）。副驾驶无法完成12；因此，156份人工智能提交的材料由教师课程协调员对其准确性进行评分，得分从0到100%不等。除了用ChatGPT完成的两个深度患者案例外，药学学生的队列得分高于所有技能实验室作业的平均AI得分。结论。药学专业学生在大多数技能实验室作业中的平均表现优于三种常用的人工智能平台。让学生了解在课堂上使用人工智能的优缺点是至关重要的。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文去求助

来源期刊