As simulation-based training becomes increasingly integrated into teacher education, this study explores how AI chatbot simulations and human-actor simulations contribute to the self-efficacy of preservice and in-service early childhood teachers, specifically in the domain of parent–teacher communication within a safe, risk-free environment. 

As parents play a central role in school life and in the daily professional practice of educators, strengthening teacher education in parent–teacher communication is essential. Studies highlight the need to equip teachers with structured tools and practical skills that facilitate effective communication and foster productive partnerships with parents across diverse educational contexts. (Addi-Raccach & Grinshtain, 2021; Azaria et al., 2024).

Simulation-based learning (SBL) has been shown to be an effective pedagogical approach for cultivating interpersonal communication skills and promoting professional development among educators (Dotger et al., 2018; Kasperski et al., 2025), particularly by providing opportunities for repeated practice in controlled settings that mirror authentic professional scenarios. In recent years, the integration of artificial intelligence has expanded the range of simulation technologies available in teacher education, offering innovative, applied tools for developing training processes and enhancing opportunities for experiential learning (Kusmawan, 2023).

There are several reasons for integrating artificial intelligence (AI) chatbots and human actors within SBL. One reason lies in the need to provide comprehensive, high-quality professional training, allowing teachers to develop communication competencies (Bandura, 1997; Frei-Landau & Levin, 2022; Tegero & Mabini, 2025). Consequently, teachers may develop a deeper understanding of the advantages and limitations of AI chatbot simulations, including their implications for assessment practices, the provision of real-time feedback, and processes of professional development. An additional reason for the integration of human actor simulations and AI chatbot simulations in teacher training programs stems from the recognized importance of practicing parent–teacher communication, among other topics such as classroom management and building resilience, within controlled, supportive experiential settings.  Research indicates a significant gap in the practical training of preservice teachers for managing parent–teacher communication; while existing programs provide substantial theoretical knowledge, they offer limited opportunities for safe, repeated, and personalized practical experiences in this complex communicative domain (Theelen et al., 2019). As a result, many preservice teachers enter the profession experiencing stress and low self-confidence, often without access to a protected learning environment that provides opportunities for practice, reflection, and learning supported by feedback and debriefing.

By examining simulation-based approaches that integrate AI chatbot and human-actor simulations, this research provides insights into teachers’ perspectives, experiences, and challenges in developing parent–teacher communication competencies.

The aim of this study is to examine self-efficacy differences: between in-service and preservice teachers, and between those practicing with human-actor simulations versus AI chatbot simulations, measured at three time points. 

 The following research questions guided this study: 

1. To what extent are there differences in self-efficacy regarding parent–teacher communication between the two simulation modalities (human-actor versus AI chatbot), within each study group (preservice and in-service teachers)

2. To what extent, are there differences in self-efficacy regarding parent–teacher communication between preservice and in-service early childhood teachers who practiced SBL with human actors and those who engaged with AI chatbot simulation during the course, within each study group?

3. How do in-service and pre-service teachers perceive the contribution of different types of feedback (AI chatbot vs. human actor) and the nature of interaction in the simulation to the development of their sense of self-efficacy?

This study employed an Explanatory Sequential Mixed-methods Design (Creswell & Plano Clark, 2018) to collect quantitative and qualitative data. Forty-eight participants (26 preservice, 22 in-service teachers) completed a semester-long course in the Education Department, at a teacher training college, during the 2024-2025 academic year. All participants practiced parent–teacher communication scenarios using both AI chatbot simulations and human-actor SBL sessions in a counterbalanced order. Quantitative data included self-efficacy questionnaires demonstrating high internal consistency (Cronbach’s α = .80–.90), and were analyzed using statistical methods. The qualitative analysis included teachers’ reflective comments, enabling statistical comparisons and thematic exploration. In the quantitative phase of the research, data were collected using three online self-report questionnaires administered via Google Forms, designed specifically for this study (Shemer-Elkaim & Landler-Pardo, 2018). The questionnaires included 18 closed questions about teachers' self-efficacy measured on a 5-point Likert-type scale ranging from 1(completely disagree) to 5 (completely agree). The questionnaires were: a pre-simulation questionnaire, administered to all participants prior to practices; a mid-course questionnaire administered between the two simulation sessions and a final questionnaire completed by all participants after practicing both simulation types. Additionally, open-ended questions were added at the end of the questionnaires, allowing participants opportunities to provide further views regarding the simulations. Questionnaires were based on a review of empirical literature pertaining to simulation-based training and self-efficacy. Data collection took place during the second semester of the 2024–2025 academic year, at predetermined days and hours. In the qualitative stage, data were based on teachers’ reflective comments provided in response to open-ended questions. These questions invited the participants to elaborate on their experiences and perceptions regarding the simulation practices. Special attention was given to the choice of words used. Responses to the open questions were read by each researcher to obtain a holistic view of the comments. Recursive reading by the researchers produced a common ground of themes. This study employed a purposive sampling approach to identify self-efficacy patterns within a small-scale group (N=48) enrolled in a specific program and arrive at propositions that can be examined in comparable contexts. This sample is purposeful as participants were chosen based on accessibility, field of study, and experience, which provided data. Students were notified about the purpose of the research and its voluntary basis. Anonymity was secured, and identification details were not included. The research abides by all the ethical protocols certified by the Ethics Committee of the college.

Results revealed no significant differences in self-efficacy between human-actor and AI chatbot simulations across all measurement points (pre-course, mid-semester, and post-course). However, in-service teachers consistently reported significantly higher self-efficacy than preservice teachers at both pre and post-course measures. Qualitative findings showed that both simulation modalities supported professional identity development, participants' practical competence, and enhanced communication strategies in practicing complex teacher-parent communication. AI chatbots were particularly valued for accessibility, and opportunities for repeated practice, whereas human-actor simulations were perceived as providing greater authenticity and emotional depth. Differences emerged between groups as in-service teachers, with greater teaching experience, emphasized increased confidence and self-efficacy engaging with parents, while preservice teachers reported stronger perceived gains in practical communication competencies and skill acquisition. Overall, 93% of participants recommended the combined approach. Participants expressed strong overall support for the dual-modality approach as a means of bridging the existing gap between theoretical knowledge and the complex realities of the classroom. Findings support a hybrid simulation model combining AI chatbot and human actor simulations to strengthen teacher–parent communication, capitalize on teachers' strengths and inform scalable, cost-effective teacher preparation (Flavian et al., 2024; Spencer et al., 2019;), However, implementation of hybrid simulation models faces challenges such as the need for ongoing teacher training, resistance to adopting new technologies, and the lack of technological infrastructure in some educational institutions (Gonçalves, 2025).

Addi-Raccah, A., and Grinshtain, Y. (2021). Teachers’ professionalism and relations with parents: teachers’ and parents’ views. Research Papers in Education, 37, 1142-1164. Azaria, A., Azoulay, R., and Reches, S. (2024). ChatGPT is a remarkable tool—For experts. Data Intelligence, 6(1), 240–296. https://doi.org/10.1162/dint_a_00235 Bandura, A. (1997). Self‐efficacy: The exercise of control. Freeman. Creswell, J. W., and Plano Clark, V. L. (2018). Designing and Conducting Mixed Methods Research (3rd ed.). Thousand Oaks, CA: SAGE. Dotger, B. H., Harris, S., and Hansel, A. (2018). Revealing the complexity of parent–teacher interactions through simulated practice. Teaching and Teacher Education, 72, 1–14. Flavian, H., & Levin, O. (2024). Using simulation-based learning to inform preservice teachers’ professional development. Teaching Education, 35(2), 145–161. https://doi.org/10.1080/10476210.2023.2240716 Frei-Landau, R., and Levin, O. (2022). The virtual Sim(HU)lation model: Conceptualization and implementation in the context of distant learning in teacher education. Teaching and Teacher Education, 117, 1–14. https://doi.org/10.1016/j.tate.2022.103798 Gonçalves, B.F. (2025). Artificial intelligence in teacher training: benefits, challenges and tools. In Conference on Education and New Developments (END Conference). Lisboa: World Institute for Advanced Research and Science. p. 331-335. Kasperski, R., Levin, O., and Hemi, M. E. (2025). Systematic Literature Review of Simulation-Based Learning for Developing Teacher SEL. Education Sciences, 15(2),129. https://doi.org/10.3390/educsci15020129 Kusmawan, U. (2023). Redefining Teacher Training: The Promise of AI-Supported Teaching Practices. Journal of Advances in Education and Philosophy, 7(09):332-335 DOI:10.36348/jaep.2023.v07i09.001 Shemer-Elkaim, T., and Landler-Pardo, G. (2018). The Simulation in Educational Practice Center: Seminar research report (Internal report). Authority for Research and Evaluation, Kibbutzim College of Education. Spencer, S., Drescher, T., Sears, J., Scruggs, A. F., and Schreffler, J. (2019). Comparing the Efficacy of Virtual Simulation to Traditional Classroom Role-Play. Journal of Educational Computing Research, 57(7), 1772-1785. https://doi.org/10.1177/0735633119855613 Tegero, M.C., and Mabini, J.P. (2025). AI Chatbot Simulations in Teacher Training: Core Teaching Competencies Developed Through Virtual Practice. Journal of Teaching and Learning, 19(4), 216-232. https://doi.org/10.22329/jtl.v19i4.10087www.jtl.uwindsor.ca216 Theelen, H., van den Beemt, A., and Brok, P. D. (2019). Classroom simulations in teacher education to support preservice teachers’ interpersonal competence: a systematic literature review. Computers and Education, 129, 14-26. https://doi.org/10.1016/j.compedu.2018.10.015