Using a chatbot as a pedagogical mediator for educational coexistence in teacher education

Introduction

There is no doubt that, in recent years, the use of artificial intelligence (AI), and especially of chatbots, has been growing and becoming consolidated in different areas, particularly in the field of education (Xue & Crompton, 2024). It is stated that AI offers a strategic opportunity to transform education, since it enriches teaching, strengthens innovation and democratizes access to knowledge (Cordón García, 2023).

In Higher Education, the rapid growth of AI is grounded in the connectivist learning theory for the digital age, which conceives learning mediated by this technology as a networked mechanism in which academics, students, tools and structures intervene (Carrillo et al., 2025).

According to Edel Navarro (2025), AI is a catalyst that strengthens the quality of academic debate, since it takes the teacher out of routine and helps them discern, contextualize and make good decisions.

Previous studies point to the emerging need to strengthen digital competencies in the university context. However, they also emphasize that the empirical base on the impact and development of AI in this field is still incipient (García et al., 2022).

Moreover, its incorporation must be evaluated from a pedagogical, ethical and formative perspective, especially in teacher education. In this regard, it is worth recalling that such education is a key process that prepares future teachers to face the technological and pedagogical challenges that arise in the classroom, and guarantees quality teaching aligned with current global demands (United Nations Educational, Scientific and Cultural Organization [UNESCO], 2024; Xue & Crompton, 2024).

The progressive incorporation of these tools into multiple educational settings, and specifically in Higher Education, has facilitated personalized learning experiences, since it makes it possible to adapt content, activities and strategies to the students' level of cognitive development. Likewise, it helps foster more active, autonomous and contextualized learning (Azevedo et al., 2022; Vaca Montenegro et al., 2025).

Based on the above, the use of chatbots can enrich teaching, broaden access to knowledge and diversify ways of learning. Hence, this study seeks to answer the following question: what was the contribution of the conversational chatbot, as a pedagogical mediator, to the learning of students in training?

This resource is defined as a type of AI-based conversational chatbot capable of interacting with people through natural language (Ashfaque et al., 2020). From the educational standpoint, interaction with students supports their formative process, since it offers comprehensive didactic support (Huang et al., 2022).

The most frequent applications in education focus on resolving questions about academic content, assignments and assessments (Martínez, 2025). Thus, conversational educational chatbots emerge as a tool with the potential to offer personalized support, immediate access to information, quick feedback and reflective guidance in teaching and learning processes (Holmes et al., 2019).

One of the most relevant aspects is that the use of conversational chatbots in education should not be analyzed only from a technological viewpoint, but also from a pedagogical perspective. For this reason, it is important to incorporate this tool into the curriculum, specifically in university training courses.

In the context of teacher education, this study deliberately integrated AI through an educational chatbot in a course on educational coexistence. The resource was used as a pedagogical mediator to work on relevant concepts and aspects of the topic, which helped foster student interaction. Recent studies recognize the potential of these chatbots for educational coexistence and for generating healthy prosocial behaviors; they are based on language models that foster competencies associated with social-emotional learning, relational skills and conflict resolution, among others (Liu et al., 2026; Zou et al., 2024).

Therefore, in teacher education, the educational chatbot can be conceived as a pedagogical mediator, since it facilitates and motivates the teaching and learning process of future education professionals. From a constructivist perspective, mediation is understood as a strategic and intentional accompaniment aimed at developing higher cognitive skills, which enhances reflection, analysis and self-regulation of learning; in other words, it enables a meaningful and active construction of knowledge (Mederos Mollineda et al., 2026).

Likewise, it is essential to strengthen teacher education in the pedagogical and critical use of AI, since its effectiveness as a mediator will depend on the teacher's ability to articulate it in their professional practice (Vargas Gómez, 2025). Therefore, teachers must be prepared in the design and use of these tools to foster more meaningful learning. Asrowardi et al. (2026) highlight the importance of teachers actively participating in the didactic incorporation of these resources. In this sense, it is necessary to promote in them a dynamic and open attitude to change, which allows them to adapt their pedagogical practices to challenging and constantly transforming contexts, for which they must mobilize their own digital competencies.

When teachers guide the use of educational chatbots as an intentional pedagogical mediation, the development of critical thinking, metacognition and self-regulation of learning in future professionals is favored, which enables a better understanding of the topics (Vargas Gómez, 2025). However, chatbots do not replace the teaching role, but rather improve accessibility and personalization of learning. The key is for teachers to act as guides and mediators in the strategic integration of these resources, so that teaching does not lose its human dimension.

Likewise, it is key to train students in the use of this tool: teach them to interact critically with it, to identify when and how it can be useful and to question its answers when necessary, in order to develop strategic, critical and ethical thinking.

In short, by critically delving into teacher education, it is evident that AI and the use of chatbots as a pedagogical tool can enrich student education, increase motivation and favor comprehensive development. This demands an educational practice committed to updating teaching methods and aligned with the innovations required by current university contexts.

From this framework, the objective of the study is to analyze the contribution of an educational conversational chatbot, as a pedagogical mediator specialized in educational coexistence, in teacher training students.

Methods

Design and participants

The study was developed under a mixed methods research approach, with a descriptive-interpretive predominance (Creswell & Plano Clark, 2018; Johnson et al., 2007). This design allowed a comprehensive understanding of the phenomenon studied by combining the systematic measurement of change with an in-depth exploration of participants' meanings and experiences (Tashakkori & Teddlie, 2010).

Quantitative data were obtained at three measurement points: baseline, process and closing. These were complemented and deepened through a qualitative focus group. This articulation made it possible to describe the magnitude of change and, at the same time, to understand the mechanisms and processes associated with the results obtained (Creswell & Plano Clark, 2018).

The quantitative component adopted a longitudinal repeated-measures design, aimed at describing the evolution of self-perceived pedagogical readiness, chatbot usage patterns and evaluation of the resource.

Regarding the qualitative component, it was approached through an inductive-deductive thematic analysis, which made it possible to delve into the emerging categories drawn from the students' own voices.

The study was carried out at a Chilean university during the first semester of the 2026 academic year. The population consisted of 70 first-year students in the Primary Education Pedagogy program with a specialization, who were taking the course Educational Psychology. Within this course, the content related to educational coexistence was worked on in particular.

Response rates by instrument were as follows: baseline survey (n = 50; 71.4 %), process survey (n = 47; 67.1 %) and exit survey (n = 48; 68.6 %). The focus group consisted of seven students from the program.

The inclusion criteria were: (a) having active enrollment in the course; (b) participating voluntarily, after providing informed consent; and (c) having used the chatbot at least once. All participants signed the informed consent before the start of the study.

Pedagogical and technical aspects considered in the design of the conversational chatbot

The pedagogical intervention was based on the design, implementation and evaluation of an educational chatbot specialized in educational coexistence, called Convivencia educativa: Aula y emoción (Educational coexistence: Classroom and emotion). This conversational agent was organized and validated according to the characteristics and contextual needs of the course. To this end, a disciplinary corpus was integrated, made up of theoretical foundations on educational coexistence, international and national regulatory frameworks, as well as a compilation of situated and meaningful pedagogical strategies for practical work with children.

The specific validation of the resource was carried out through a user test, corresponding to a pilot, administered to a group of 10 first-year Pedagogy students.

The educational conversational chatbot was designed taking into consideration several elements, among them: the pedagogical objective, the modeling of the conversation according to a Socratic, directive or collaborative pedagogical tone, the AI technology employed and the evaluation of its effectiveness (Smutny & Schreiberova, 2020). These elements constitute relevant aspects for the design of this type of digital resource.

From a pedagogical perspective, an educational conversational chatbot specialized in educational coexistence can be understood as a resource supported by AI that, through intentional pedagogical interactions, favors analysis, reflection and learning about aspects related to coexistence, such as interpersonal relationships, empathy, conflict resolution and socio-emotional competencies. In this sense, its purpose is not limited to providing information, but it can also play a role as a pedagogical mediator by promoting active student participation, fostering critical reflection on their experiences and favoring the identification of prosocial responses and concrete strategies to address situations of educational coexistence in challenging educational contexts.

Likewise, it is necessary to differentiate this type of chatbot from others of a more generic nature. Although the latter may also present conversational capabilities, an educational chatbot specialized in educational coexistence uses conversational interaction as an intentional pedagogical mediation strategy aimed at reflection and analysis of specific coexistence situations. To this end, it can articulate theoretical elements, public policies, conflict resolution strategies and other content related to the topic.

Technical structure and platform used

Regarding the technical structure, according to its scope and capacity, this resource can be placed within the category of narrow or weak artificial intelligence (Artificial Narrow Intelligence, ANI), designed to perform specific tasks efficiently. For the purposes of this study, the ChatGPT platform was used, through the modality of creating a personalized chatbot available in the GPT Store for direct interaction with students. The language model used was GPT-4o, whose capabilities include reasoning, multimodal processing and programming.

Pedagogical foundation of the chatbot

From a pedagogical perspective, the educational conversational chatbot is relevant to address educational coexistence in the training of future teachers, because it allows generating focused, reflective and contextualized interactions based on situations related to the school system. Through dialogue with the chatbot, students can analyze conflicts, explore theoretical foundations, make decisions and reflect on different alternatives for action, which favors the development of socio-emotional competencies, among others.

In this way, the resource operates as a pedagogical mediator that articulates teaching with the promotion of active, reflective and autonomous learning about educational coexistence in current training scenarios.

Among the pedagogical principles considered is socio-constructivist learning. In this regard, Holmes (2020) states that, in the university context, this approach conceives the student as an active participant in the construction of knowledge, by favoring methodologies that enhance active learning, reflection and participation.

Likewise, autonomous learning was considered. According to Zimmerman (2002), this requires future teachers to develop self-regulation processes that allow them to set goals, select strategies, monitor their performance and reflect on the results obtained.

In the same way, mediated learning was incorporated, understanding pedagogical mediation as an element that fosters active, collaborative and student-centered learning processes, particularly when resources and pedagogical strategies are intentionally aligned with the program's training objectives (Méndez Forns & Vázquez Zubizarreta, 2025).

Finally, situated learning was considered, which can contribute to generating authentic training experiences by linking the proposed content with situations and competencies relevant to professional performance (Hamman-Fisher & McGhie, 2023).

Examples of interaction with students

Some examples of interaction between students and the conversational chatbot were the following:

Example 1. Searching for and analyzing information on educational coexistence

Student: "Could you find me articles that talk about the importance of educational coexistence in the classroom and how it contributes to teacher education?"

Chatbot: "Yes, I found several articles that fit what you are looking for very well. I especially recommend the following: 1. The most relevant to your topic: Loubies, L., & Valdivieso Tocornal, P. (2023). Formación inicial docente en convivencia escolar..."

Example 2. Conceptual exploration of emotional development

Student: "Which authors write about emotional development in students?"

Chatbot: "Yes, to work on students' emotional development, there are several authors who can be useful to you, but it is worth distinguishing what each one contributes. Daniel Goleman works on emotional intelligence, especially self-awareness, self-regulation, motivation, empathy and social skills. He is useful for relating emotional development to relationships and learning. Rafael Bisquerra is especially relevant for education, because he addresses emotional education and emotional competencies. His approaches make it possible to link emotional awareness, regulation, emotional autonomy and social competencies..."

Ultimately, this type of resource can help strengthen the comprehensive education of future education professionals from a critical, reflective and ethical perspective.

Instruments

The implementation stages were: (a) administration of the baseline survey; (b) induction into the use of the chatbot with explicit ethical guidelines; (c) autonomous use linked to course activities during the semester; (d) administration of the process survey; (e) administration of the exit survey; and (f) holding of the focus group.

Three perception surveys and a focus group guide were designed, all validated through expert judgment in teacher education and educational technologies.

The baseline survey included five-level Likert questions (1 = Not at all prepared, 5 = Very prepared) on self-perceived readiness to integrate coexistence factors; and open-ended questions about known strategies, guidance resources used and prior experience with digital tools.

The process survey included closed questions about: frequency of use; perceived usefulness, relevance and contextualization of the information; types of guidance most valued, degree of complementarity with teacher education and, lastly, intention to recommend.

The exit survey maintained a structure parallel to the initial one, additionally incorporating: overall evaluation of the chatbot, new strategies learned, guidance resources used in the last month and suggestions for improvement. Ultimately, this structure allowed a comparative analysis between the start and the end.

The focus group used a guide of seven guiding questions on: overall experience of use, moments of greatest pedagogical significance, most valued guidance, understanding of coexistence factors, contributions to teacher education, difficulties and suggestions. The session was recorded, transcribed and subjected to thematic analysis.

Quantitative data were analyzed using descriptive statistics with an emphasis on: (a) distribution of absolute and relative frequencies by scale level; (b) comparative analysis of the start and exit for the variable of self-perceived pedagogical readiness, grouping high and medium-low levels; (c) description of longitudinal trends in chatbot use, usefulness and evaluation; and (d) analysis of multiple frequencies for multiple-choice variables. It should be noted that, given the descriptive nature of the study, the analysis focused on frequency distributions by scale level, without applying tests of statistical significance or additional measures of central tendency and dispersion, which is acknowledged as a methodological limitation to be considered.

Focus group data were analyzed using inductive-deductive thematic analysis (Braun & Clarke, 2006), with the support of the software atlas.ti 7.5. The process included: (a) transcription and familiarization with the data; (b) generation of initial codes; (c) searching for and reviewing thematic categories; (d) defining and naming categories; and (e) producing the analytical report. Deductive categories were established from the study's objectives and inductive ones from the emerging data, verifying saturation through iterative review.

The triangulation of instruments (initial survey, process survey, exit survey and focus group) was the main strategy for ensuring the credibility and transferability of the findings (Guba & Lincoln, 2000; Sandín, 2003). Triangulation between quantitative and qualitative data made it possible to identify patterns of convergence, complementarity and divergence among the sources (Marradi et al., 2014).

Results and discussion

The results on the use of the conversational chatbot in educational coexistence are presented through the articulation of quantitative and qualitative data, across five categories of analysis derived from the study's objective and from the patterns emerging in the data: (1) appropriation and progressive use of the chatbot, (2) contribution to learning about educational coexistence, (3) development of metacognitive and reflective processes, (4) ethical and professional education in the use of AI, and (5) technical limitations and projections for improvement.

Regarding the category of appropriation and progressive use of the chatbot, the results of the process survey (n = 47) stand out, revealing a pattern of frequent and sustained chatbot use throughout the semester. Of respondents, 66.0 % used it 2 to 3 times in the last three weeks of measurement, 17.0 % once a week or more, and 17.0 % only once.

As for perceived usefulness, 68.1 % rated it as very useful and 25.5 % as somewhat useful, accumulating 93.6 % positive ratings, and 68.1 % indicated that the chatbot always provided relevant and contextualized information, with an additional 29.8 % reporting most of the time.

In the exit survey, 81.3 % of students (n = 39 of 48 participants) mention the chatbot as one of their main pedagogical support resources during the last month, being the most frequently reported resource alone or in combination with others, followed by the teacher (47.9 %), peers (43.8 %), books (22.9 %) and social media (22.9 %). This result contrasts markedly with the initial survey, in which the chatbot did not appear among the declared guidance resources, which were teachers, peers and the internet.

Furthermore, as a comparative summary that does not duplicate the figures just reported in Table 1, the results of the three surveys applied (initial, process and exit) are compared in relation to the evolution in the use of the chatbot.

Likewise, the thematic analysis confirms and deepens the quantitative pattern, describing a trajectory from an initial exploration marked by technical difficulties and difficulties in formulating prompts toward an autonomous and strategic incorporation of the resource. One participant stated: "At first it was kind of complicated, but then I used another device and it worked perfectly, it was a big help because not only were citations found for this course, but it also helped me with my other assignments, and to orient myself better in other activities in other subjects". Another noted: "At first I didn't really understand how to use it, but then with a good prompt it gave me good suggestions, especially for searching for articles, which takes a lot of time to look for, and when I asked it, it gave me several quality articles".

In short, the pattern of progressive appropriation identified achieved a sustained increase in frequency of use and evaluation, which is consistent with what Wollny et al. (2021) state, who affirm that the pedagogical potential of educational chatbots is realized gradually, as students develop skills to interact strategically with the resource. Along the same lines, Zawacki Richter et al. (2019) point out that the perception of AI's usefulness in Higher Education depends on both instructional design and the teacher's pedagogical support.

The fact that 93.6 % of students rated the chatbot's usefulness positively in the process survey and that 81.3 % identified it as their main support resource at the end suggests that intentional and guided incorporation, with explicit ethical guidelines from the outset, is key to effective appropriation. This resonates with Kasneci et al. (2023) regarding the need to integrate AI into educational settings in a pedagogically grounded way, and not only as a tool for accessing information.

Regarding the contribution to learning about coexistence, the comparison between the initial and exit surveys reveals a significant shift in self-perceived pedagogical readiness to integrate socio-affective factors. In the initial survey, 8.0 % reported feeling poorly prepared (option 2), 24.0 % moderately prepared (option 3), 54.0 % fairly prepared (option 4) and 14.0 % very prepared (option 5). In the exit survey, no student reported low levels of readiness (options 1 or 2), 8.3 % were at a medium level, 47.9 % at a high level and 41.7 % at a very high level (see Table 2).

This result reveals an increase of 23.5 percentage points in high levels of readiness (from 68.0 % to 91.5 %) and the complete disappearance of low levels. Regarding the type of guidance most valued during the process, 66.0 % highlighted theoretical references or authors as the most useful, followed by suggestions of pedagogical strategies (48.9 %), and support in regulations and public policies (21.3 %).

Discussing this aspect delves into the mechanisms through which the chatbot contributed to understanding educational coexistence. Participants highlight three main uses: theoretical references and authors, understanding of the regulatory framework, and development of pedagogical strategies. One of them stated: "For the integrated project, which is for another course, I searched a lot about policies and also about the socioemotional topic. It gave me a lot of information, and then I would go to the pages, read, and that again made it easier and left the idea much clearer for me". Another commented: "I used the chatbot to see what the authors said and at the same time to reinforce the concepts [...]". A third expressed: "The chatbot is kind of at the service of these learnings ... it was a help to understand them more deeply. Of course, I didn't depend 100 % on the chatbot, but it was a help."

In short, the chatbot contributed as scaffolding for understanding coexistence. The increase of 23.5 percentage points in high levels of self-perceived readiness (from 68.0 % to 91.5 %) and the complete disappearance of low levels in the exit survey constitute substantive indicators of the intervention's impact. These results are consistent with the theoretical framework of cognitive scaffolding (Luckin & Holmes, 2016; Kuhail et al., 2023), which holds that technologies can act as supports for understanding complex concepts when articulated with teachers' pedagogical mediation.

The distinction that students make between the teacher's "human and contextualized language" and the chatbot's "technical and general language" is especially revealing: it suggests that both modes of mediation are perceived as complementary rather than mutually exclusive. This finding coincides with Hwang and Chang (2021), who identify complementarity with teacher mediation as one of the conditions that enhance the pedagogical value of chatbots, and Holmes et al. (2019), who underscore the importance of the intentional design of their curricular integration.

Regarding the development of metacognitive and reflective processes in the process survey, 97.9 % stated that the chatbot complements their teacher education (strongly agree: 55.3 %; agree: 42.6 %). The open-ended responses in the exit survey revealed a growing articulation among theory, regulations and pedagogical practice, as well as the use of the chatbot to organize prior ideas, contrast information and deepen understanding.

The thematic analysis identified emerging metacognitive processes that were not explicitly planned in the design of the intervention. Students describe processes of planning, monitoring and evaluating their own learning: "I searched a lot about coexistence and many authors came up, some older ones that are the foundation, and others more up to date. That helped me realize that reading is important, but so are understanding, comparing and reflecting on a topic". Another expression was: "I did kind of a combination between the chatbot and AI in general: the chatbot gave me the ideas and then general AI was more for going deeper".

Therefore, the development of metacognitive processes was a noteworthy aspect. These metacognitive processes, not explicitly planned in the intervention, constitute one of the most relevant findings of the study. This result aligns with research that highlights the potential of AI technologies to promote reflection on one's own learning when their use is pedagogically mediated (Taub et al., 2021). Future teachers described processes of planning, monitoring and evaluating their learning when interacting with the chatbot: contrasting information, verifying authors, reformulating prior ideas. These processes are constitutive of metacognition, as defined in the specialized literature (Flavell, 1979; Schraw & Dennison, 1994).

The fact that they emerge in an unplanned way suggests that the design of the experience, by incorporating the chatbot into complex and authentic tasks, generated the conditions for students to play an active role in regulating their own learning. This is consistent with the constructivist view of pedagogical mediation (Mederos Mollineda et al., 2026) and with the principles of active and self-regulated learning that underlie quality teacher education (Vaillant & Marcelo, 2015).

Regarding ethical and professional education in the use of AI in the process survey, 83.0 % responded that they would "totally" recommend the use of the chatbot, while the remaining 17 % considered that they "maybe" would. In the exit survey, 75.0 % gave very positive ratings (options 4 and 5), with an average of 4.30 out of 5. The open suggestions show a cross-cutting concern for the responsible, non-dependent use of the resource.

Likewise, ethical reflection emerges with particular force in the conversation with the students, constituting one of the most consistent and cross-cutting categories of the analysis. In a first opinion it is expressed: "Using the chatbot is not like using it always for everything, but in a responsible way. If I want to apply a certain activity and it doesn't work, I look at why it doesn't work and ask the chatbot to help me reshape it"; in another testimony: "AI is not the one who is going to be teaching classes, we are. We are the ones who have to learn, know the policies, be able to give quality classes. Learning always has to be your own. AI can be very intelligent, but we also have to do it ourselves, especially when it comes to emotions and coexistence"; a third opinion points out that "although it is a super good resource, we must be clear that we are teachers in training and it is a tool, but not one to abuse". Another stated: "It is essential that we be ethical, because we all use AI, and as a teacher I will have to be aware. A good educational chatbot is useful for teaching students to use AI as a support tool and not as something that does your homework or makes things too easy for you." In this way, the ethical dimension stands out as a central formative component in the use of AI.

The high coherence among the instruments regarding ethical concern for the responsible use of AI is one of the study's most significant results. In contrast with research reporting uncritical tendencies in the use of generative AI in Higher Education (Tlili et al., 2023), the students in this experience show a clear understanding of AI as a support tool, with an emphasis on professional judgment, source verification and plagiarism prevention.

This result is consistent with UNESCO's (2023) guidelines on AI literacy, which state that education systems need to train people in the critical and responsible use of these technologies. From the perspective of teacher education, the finding takes on special relevance. Specifically, future teachers must be not only competent users of AI but also trainers capable of guiding its use pedagogically and ethically in school settings (Celik et al., 2022; Darling Hammond et al., 2020; Asrowardi et al., 2026).

Lastly, regarding technical limitations and projections for improvement, the exit survey suggestions mainly point to: the need for greater contextualization with situated examples, incorporation of visual supports and further development of content. These limitations do not call into question the relevance of the resource but rather guide its refinement.

Likewise, the limitations relate to incompatibility on mobile devices and loss of session history. In this regard, participants expressed the following opinions: "When I tried to use the resource on the phone it didn't work, but on the computer it did, so we changed our strategy and tried to always use it on the computer so that it would work". "It showed me it was loading the information, but it didn't give me anything and I had to restart, and start a new chat, without it being saved. I had to be very precise and very specific with my responses".

The proposals for improvement reveal the resource's potential for scalability, as they express: "I would like it to offer a conflict simulation mode, where the chatbot gives you conflict scenarios and certain strategies that could be carried out, always grounded in reality". Another opinion was: "It would be good to hold workshops on how to create a good prompt, I think that would be important so that students know how to use this tool well and not take advantage of it".

Consequently, limitations and conditions for scalability are identified through incompatibility on mobile devices and loss of session history. All of the above is consistent with what has been reported in studies on the implementation of educational chatbots in early stages (Zawacki-Richter et al., 2019; Okonkwo & Ade-Ibijola, 2021). However, the students' adaptive capacity and the absence of questioning about the relevance of the resource suggest a positive and resilient disposition. The students' proposals for extending the model provide concrete inputs for the scalability of the innovation and coincide with the literature on AI integration in teacher education (Darling Hammond et al., 2020).

Ultimately, the empirical findings support a multidimensional evaluation of the contributions of this conversational agent specialized in coexistence, considering its degree of influence on improving university academic practices and strengthening the key competencies in the graduate profile of future teachers.

However, the interpretation of these positive results regarding the contribution of this type of conversational chatbot to students' formative process must be carried out with caution, because the study design did not make it possible to control certain factors possibly associated with the results obtained, such as: variability in the frequency and duration of chatbot use, the characteristics of the interactions established, the prior knowledge and socioemotional competencies of the teachers in training, digital literacy and their motivation. Likewise, it is not possible to specify to what extent the results are attributable to the characteristics of the chatbot itself, to the pedagogical mediation developed, or to the interrelation between both elements.

Conclusion

The present study provides empirical evidence on the use of a conversational chatbot specialized in educational coexistence as a mediating resource in teacher education. When the resource is integrated in an intentional, guided and ethically oriented manner, it can help improve self-perceived pedagogical readiness to integrate coexistence elements, strengthen their scaffolding through conceptual and regulatory understanding of the topic, foster metacognitive processes in an emergent way and develop a critical and responsible attitude toward the use of AI in teacher education.

The study allows us to maintain that the conversational chatbot operated as a pedagogical mediation resource by fostering processes of reflection, analysis and transfer to teaching practice. In this sense, the resource used constituted a contribution to students' learning process, by mobilizing not only conceptual but also procedural knowledge. However, the positive results should be interpreted with caution, since it was not possible to control individual and contextual factors that could have influenced the findings. Likewise, it cannot be determined exactly whether the results are due to the chatbot, to the pedagogical mediation or to the interaction between both elements. Furthermore, from a methodological standpoint, as noted above, the analysis focused on frequency distributions by scale level, without applying tests of statistical significance or additional measures of central tendency and dispersion, which constitutes a limitation of the study.

For future research, it is suggested to incorporate quasi-experimental designs with a control group, include measurements of actual learning that go beyond students' self-perception, implement prior workshops on formulating and using prompts, extend the experience to multiple courses in the program and explore the potential of chatbots with case-simulation capabilities for developing relevant and meaningful competencies related to educational coexistence.

In short, to the extent that more actors become involved in the processes of implementing, reflecting on and validating the use of AI-based resources, it will be possible to strengthen knowledge generation in this area and to understand more deeply the conditions under which these resources can contribute to the comprehensive education of future teachers in the current higher education scenarios at the national and international levels.

References

• Ashfaque, M. W., Tharewal, S., Iqbal, S., & Kayte, C. N. (2020). A review on techniques, characteristics and approaches to an intelligent tutoring chatbot system. In 2020 International Conference on Smart Innovations in Design, Environment, Management, Planning and Computing (ICSIDEMPC) (pp. 258-262). IEEE. https://doi.org/10.1109/ICSIDEMPC49020.2020.9299583
• Asrowardi, I., Abd Rahman, T. K., Putra, S. D., Abd Rahman, A., Faisal, M., Alanda, A., & Anwar, N. (2026). A comparative analysis of AI maturity frameworks: Toward transformative applications in higher education. JOIV: International Journal on Informatics Visualization, 10(2), 727-738. https://doi.org/10.62527/joiv.10.2.3563
• Azevedo, R., Taub, M., & Mudrick, N. V. (2018). Understanding and reasoning about real-time cognitive, affective, and metacognitive processes to foster self-regulation with advanced learning technologies. In D. H. Schunk & J. A. Greene (Eds.), Handbook of self-regulation of learning and performance (2nd ed., pp. 254-270). Routledge. https://doi.org/10.4324/9781315697048-17
• Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), 77-101. https://doi.org/10.1191/1478088706qp063oa
• Carrillo, C., Flores, M. A., & Mena, J. (2025). Artificial intelligence in the context of teacher education: Emerging themes and critical issues. European Journal of Teacher Education, 48(5), 1143-1162. https://doi.org/10.1080/02619768.2025.2584283
• Celik, I., Dindar, M., Muukkonen, H., & Järvelä, S. (2022). The promises and challenges of artificial intelligence for teachers: A systematic review of research. TechTrends, 66(4), 616-630. https://doi.org/10.1007/s11528-022-00715-y
• Cordón García, O. (2023). Inteligencia artificial en educación superior: Oportunidades y riesgos. RiiTE Revista Interuniversitaria de Investigación en Tecnología Educativa, (15), 16-27. https://doi.org/10.6018/riite.591581
• Creswell, J. W., & Plano Clark, V. L. (2018). Designing and conducting mixed methods research (3rd ed.). SAGE.
• Darling Hammond, L., Flook, L., Cook-Harvey, C., Barron, B., & Osher, D. (2020). Implications for educational practice of the science of learning and development. Applied Developmental Science, 24(2), 97-140. https://doi.org/10.1080/10888691.2018.1537791
• Edel Navarro, R. (2025). La inteligencia artificial como catalizador en la transformación de la educación superior. Educación Superior, (40), 145-155. https://doi.org/10.56918/es.2025.i40.pp145-155
• Flavell, J. H. (1979). Metacognition and cognitive monitoring: A new area of cognitive-developmental inquiry. American Psychologist, 34(10), 906-911. https://doi.org/10.1037/0003-066X.34.10.906
• García, M., Morales González, M. J., & Gisbert Cervera, M. (2022). El desarrollo de la competencia digital docente en educación superior. Una revisión sistemática de la literatura. RiiTE Revista Interuniversitaria de Investigación en Tecnología Educativa, (13), 173-199. https://doi.org/10.6018/riite.543011
• Guba, E. G., & Lincoln, Y. S. (2000). Paradigmas en competencia en la investigación cualitativa. In C. A. Denman & J. A. Haro (Comps.), Por los rincones: Antología de métodos cualitativos en la investigación social (pp. 113-145). El Colegio de Sonora.
• Hamman-Fisher, D., & McGhie, V. (2023). Towards decoloniality of the education training and development third-year curriculum: Employing situated learning characteristics to facilitate authentic learning. Cogent Education, 10(2), Article 2237301. https://doi.org/10.1080/2331186X.2023.2237301
• Holmes, W., Bialik, M., & Fadel, C. (2019). Artificial intelligence in education: Promises and implications for teaching and learning. Center for Curriculum Redesign. https://discovery.ucl.ac.uk/id/eprint/10139722/
• Holmes, A. G. D. (2020). Constructivist learning in university undergraduate programmes. Has constructivism been fully embraced? Is there clear evidence that constructivist principles have been applied to all aspects of contemporary university undergraduate study? Shanlax International Journal of Education, 8(1), 1-12. https://doi.org/10.34293/education.v8i1.819
• Huang, J., Saleh, S., & Liu, Y. (2022). A review on artificial intelligence in education. Academic Journal of Interdisciplinary Studies, 10(3), 206-217. https://doi.org/10.36941/ajis-2021-0077
• Hwang, G.-J., & Chang, C.-Y. (2021). A review of opportunities and challenges of chatbots in education. Interactive Learning Environments, 31(7), 4099-4112. https://doi.org/10.1080/10494820.2021.1952615
• Johnson, R. B., Onwuegbuzie, A. J., & Turner, L. A. (2007). Toward a definition of mixed methods research. Journal of Mixed Methods Research, 1(2), 112-133. https://doi.org/10.1177/1558689806298224
• Kasneci, E., Seßler, K., Küchemann, S., Bannert, M., Dementieva, D., Fischer, F., Gasser, U., Groh, G., Günnemann, S., Hüllermeier, E., Krusche, S., Kutyniok, G., Michaeli, T., Nerdel, C., Pfeffer, J., Poquet, O., Sailer, M., Schmidt, A., Seidel, T., ... Kasneci, G. (2023). ChatGPT for good? On opportunities and challenges of large language models for education. Learning and Individual Differences, 103, e102274. https://doi.org/10.1016/j.lindif.2023.102274
• Kuhail, M. A., Alturki, N., Alramlawi, S., & Alhejori, K. (2023). Interacting with educational chatbots: A systematic review. Education and Information Technologies, 28(1), 973-1018. https://doi.org/10.1007/s10639-022-11177-3
• Luckin, R., Holmes, W., Griffiths, M., & Forcier, L. B. (2016). Intelligence unleashed: An argument for AI in education. Pearson. https://discovery.ucl.ac.uk/id/eprint/1475756/
• Liu, H., Fang, X., Cui, Q., & Chiu, T. K. F. (2026). A systematic mapping review on how generative artificial intelligence impacts social and emotional learning: A case of large language model chatbots. Review of Education, 14(1), e70141. https://doi.org/10.1002/rev3.70141
• Marradi, A., Archenti, N., & Piovani, J. (2014). Manual de metodología de las ciencias sociales. Siglo XXI. https://desarrollomedellin.wordpress.com/wp-content/uploads/2017/03/marradi-a-archenti-n-piovani-j-2007.pdf
• Martínez, M. (2025). Inteligencia artificial y educación. Revista Docentes 2.0, 18(1), 245-257. https://doi.org/10.37843/rted.v18i1.614
• Mederos Mollineda, K., Tabares Cruz, Y. B., & Domínguez Brito, L. D. (2026). ChatGPT como mediador cognitivo en la educación superior: Impacto en el pensamiento crítico y la metacognición del estudiante universitario. Reincisol, 5(9), Article 1388. https://doi.org/10.59282/reincisol.V5(9)1388
• Méndez Forns, I. M., & Vázquez Zubizarreta, G. (2025). La mediación pedagógica en el proceso de enseñanza aprendizaje con el modelo híbrido, en la Universidad de La Habana. Nobilis, 2(1), 33-42. http://share.google/ocDxU2H5Sqaixf5a
• Okonkwo, C. W., & Ade-Ibijola, A. (2021). Chatbots applications in education: A systematic review. Computers and Education: Artificial Intelligence, 2, Article 100033. https://doi.org/10.1016/j.caeai.2021.100033
• Sandín Esteban, M. P. (2003). Investigación cualitativa en educación: Fundamentos y tradiciones. McGraw-Hill. https://ve.scielo.org/scielo.php?script=sci_arttext&pid=S0798-97922005000300007
• Schraw, G., & Dennison, R. S. (1994). Assessing metacognitive awareness. Contemporary Educational Psychology, 19(4), 460-475. https://doi.org/10.1006/ceps.1994.1033
• Smutny, P., & Schreiberova, P. (2020). Chatbots for learning: A review of educational chatbots for Facebook Messenger. Computers & Education, 151, Article 103862. https://doi.org/10.1016/j.compedu.2020.103862
• Tashakkori, A., & Teddlie, C. (Eds.). (2010). SAGE handbook of mixed methods in social & behavioral research (2nd ed.). SAGE. https://doi.org/10.4135/9781506335193
• Taub, M., Azevedo, R., Rajendran, R., Cloude, E., Biswas, G., & Price, M. (2021). How are students' emotions related to the accuracy of cognitive and metacognitive processes during learning with an intelligent tutoring system? Learning and Instruction, 72, Article 101200. https://doi.org/10.1016/j.learninstruc.2019.04.001
• Tlili, A., Shehata, B., Adarkwah, M. A., Bozkurt, A., Hickey, D. T., Huang, R., & Agyemang, B. (2023). What if the devil is my guardian angel: ChatGPT as a case study of using chatbots in education. Smart Learning Environments, 10(1), Article 15. https://doi.org/10.1186/s40561-023-00237-x
• UNESCO. (2023). Guidance for generative AI in education and research. https://doi.org/10.54675/EWZM9535
• UNESCO. (2024). AI competency framework for teachers. https://doi.org/10.54675/ZJTE2084
• Vaca Montenegro, M. L., Campos Ortiz, J. M., Anrango Sosa, M. A., Quilumbaquin Chicaiza, M. C., Patiño Méndez, P. F., & Arias Benalcázar, D. V. (2025). Las tecnologías emergentes en los procesos de enseñanza-aprendizaje: Una revisión sistemática. Revista Latinoamericana de Calidad Educativa, 2(3), 34-41. https://doi.org/10.70625/rlce/98
• Vaillant, D., & Marcelo, C. (2015). El ABC y D de la formación docente. Narcea.
• Vargas Gómez, Á. M. (2025). La inteligencia artificial generativa como mediación didáctica crítica en la educación superior. Revista Avancemos en Educación, 1(1), 14-27. https://mirevista.avancemos.edu.co/index.php/revistaavancemoseneducacion/article/view/2
• Wollny, S., Schneider, J., Di Mitri, D., Weidlich, J., Rittberger, M., & Drachsler, H. (2021). Are we there yet? – A systematic literature review on chatbots in education. Frontiers in Artificial Intelligence, 4, Article 654924. https://doi.org/10.3389/frai.2021.654924
• Xue, S., & Crompton, H. (2024). Educational technology research during the COVID-19 pandemic. Interactive Technology and Smart Education, 21(1), 83-107. https://doi.org/10.1108/ITSE-05-2022-0067
• Zawacki Richter, O., Marín, V. I., Bond, M., & Gouverneur, F. (2019). Systematic review of research on artificial intelligence applications in higher education – where are the educators? International Journal of Educational Technology in Higher Education, 16, Article 39. https://doi.org/10.1186/s41239-019-0171-0
• Zou, W., Yang, Q., DiFranzo, D., Chen, M., Hui, W., & Bazarova, N. N. (2024). Social Media Co-pilot: Designing a chatbot with teens and educators to combat cyberbullying. International Journal of Child-Computer Interaction, 41, e100680. https://doi.org/10.1016/j.ijcci.2024.100680