Scholarlink

Background <p>Large language models (LLMs) are being gradually integrated into dental education, but more attention was paid in perspective of educators rather than students. Evidence regarding how students employ them for self-learning in Preventive Dentistry, and whether specific usage patterns are related to academic performance, remains scarce. This study examined dental students’ self-perceived LLM knowledge, attitudes, and behaviors; evaluated perceived usefulness across Preventive Dentistry modules; and explored associations between LLM usage (frequency, type, motivation) and academic performance.</p> Methods <p>This cross-sectional pilot study recruited all third-year dental undergraduates enrolled in a Preventive Dentistry course at a university in eastern China. A total of 117 students provided valid responses and completed final examination. Academic performance was assessed using final examination scores as a continuous outcome variable. Linear regression analysis was performed to examine associations between LLM usage and examination scores. The data were analyzed using SPSS 22.0, and the level of significance was set at 0.05.</p> Results <p>LLM use was reported by all participants, with eighty-four students (71.8%) considering themselves as proficient users. For self-learning in Preventive Dentistry, 49.6% of participants reported using LLMs one to three times per month and DeepSeek (99.1%) was most popular. Students perceived LLMs as more helpful for theoretical learning than for practical skill acquisition (<i>p</i> &lt; 0.001). The most frequently reported motivation for LLM use was to broaden knowledge and facilitate comprehension (76.1%), compared with other motivations. Importantly, the linear regression model (adjusted R<sup>2</sup> = 0.274) showed that an association was observed between the motivation “to broaden knowledge and facilitate comprehension” and academic performance (B = 5.321 (95%CI: 1.669 to 8.973), β = 0.233, <i>p</i> = 0.005), indicating a small-to-moderate positive effect size.</p> Conclusions <p>In this pilot sample, LLMs were widely reported to be used and were perceived by these students as more helpful for learning theoretical rather than for practical knowledge during self-learning in Preventive Dentistry. Using LLMs to broaden knowledge and facilitate comprehension was the most frequently reported motivation and showed a modest positive association with academic performance, although causality cannot be inferred due to the cross-sectional design. These findings provide preliminary insights into students’ use of LLMs and may help inform future research on their role in dental education.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Exploring the use of large language models for self-learning in preventive dentistry and their associations with academic performance: a cross-sectional pilot study among dental students

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月20日

查看全文
  • Jiayi Yang,
  • Mingming Li,
  • Shuhua Wang,
  • Zhuojin Shi,
  • Jinhua Fang,
  • Pei Zheng,
  • Zhiwu Wu

摘要

Background

Large language models (LLMs) are being gradually integrated into dental education, but more attention was paid in perspective of educators rather than students. Evidence regarding how students employ them for self-learning in Preventive Dentistry, and whether specific usage patterns are related to academic performance, remains scarce. This study examined dental students’ self-perceived LLM knowledge, attitudes, and behaviors; evaluated perceived usefulness across Preventive Dentistry modules; and explored associations between LLM usage (frequency, type, motivation) and academic performance.

Methods

This cross-sectional pilot study recruited all third-year dental undergraduates enrolled in a Preventive Dentistry course at a university in eastern China. A total of 117 students provided valid responses and completed final examination. Academic performance was assessed using final examination scores as a continuous outcome variable. Linear regression analysis was performed to examine associations between LLM usage and examination scores. The data were analyzed using SPSS 22.0, and the level of significance was set at 0.05.

Results

LLM use was reported by all participants, with eighty-four students (71.8%) considering themselves as proficient users. For self-learning in Preventive Dentistry, 49.6% of participants reported using LLMs one to three times per month and DeepSeek (99.1%) was most popular. Students perceived LLMs as more helpful for theoretical learning than for practical skill acquisition (p < 0.001). The most frequently reported motivation for LLM use was to broaden knowledge and facilitate comprehension (76.1%), compared with other motivations. Importantly, the linear regression model (adjusted R2 = 0.274) showed that an association was observed between the motivation “to broaden knowledge and facilitate comprehension” and academic performance (B = 5.321 (95%CI: 1.669 to 8.973), β = 0.233, p = 0.005), indicating a small-to-moderate positive effect size.

Conclusions

In this pilot sample, LLMs were widely reported to be used and were perceived by these students as more helpful for learning theoretical rather than for practical knowledge during self-learning in Preventive Dentistry. Using LLMs to broaden knowledge and facilitate comprehension was the most frequently reported motivation and showed a modest positive association with academic performance, although causality cannot be inferred due to the cross-sectional design. These findings provide preliminary insights into students’ use of LLMs and may help inform future research on their role in dental education.