- Physical Review Journals
All JournalsPhysics Magazine

Physical Review Physics Education Research

Reuse & Permissions

It is not necessary to obtain permission to reuse this article or its components as it is available under the terms of the Creative Commons Attribution 4.0 International license. This license permits unrestricted use, distribution, and reproduction in any medium, provided attribution to the author(s) and the published article's title, journal citation, and DOI are maintained. Please note that some figures may have been included with permission from other third parties. It is your responsibility to obtain the proper permission from the rights holder directly for these figures.

Export citation

Export citation

Choose format for download:

Download Citation
  • Open Access

Comparing teacher-centered and student-centered agents based on prompt engineering: Effects on learning performance, cognitive load, flow experience, and empathy perception in physics learning

Yimin Wang1,3, Xiaotong Chen1,2, Yushan Xiong1,2, Shaorui Xu1,4,*, Qiuye Li1,2,†, and Shaona Zhou1,2,‡

  • 1Guangdong Provincial Key Laboratory of Quantum Engineering and Quantum Materials, National Demonstration Center for Experimental Physics Education, School of Physics, South China Normal University, Guangzhou, People’s Republic of China
  • 2Guangdong Basic Research Center of Excellence for Structure and Fundamental Interactions of Matter, Key Laboratory of Atomic and Subatomic Structure and Quantum Control (Ministry of Education), South China Normal University, Guangzhou, People’s Republic of China
  • 3Guangdong University of Education, Guangzhou, People’s Republic of China
  • 4School of Electronics and Communication, Guangdong Mechanical & Electrical Polytechnic, Guangzhou, Guangdong, People’s Republic of China

  • *Contact author: jayxee@hotmail.co.uk
  • †Contact author: 2019021855@m.scnu.edu.cn
  • ‡Contact author: zhou.shaona@m.scnu.edu.cn
PDF

Phys. Rev. Phys. Educ. Res. 22, 020131 – Published 9 September, 2026

DOI: https://doi.org/10.1103/9t5b-twsb

Abstract

Large language models (LLMs) often face challenges in generating accurate and coherent explanations when solving physics questions. Prompt engineering serves as an effective strategy for human-artificial intelligence communication, guiding LLMs to generate precise and contextually relevant responses with well-designed instructions. The study utilized prompt engineering on the Coze platform to construct two types of instructional agents: a teacher-centered agent and a student-centered agent. They were designed to respond to students’ questions in accordance with their respective roles and communication styles. Specifically, the teacher-centered agent tended to provide authoritative, professionally grounded, and content-focused responses, whereas the student-centered agent was designed to be more attentive to students’ prior knowledge and learning needs, offering adaptive and supportive guidance. The study compared the effects of these two agents on students’ learning performance, cognitive load, flow experience, and empathy perception. The results revealed differential effects between the two prompt-engineered agent conditions. Compared with the teacher-centered agent, students interacting with the student-centered agent condition showed higher learning performance, lower extraneous cognitive load, stronger germane cognitive engagement, greater flow experience, and higher empathy perception.

View figure in article

Physics Subject Headings (PhySH)

Article Text

References (84)

  1. A. Extance, ChatGPT has entered the classroom: How large language models could transform education, Nature (London) 623, 474 (2023).
  2. Y. Zhu, C. Zhu, T. Wu, S. Wang, Y. Zhou, J. Chen, F. Wu, and Y. Li, Impact of assignment completion assisted by large language model-based chatbot on middle school students’ learning, Educ. Inf. Technol. 30, 2429 (2025).
  3. A. Coban, D. Dzsotjan, S. Küchemann, J. Durst, J. Kuhn, and C. Hoyer, AI support meets AR visualization for Alice and Bob: Personalized learning based on individual ChatGPT feedback in an AR quantum cryptography experiment for physics lab courses, EPJ Quantum Technol. 12, 15 (2025).
  4. R. Malik, D. Abdi, R. Wang, and D. Demszky, Scaffolding middle school mathematics curricula with large language models, Br. J. Educ. Technol. 56, 999 (2025).
  5. B. Gregorcic and A.-M. Pendrill, ChatGPT and the frustrated Socrates, Phys. Educ. 58, 035021 (2023).
  6. G. Polverini, J. Melin, E. Önerud, and B. Gregorcic, Performance of ChatGPT on tasks involving physics visual representations: The case of the brief electricity and magnetism assessment, Phys. Rev. Phys. Educ. Res. 21, 010154 (2025).
  7. R. Loubet, P. Zittlau, L. Vollmer, M. Hoffmann, S. Fellenz, F. Jirasek, H. Leitte, and H. Hasse, Using large language models for solving textbook-style thermodynamic problems, Comput. Chem. Eng. 204, 109333 (2026).
  8. H. Liu, H. Yin, Z. Luo, and X. Wang, Integrating chemistry knowledge in large language models via prompt engineering, Synth. Syst. Biotechnol. 10, 23 (2025).
  9. N. Xu and X. Ma, DecoPrompt: Decoding prompts reduces hallucinations when large language models meet false premises, arXiv:2411.07457.
  10. J. Cheng, T. Su, J. Yuan, G. He, J. Liu, X. Tao, J. Xie, and H. Li, Chain-of-thought prompting obscures hallucination cues in large language models: An empirical evaluation, arXiv:2506.17088.
  11. D. Tong, B. Jin, Y. Tao, H. Ren, A. Y. M. Atiquil Islam, and L. Bao, Exploring the role of human-AI collaboration in solving scientific problems, Phys. Rev. Phys. Educ. Res. 21, 010149 (2025).
  12. L. Gong, J. Chen, and F. Wu, Is ChatGPT a competent teacher? Systematic evaluation of large language models on the competency model, IEEE Trans. Learn. Technol. 18, 530 (2025).
  13. E. Tufino and B. Gregorcic, Creating a customisable Socratic AI physics tutor, Phys. Educ. 60, 065037 (2025).
  14. A. Sirnoorkar and N. S. Rebello, Feedback that clicks: Introductory physics students’ valued features in AI feedback generated from self-crafted and engineered prompts, arXiv:2509.08516.
  15. G. Kortemeyer, Could an artificial-intelligence agent pass an introductory physics course?, Phys. Rev. Phys. Educ. Res. 19, 010132 (2023).
  16. C. G. West, Advances in apparent conceptual physics reasoning in GPT-4, arXiv:2303.17012.
  17. Y. Liang, D. Zou, H. Xie, and F. L. Wang, Exploring the potential of using ChatGPT in physics education, Smart Learn. Environ. 10, 52 (2023).
  18. Z. Chen and T. Wan, Grading explanations of problem-solving process and generating feedback using large language models at human-level accuracy, Phys. Rev. Phys. Educ. Res. 21, 010126 (2025).
  19. T. Wan and Z. Chen, Exploring generative AI assisted feedback writing for students’ written responses to a physics conceptual question with prompt engineering and few-shot learning, Phys. Rev. Phys. Educ. Res. 20, 010152 (2024).
  20. S. Küchemann, S. Steinert, N. Revenga, M. Schweinberger, Y. Dinc, K. E. Avila, and J. Kuhn, Can ChatGPT support prospective teachers in physics task development?, Phys. Rev. Phys. Educ. Res. 19, 020128 (2023).
  21. J. White, Q. Fu, S. Hays, M. Sandborn, C. Olea, H. Gilbert, A. Elnashar, J. Spencer-Smith, and D. C. Schmidt, A prompt pattern catalog to enhance prompt engineering with ChatGPT, arXiv:2302.11382.
  22. L. Giray, Prompt engineering with ChatGPT: A guide for academic writers, Ann. Biomed. Eng. 51, 2629 (2023).
  23. J. Wei, X. Wang, D. Schuurmans, M. Bosma, B. Ichter, F. Xia, E. H. Chi, Q. V. Le, and D. Zhou, Chain-of-thought prompting elicits reasoning in large language models, Adv. Neural Inf. Process. Syst. 35, 24824 (2022).
  24. J. Wu, J. Li, Z. Ge, M. Xu, L. Lin, and R. Zhang, Effectiveness of generative AI in automated written corrective feedback with prompting, J. Educ. Comput. Res. 63, 1493 (2025).
  25. T. Kojima, S. S. Gu, M. Reid, Y. Matsuo, and Y. Iwasawa, Large language models are zero-shot reasoners, arXiv:2205.11916.
  26. OpenAI, Reasoning models, https://developers.openai.com/api/docs/guides/reasoning?utm_source=chatgpt.com.
  27. C. Cohn, N. Hutchins, T. Le, and G. Biswas, A chain-of-thought prompting approach with LLMs for evaluating students, Formative Assess. Responses Sci., AAAI 38, 23182 (2024).
  28. Z. Lai, G. Dai, Y. Lu, Y. Xu, J. Li, and X. Zhu, DeepSyn: Operable chemical retrosynthetic design with the DeepSeek R1 model, J. Phys. Chem. Lett. 16, 4962 (2025).
  29. Y. Walter, Embracing the future of artificial intelligence in the classroom: The relevance of AI literacy, prompt engineering, and critical thinking in modern education, Int. J. Educ. Technol. High. Educ. 21, 15 (2024).
  30. G. Polverini and B. Gregorcic, How understanding large language models can inform the use of ChatGPT in physics education, Eur. J. Phys. 45, 025701 (2024).
  31. Z. Zheng, O. Zhang, C. Borgs, J. T. Chayes, and O. M. Yaghi, ChatGPT chemistry assistant for text mining and the prediction of MOF synthesis, J. Am. Chem. Soc. 145, 18048 (2023).
  32. B. Lyu, C. Li, H. Li, H. Oh, Y. Song, W. Zhu, and W. Xing, The role of teachable agents’ personality traits on student-AI interactions and math learning, Comput. Educ. 234, 105314 (2025).
  33. B. Xu, A. Yang, J. Lin, Q. Wang, C. Zhou, Y. Zhang, and Z. Mao, ExpertPrompting: Instructing large language models to be distinguished experts, arXiv:2305.14688.
  34. W. Xing, Y. Song, C. Li, Z. Liu, W. Zhu, and H. Oh, Development of a generative AI -powered teachable agent for middle school mathematics learning: A design-based research study, Br. J. Educ. Technol. 56, 2043 (2025).
  35. N. L. Schroeder, W. L. Romine, and S. D. Craig, Measuring pedagogical agent persona and the influence of agent persona on learning, Comput. Educ. 109, 176 (2017).
  36. R. Xiao, X. Hou, R. Ye, M. Kazemitabaar, N. Diana, M. Liut, and J. Stamper, Improving student-AI interaction through pedagogical prompting: An example in computer science education, arXiv:2506.19107.
  37. Y. Ding, H. Hu, J. Zhou, Q. Chen, B. Jiang, and L. He, Boosting large language models with Socratic method for conversational mathematics teaching, arXiv:2407.17349.
  38. G. Kestin, K. Miller, A. Klales, T. Milbourne, and G. Ponti, AI tutoring outperforms in-class active learning: An RCT introducing a novel research-based design in an authentic educational setting, Sci. Rep. 15, 17458 (2025).
  39. S. Küchemann, S. Steinert, J. Kuhn, K. Avila, and S. Ruzika, Large language models—Valuable tools that require a sensitive integration into teaching and learning physics, Phys. Teach. 62, 400 (2024).
  40. Significant Gravitas, AutoGPT, https://github.com/Significant-gravitas/AutoGPT.
  41. LangChain, https://www.langchain.com/agents.
  42. B. Ni and M. J. Buehler, MechAgents: Large language model multi-agent collaborations can solve mechanics problems, generate new data, and integrate knowledge, Extreme Mech. Lett. 67, 102131 (2024).
  43. Coze, https://www.coze.cn/docs/guides/welcome.
  44. Cherry Studio, https://www.cherry-ai.com/.
  45. M. Liu, M. Yang, Z.-M. Wu, and J. Liao, Development, application status and future prospect of large language model agents for education, Mod. Educ. Technol. 34, 5 (2024).
  46. F. Fan, S. Wang, M. D. M. Hemuti, X. Nie, and L. T. Yang, Enhancing students’ metacognition with innovative IA-based metacognitive reflective learning tool, IEEE Trans. Learn. Technol. 18, 699 (2025).
  47. J. Sweller, Cognitive load during problem solving: Effects on learning, Cogn. Sci. 12, 257 (1988).
  48. P. Chandler and J. Sweller, Cognitive load theory and the format of instruction, Cogn. Instr. 8, 293 (1991).
  49. J. Sweller, Cognitive load theory, learning difficulty, and instructional design, Learn. Instr. 4, 295 (1994).
  50. J. Sweller, J. J. G. van Merrienboer, and F. G. W. C. Paas, Cognitive architecture and instructional design, Educ. Psychol. Rev. 10, 251 (1998).
  51. Q. Liu, S. Yu, W. Chen, Q. Wang, and S. Xu, The effects of an augmented reality based magnetic experimental tool on students’ knowledge improvement and cognitive load, J. Comput. Assist. Learn. 37, 645 (2020).
  52. D. Baidoo-Anu, and L. Owusu Ansah, Education in the era of generative artificial intelligence (AI): Understanding the potential benefits of ChatGPT in promoting teaching and learning, J. AI 7, 52 (2023).
  53. L. Yan, R. Martinez-Maldonado, Y. Jin, V. Echeverria, M. Milesi, J. Fan, L. Zhao, R. Alfredo, X. Li, and D. Gašević, The effects of generative AI agents and scaffolding on enhancing students’ comprehension of visual learning analytics, Comput. Educ. 234, 105322 (2025).
  54. W. K. Monib, A. Qazi, and M. M. Mahmud, Exploring learners’ experiences and perceptions of ChatGPT as a learning tool in higher education, Educ. Inf. Technol. 30, 917 (2025).
  55. L. P. Patac, A. V. Patac Jr. Jr., Using ChatGPT for academic support: Managing cognitive load and enhancing learning efficiency–A phenomenological approach, Soc. Sci. Humanit. Open 11, 101301 (2025).
  56. J. Lademann, J. Henze, and S. Becker-Genschow, Augmenting learning environments using AI custom chatbots: Effects on learning performance, cognitive load, and affective variables, Phys. Rev. Phys. Educ. Res. 21, 010147 (2025).
  57. Z. Ji et al., Survey of hallucination in natural language generation, ACM Comput. Surv. 55, 1 (2023).
  58. D. J. Woo, D. Wang, K. Guo, and H. Susanto, Teaching EFL students to write with ChatGPT: Students’ motivation to learn, cognitive load, and satisfaction with the learning process, Educ. Inf. Technol. 29, 24963 (2024).
  59. M. Csikszentmihalyi, Flow: The psychology of optimal experience (1991).
  60. C.-C. Wang and M.-C. Hsu, An exploratory study using inexpensive electroencephalography (EEG) to understand flow experience in computer-based instruction, Inf. Manage. 51, 912 (2014).
  61. U. Konradt, R. Filip, and S. Hoffmann, Flow experience and positive affect during hypermedia learning, Br. J. Educ. Technol. 34, 309 (2003).
  62. Q. Li, S. Xu, Y. Xiong, W. He, and S. Zhou, Comparing the perception of emergency remote teaching experience between physics and nonphysics students, Phys. Rev. Phys. Educ. Res. 19, 020131 (2023).
  63. F. S. Al-Hafdi and S. M. AlNajdi, The effectiveness of using chatbot-based environment on learning process, students’ performances and perceptions: A mixed exploratory study, Educ. Inf. Technol. 29, 20633 (2024).
  64. M. Wang, D. Zhang, J. Zhu, and H. Gu, Effects of incorporating a large language model-based adaptive mechanism into contextual games on students’ academic performance, flow experience, cognitive load and behavioral patterns, J. Educ. Comput. Res. 63, 662 (2025).
  65. S. van Hooft, The Handbook of Virtue Ethics (Acumen, Durham, 2014).
  66. S. Stojiljković, G. Djigić, and B. Zlatković, Empathy and teachers’ roles, Procedia - Soc. Behav. Sci. 69, 960 (2012).
  67. Z. Zhang, Toward the role of teacher empathy in students’ engagement in English language classes, Front. Psychol. 13, 880935 (2022).
  68. S. Meyers, K. Rowell, M. Wells, and B. C. Smith, Teacher empathy: A model of empathy for teaching for student success, Coll. Teach. 67, 160 (2019).
  69. L. Zhi, GuidedEmpathy: Guiding large language models for empathetic response generation with situational awareness, in Proceedings of the 2024 8th International Conference on Computer Science and Artificial Intelligence (ACM, Beijing, China, 2024), pp. 354–360.10.1145/3709026.3709029
  70. M. T. Shih, M. Y. Hsu, and S. C. Lee, Empathy-GPT: Leveraging large language models to enhance emotional empathy and user engagement in embodied conversational agents, in The 37th Annual ACM Symposium on User Interface Software and Technology (ACM, Pittsburgh PA, USA, 2024), pp. 1–3.10.1145/3672539
  71. X. Bai, G. Chen, T. He, C. Zhou, and C. Guo, A holistic comparative study of large language models as emotional support dialogue systems, Cogn. Comput. 17, 71 (2025).
  72. Physics: Grade 8, edited by Q.-C. Peng and S.-B. Huang (People’s Education Press, Beijing, 2016), Vol. 1, p. 147.
  73. H. Liu, Z.-W. Chen, J.-J. Li, and Z.-K. Zhai, Thermal comfort research of air conditioner outlet locations, Home Appliance Sci. Technol. 2, 48 (2022).
  74. S. Park and Y.-C. Chen, Mapping out the integration of the components of pedagogical content knowledge (PCK): Examples from high school biology classrooms, J. Res. Sci. Teach. 49, 922 (2012).
  75. Z. Deng, W. Ma, Q.-L. Han, W. Zhou, X. Zhu, S. Wen, and Y. Xiang, Exploring DeepSeek: A survey on advances, applications, challenges and future directions, IEEECAA J. Autom. Sin. 12, 872 (2025).
  76. C.-C. Chang, C. Liang, P.-N. Chou, and G.-Y. Lin, Is game-based learning better in flow experience and various types of cognitive load than non-game-based learning? Perspective from multimedia and media richness, Comput. Hum. Behav. 71, 218 (2017).
  77. J. Pearce, M. Ainley, and S. Howard, The ebb and flow of online learning, Comput. Hum. Behav. 21, 745 (2005).
  78. M. H. Davis, A multidimensional approach to individual differences in empathy, JSAS Catalog Sel. Doc. Psychol. 10, 85 (1980).
  79. Z. Jiang and M. Jiang, Beyond answers: Large language model-powered tutoring system in physics education for deep learning and precise understanding, arXiv:2406.10934.
  80. C. Heeter, Being there: The subjective experience of presence, in Presence: Teleoperators and Virtual Environments (MIT Press, Cambridge, MA, 1992), pp. 262271.
  81. J. Sweller, Cognitive load theory and educational technology, Educ. Technol. Res. Dev. 68, 1 (2020).
  82. J. Sweller, J. J. G. van Merriënboer, and F. Paas, Cognitive architecture and instructional design: 20 years later, Educ. Psychol. Rev. 31, 261 (2019).
  83. J. Hamari, D. J. Shernoff, E. Rowe, B. Coller, J. Asbell-Clarke, and T. Edwards, Challenging games help students learn: An empirical study on engagement, flow and immersion in game-based learning, Comput. Hum. Behav. 54, 170 (2016).
  84. E. Ortega-Ochoa, M. Arguedas, and T. Daradoumis, Empathic pedagogical conversational agents: A systematic literature review, Br. J. Educ. Technol. 55, 886 (2024).

Outline

Information

Phys. Rev. Phys. Educ. Res. 22, 020131– Published 9 September, 2026
  • Received 30 October 2025
  • Accepted 1 July 2026
Crossmark - Check for updates button
Creative Commons Logo

Published by the American Physical Society under the terms of the Creative Commons Attribution 4.0 International license. Further distribution of this work must maintain attribution to the author(s) and the published article’s title, journal citation, and DOI.

Published by the American Physical Society

Sign In to Your Journals Account

Filter

Filter

Article Lookup

Enter a citation