Reflective writing is an essential component of the pre-service teacher preparation process as they transition into the teaching profession. Reflection enables pre-service teachers to identify their strengths, recognize areas for improvement, and develop professionally (Cohen-Sayag & Fischl, 2012; Boud, 2001; Dumlao & Pinatacan, 2019). Reflective writing in teacher education is most beneficial when it reaches deeper levels of thinking and engages the highest forms of critical reflection, using reflection-based interpretations to planfuture actions (Mezirow, 1990). Research has shown that reflective writing often becomes merely a descriptive account of events, lacking depth and critical judgment (Chen et al., 2016). Students may struggle to clearly articulate what they have learned, fail to think critically, and neglect to evaluate their experiences(Spaulding and Wilson, 2002).

Deep reflection can be achieved through dialogues based on questions. Ghaye (2010, p. 42) calls this type of dialogues “reflective conversations,” which may lead to teachers to learn from their own educational experience and question existing educational values. In reflective conversations, the one who asks questions plays a crucial part. According to Pendlebury (1995) this person is ‘dialogical other’, and his or her role is to elicit the practical arguments and ‘theories’ that underpin the teacher’s work (Cited in Ghaye, 2010).The ‘dialogical other’ is supportive and adds structure to the reflective conversation; he/she does not necessarily focus on weakness but ensure balance in judgement and strengthen the positive things. In Bain et al. (1999) study with dialogue reflections 66% of the subjects improved their reflections from reporting to reasoning.

Since the emergence of ChatGPT, AI has revolutionized education; it has been used to produce learning materials, assess, give feedback, and act as a virtual conversation assistant in higher education (Dempereet al., 2023; Essel et al. 2022; Wang et al., 2017). Most research on ChatGPT focuses on lesson planning, personalized learning, feedback, and content creation, but little is known about its role in reflective writing. It is unclear how interacting with ChatGPT affects pre-service teachers’ reflective writing. This study will explore this gap.

This study is grounded in reflective practice theory. ChatGPT is viewed as a dialogical other (Pendlebury, 1995 cited in Ghaye, 2010) that supports reflection through asking questions, providing feedback, and dialogue. ChatGPT serves as a scaffold that helps pre-service teachers express their thoughts, deepen metacognitive awareness, which aligns with guided reflective dialogue models. Combining the potential of ChatGPT to brainstorm ideas into reflective writing, can address the challenges in reflective thinking.

This study explores ChatGPT’s potential to support pre-service teachers in developing their reflective writing by generating ideas, offering personalized assistance, and providing feedback. Our study aims to contribute to the growing understanding of how AI can enhance reflective practices in teacher education. Specifically, the study investigates how ChatGPT influences pre-service teachers’ reflective writing and their vocabulary use. The research questions are as follows:

RQ1: How do pre-service teachers communicate with ChatGPT for reflective writing?

RQ2: How does this interaction with ChatGPT as a writing assistant influence the content and structure of their reflective writing?

These research findings provide insight into the use of ChatGPT in the reflective writing process, as well as its potential for future development and practical application. For researchers, the findings can serve as a foundation for further studies and experimentation. The results may benefit teacher education programs; program coordinators and educators may consider integrating ChatGPT into pre-service teachers’ reflective practice to enhance their level of reflection and focus. Additionally, the results could be interesting for AI developers to create special AI tools for reflective writing.

To answer the research questions, a mixed-method study was conducted. 15 pre-service teachers, who were enrolled in the BA–MA integrated Primary Teacher Education Program at Iakob Gogebashvili Telavi State University (Georgia), were involved in the study. Pre-service teachers' reflective essays—first written independently (Reflective essay 1, RE1), then written with the help of ChatGPT's guiding questions on the same topic (Reflective essay 2, RE2) — were analysed. Both sets of reflective essays, along with the participants' conversation logs with ChatGPT, were collected and examined. The quantitative study design focused on analyzing the following: • Word Count analysis of RE1 & RE2: This comparison aimed to determine how many words the participants wrote independently versus with the help of AI, assessing whether the word count increased when ChatGPT was involved. SPSS 24 was used to calculate descriptive statistics (mean, median, standard deviation) and conduct inferential statistics (t-test and p-value). • Prompts Count: Quantitative analysis of the prompts evaluated how frequently participants interacted with ChatGPT. • Count of the same lexical units in RE2 & logs; Overlap of lexical units between RE2 and ChatGPT conversation logs: To measure the influence of AI on vocabulary and ideas, repeated lexical units were examined (Voyant Tools https://voyant-tools.org) in RE2 that appeared in the ChatGPT logs, indicating the extent to which participants incorporated AI-generated language and concepts (Poldner et al, 2014). A qualitative study design was used to analyze the following: • Content analysis of RE2 and logs; Reflection topics suggested by ChatGPT and their occurrence in RE2: For content analysis, it was necessary to identify the questions and topics in the ChatGPT logs and then trace these topics in the participants’ Reflective Essay 2 (RE2) (Chen et al., 2016; Zhang et al., 2023). • Prompts Analysis: Understanding how participants communicated with AI and how this interaction related to their writing was essential. This included examining whether participants addressed ChatGPT with requests or questions and whether they responded to AI’s questions within the conversation logs. To analyze reflection topics in the ChatGPT conversations, all logs were coded and examined using thematic analysis by two researchers, one of whom was an external researcher. To reduce potential bias in coding and qualitative judgment, both researchers independently reviewed and coded the reflective essays and ChatGPT logs. The types of prompts (questions, requests, responses/dialogic exchanges), as well as participants’ introductions to AI, were also thematically analyzed.

The aim of the study was to explore ChatGPT’s impact on pre-service teachers’ reflective writing to understand how AI can be used to improve pre-service teachers' reflective writing. The word count comparison of independently written essays and ChatGPT-assisted showed that the effect might be not only significant but also meaningful in practice, as it suggests that AI support leads to relatively longer reflections. The standardized AI output of a set of questions heavily relies on its training data, as also noted by Buchholtz and Huget (2024). ChatGPT acts as a “dialogic other”, asks questions, provides feedback and adds structure to the writing. ChatGPT’s questions might inspire students to reflect by introducing new ideas that influence their writing and thinking. Students may adopt concepts, structures, and vocabulary from the AI. ChatGPT may provide a set of questions personalized to pre-service teachers’ needs. However, reviewing all questions generated by ChatGPT may be difficult, as it is time-consuming and cognitively demanding. Additionally, AI could offer instant feedback and supply with model sentences to support reflective thinking. Students should receive proper guidance before writing a reflective essay with AI. They need to understand that the scope of the AI’s responses can be controlled through carefully structured prompts. A clear introduction to the task is crucial for ensuring effective guidance from ChatGPT. Using ChatGPT as a guiding partner for pre-service teachers’ reflective writing requires careful consideration. On one hand, ChatGPT might enhance reflection with different topics related to teaching profession, on the other hand it could impact pre-service teachers’ writing in many ways. Students may lose their voice, copy the sentences from the logs, or follow suggested structure and opt out answering deep questions. It is important to study what type of AI guidance help pre-service teachers achieve deeper layers of thinking.

Bain, J.D., Ballantyne, R., Packer, J., & Mills, C. (1999). Using journal writing to enhance student teachers’ reflectivity during field experience placements. Teachers and Teaching: Theory and Practice, vol. 5, no. 1, 51-73. Boud, D. (2001). Peer Learning in Higher Education: Learning from and with Each Other. London: Kogan Page. Buchholtz, N. & Huget, J. (2024). ChatGPT as a reflection tool to promote the lesson planning competencies of pre-service teachers. 2024, Conference contribution - Article for conference in: Proceedings of the 17th ERME Topic Conference MEDA 4 – University of Bari Aldo Moro, Italy, 3-6 September 2024 : 129-136, University of Bari Aldo Moro Chen, Y., Yu, B., Zhang, X., & Yu, Y. (2016). Topic modeling for evaluating students' reflective writing: A case study of pre-service teachers' journals. ACM International Conference Proceeding Series, 25-29-April-2016, 1–5. https://doi.org/10.1145/2883851.2883951 Cohen-Sayag, E., & Fischl, D. (2012). Reflective writing in pre-service teachers' teaching: What does it promote? Australian Journal of Teacher Education, 37(10). https://doi.org/10.14221/ajte.2012v37n10.1 Dempere, J., Modugu, K., Hesham, A., and Ramasamy, L. K. (2023). The impact of ChatGPT on higher education. Front. Educ. 8:1206936. doi: 10.3389/feduc.2023.1206936 Dumlao, R. P., & Pinatacan, J. R. (2019). From practice to writing: Using reflective journal instruction in enhancing pre-service teachers' professional development. International Journal of Instruction, 12(4), 459–478. https://doi.org/10.29333/iji.2019.12430a Essel, H., Vlachopoulos, D., Tachie-Menson, A., Johnson, E., and Baah, P. (2022). The impact of a virtual teaching assistant (chatbot) on students’ learning in Ghanaian higher education. Int. J. Educ. Technol. High. Educ. 19, 1–19. doi: 10.1186/s41239-022-00362-6 Ghaye, T. (2010). Teaching and Learning through Reflective Practice: A Practical Guide for Positive Action (2nd ed.). Routledge. https://doi.org/10.4324/9780203833322 Mezirow, J. (1990). How Critical Thinking Activates Transformative Learning. San Francisco: Jossey-Bass. Poldner, E., Van der Schaaf, M., Simons, P. R. J., Van Tartwijk, J., & Wijngaards, G. (2014). Assessing student teachers' reflective writing through quantitative content analysis. European Journal of Teacher Education, 37(3), 348–373. https://doi.org/10.1080/02619768.2014.892479 Spalding, E., &Wilson, A. (2002). Demystifying Reflection: A Study of Pedagogical Strategies that Encourage Reflective Journal Writing. Teachers College Record, 104, 7, 1393-1421 Wang, Y., Petrina, S., and Feng, F. (2017). VILLAGE—virtual immersive language learning and gaming environment: immersion and presence. Br. J. Educ. Technol. 48, 431–450. doi: 10.1111/bjet.12388 Zhang, C., Schießl, J., Plößl, L., Hofmann, F., & Gläser-Zikuda, M. (2023). Evaluating Reflective Writing in Pre-Service Teachers: The Potential of a Mixed-Methods Approach. Education Sciences, 13(12). https://doi.org/10.3390/educsci13121213