Article Per Year (5 Year)
p-Index From 2021 - 2026
0.23
P-Index
This Author published in this journals
All Journal Jurnal Pendidikan Sains Indonesia (Indonesian Journal of Science Education)
Nuri Balta
Suleyman Demirel University, Kaskelen, Kazakhstan

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

 1 
Comparative Analysis of High School Students’ and ChatGPT-5 Evaluation Ability on Work and Energy Topics Fausta Theresiana; Judyanto Sirait; Lanang Maulana Aminullah; Nuri Balta
Jurnal Pendidikan Sains Indonesia Vol. 14 No. 3: JULY 2026
Publisher : Universitas Syiah Kuala

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24815/jpsi.v14i3.1985

Abstract

The development of science and technology emphasizes the importance of scientific practices, particularly the ability to evaluate scientific arguments; however, students’ evaluation abilities have not yet developed optimally. On the other hand, the use of artificial intelligence (AI), such as ChatGPT, in education is increasing. Nevertheless, studies comparing the evaluation abilities of high school students and ChatGPT in solving physics problems remain very limited. This study aims to comparatively analyze the evaluation abilities of high school students and ChatGPT-5 on the topic of work and energy. A descriptive quantitative approach was employed to analyze in depth the responses of both high school students and ChatGPT. The study involved 92 eleventh-grade high school students selected through purposive sampling and ChatGPT-5 as research subjects. Both the students and ChatGPT were asked to complete a work and energy test consisting of four evaluation-based essay questions. ChatGPT-5 was given the test along with specific structured prompts. The results indicate that the evaluation ability of high school students is in the moderate category (mean score of 68.65), with most students able to evaluate claims and statements, although misconceptions and errors in formula usage were still found. Meanwhile, ChatGPT-5 demonstrated a high level of evaluation ability (score of 100) when given structured prompts, showing excellent performance in providing scientific arguments and applying formulas in accordance with the concepts. These findings suggest that ChatGPT-5 can serve as a supplementary learning tool, while teachers need to strengthen students’ scientific reasoning skills through structured evaluation practices
Co-Authors Fausta Theresiana Judyanto Sirait Lanang Maulana Aminullah