Conference Agenda
Overview and details of the sessions of this conference. Please select a date or location to show only sessions at that day or location. Please select a single session for detailed view (with abstracts and downloads if available).
Please note that all times are shown in the time zone of the conference. The current conference time is: 19th Aug 2026, 21:28:00 EET
|
Daily Overview |
| Session | ||
16 SES 15 B: Teachers Navigating AI: Agency, Ethics, and Readiness
Paper Session | ||
    | ||
16. ICT in Education and Training
Paper The Interplay of Teachers' intelligent-TPACK, Agency and Fear: A Large-Scale Survey on AI in Education in Finland 1: University of Oulu, Finland; 2: University of Eastern Finland, FInland Presenting Author:The rapid integration of artificial intelligence (AI) into educational contexts has fundamentally altered teachers’ professional work (Laru et al., 2025). AI-based tools increasingly support lesson planning, assessment, feedback, and instructional decision-making, creating new opportunities for personalization and efficiency in teaching (Celik et al., 2025). At the same time, these technologies raise substantial concerns related to ethical use, trust in algorithmic recommendations, and the preservation of teachers’ professional agency. Unlike earlier educational technologies, AI systems actively generate content and recommendations, which may influence pedagogical decisions in uncertain ways. As a result, teachers are required not only to use AI tools, but also to critically evaluate AI-generated outputs and determine when and how such tools should inform instructional practice. Existing research on AI in education has often focused on technical adoption or attitudes toward AI, with limited attention to teachers’ integrated professional knowledge and decision-making processes. While the Technological Pedagogical Content Knowledge (TPACK) framework has been widely used to conceptualize teachers’ technology integration, it does not fully account for the autonomous and data-driven nature of AI systems. The Intelligent-TPACK framework addresses this gap by extending TPACK to include AI-specific knowledge dimensions and by emphasizing ethics, trust, and teacher agency as essential components of responsible AI use (Celik, 2023). However, large-scale empirical evidence examining these dimensions simultaneously remains scarce. This study responds to this need by providing a comprehensive examination of teachers’ AI use based on data from 1,505 teachers in Finland Methodology, Methods, Research Instruments or Sources Used The primary aim of this study was to investigate teachers’ use of AI through the lens of the Intelligent-TPACK framework, with a particular focus on how AI-related knowledge dimensions relate to ethical judgment, trust in AI systems, and teacher agency. Specifically, the study sought to examine (a) teachers’ levels of Intelligent-Technological Knowledge (Intelligent-TK), Intelligent-Technological Pedagogical Knowledge (Intelligent-TPK), Intelligent-Technological Content Knowledge (Intelligent-TCK), and their integrative Intelligent-TPACK; (b) the role of ethics and trust in shaping teachers’ engagement with AI; and (c) the extent to which teacher agency supports responsible and reflective AI use in pedagogical decision-making. Conclusions, Expected Outcomes or Findings The findings indicate that teachers reported moderate to high levels of AI-related knowledge across the Intelligent-TPACK dimensions. Intelligent-TK emerged as a foundational component, supporting more advanced pedagogical and content-oriented uses of AI. Both Intelligent-TPK and Intelligent-TCK were strong predictors of Intelligent-TPACK, underscoring the importance of aligning AI tools with pedagogical strategies and subject-specific practices rather than using AI in a purely technical manner. Ethical judgment was strongly associated with Intelligent-TPACK, suggesting that teachers with higher integrative AI knowledge were more likely to consider issues such as privacy, fairness, and inclusivity when using AI in teaching. Trust in AI systems was positively related to AI-related knowledge; however, the results indicate that trust alone is insufficient for responsible AI use. Teachers who reported high trust without corresponding agency were more vulnerable to uncritical acceptance of AI recommendations. Teacher agency played a crucial moderating role. Teachers with stronger agency were better able to critically evaluate AI-generated outputs, adapt recommendations to their instructional context, and retain control over pedagogical decisions. This finding highlight agency as a key protective factor against AI overreliance. This study demonstrates that effective and responsible AI integration in education depends on more than technical competence. A balanced combination of Intelligent-TPACK, ethical judgment, calibrated trust, and strong teacher agency is essential for human-centred AI use in teaching. The findings emphasize the need for teacher education and professional development programs that explicitly address ethical reasoning, trust calibration, and agency alongside AI-related knowledge. By doing so, such programs can support teachers in leveraging the benefits of AI while maintaining professional autonomy and pedagogical responsibility. References Celik, I. (2023). Towards Intelligent-TPACK: An empirical study on teachers’ professional knowledge to ethically integrate artificial intelligence (AI)-based tools into education. Computers in human behavior, 138, 107468. Celik, I., Kontkanen, S., Laru, J., & Dalyanci, A. A. (2025). Co-constructing adaptive lesson plans with GenAI: Pre-service teachers' Intelligent-TPACK and prompt engineering strategies. Computers & Education, 105485. Laru, J., Celik, I., Jokela, I., & Mäkitalo, K. (2025). The antecedents of pre-service teachers’ AI literacy: perceptions about own AI driven applications, attitude towards AI and knowledge in machine learning. European Journal of Teacher Education, 48(5), 964-986. 16. ICT in Education and Training
Paper Faculty as Street-Level Bureaucrats: A Multiple Case Study on Ethical Generative Artificial Intelligence Integration in a Leading International Turkish Research University Middle East Technical University, Turkey (Türkiye) Presenting Author:Generative artificial intelligence (GenAI) has entered the global education landscape with the expectation that it will enhance the performance of students, teaching staff, and institutions, as noted in the OECD's latest Digital Education Outlook report (2026). One of the key findings in this report is that meeting these expectations requires strengthening student agency in the use of GenAI, ensuring educators engage collaboratively, and recommending strong pedagogy and a human-centered approach for system and institutional management. In higher education, as discussed by Bozkurt and colleagues (2021), faculty members play a critical role in this process because they are directly responsible for deciding how AI tools will be used in classroom applications. Therefore, understanding faculty perceptions and ethical decision-making processes regarding GenAI integration is considered a fundamental component of successful pedagogical transformation. This understanding is crucial for establishing how GenAI can be systemically integrated and ethically practiced within higher education. In this context, as highlighted by the European Commission (2025), this research aims to concretize the goal of ensuring that the classroom strategies employed by teaching staff are fully compatible with democratic values and legal frameworks in the field of education in light of recent technological developments. The purpose of proposed research is to document the classroom strategies and ethical solutions generated by faculty members regarding students' use of GenAI, specifically within Türkiye's leading state research university where institutional ethical guidelines related to the use of GenAI are currently absent. By building on the individual practices observed in this premier public institution, the study lays the groundwork for developing an inclusive institutional policy for ethical AI use. In this regard, the study will reveal bottom-up strategies that fill institutional gaps in a poly-crisis environment, offering a local and practical response to ensure trust in academic knowledge. This research analyzes the gap between institutional policy rhetoric and the operational reality of GenAI integration through the lens of Lipsky’s (1980) Street-Level Bureaucrats Theory. By highlighting the tensions between the emerging global policy agenda on GenAI and faculty members' daily classroom practices, it will document how institutional gaps are filled with individual strategies. Research on GenAI use in Turkish universities is still limited, but existing studies provide important insights into GenAI use in higher education in Türkiye. The literature shows that GenAI is mostly used at an individual level by academics rather than through institutional strategies. Studies highlight that faculty members approach GenAI cautiously and focus on its effects on teaching, assessment, and academic values (Bozkurt et al., 2021). Faculty acceptance of GenAI is strongly influenced by ethical concerns and professional responsibility, which is especially relevant in the Turkish higher education context where instructors have high autonomy (Bilgin & Güngören, 2025) Ethical use of AI is a central theme in the literature. The main concerns include academic integrity, data privacy, and professional responsibility. Faculty members worry that GenAI may increase plagiarism and weaken assessment validity (Marín et al., 2025), while students are often unsure about ethical boundaries (Ahmed, 2024). Shared documents and frameworks emphasize transparency, risk-based regulation, and institutional guidance to support responsible GenAI use in higher education (Temper et al., 2025). Methodology, Methods, Research Instruments or Sources Used This qualitative research adopts an interpretivist paradigm to understand faculty members' ethical strategies regarding GenAI integration within their contextual realities. A multiple case study design was selected because it is the most suitable approach when the boundaries between the phenomenon and its real-life context are ambiguous (Yin, 2018). Guided by the research question, the study utilizes replication logic for case selection. Since strategies were anticipated to vary across disciplines, the study follows a theoretical replication structure. Participants comprise senior academic staff engaged in both teaching and research across five distinct faculties and the Department of Modern Languages at a leading Turkish university. Crucially, each participant is treated as a separate case within this design. Participants were selected through criterion-based purposive sampling by identifying faculty knowledgeable about GenAI ethics (Patton, 2014). Interview protocol was developed through a literature review and finalized with expert validation. To reduce recall bias, participants receive questions in advance. Semi-structured interviews last 30 to 40 minutes and take place in person or online depending on preference. Informed consent is obtained for voluntary participation and audio recording. All recordings are transcribed verbatim and shared with participants for member checking to ensure accuracy. To capture more depth, two researchers attend each session where one acts as a moderator and the other records field notes. The data collection process begins with a target of 2-3 participants per faculty and is completed upon reaching data saturation. Methodological triangulation is ensured through the analysis of course syllabi and policy documents. National texts published in 2024 identified for analysis include the Council of Higher Education’s "Ethical Guide on Generative AI Usage," TÜBİTAK’s "Guide on Responsible and Reliable Use of Generative AI. Internationally, the "EU AI Act" in the same year serves as the key regulatory benchmark. Analyzing these documents reveals macro expectations while interviews uncover micro-level strategies. Data analysis is conducted using the reflexive thematic analysis approach developed by Braun and Clarke (2021). This method emphasizes the researcher's active role in constructing themes through a six-stage iterative process. Preliminary document review has informed the interview guide. Since document analysis offers a top-down perspective, interviews are essential to examine bottom-up practices and assess the extent to which classroom realities deviate from official policies. Conclusions, Expected Outcomes or Findings The preliminary findings indicate that the absence of a clear and comprehensive GenAI policy within the scope of higher education at the institutional level has led faculty members to come up with individual strategies to address the issue, which has resulted in an obvious discrepancy between the macro level policy expectations and micro level classroom practices. Document analysis shows that the policy texts ascribe strategic roles to faculty members, harbor specific expectations regarding ethical oversight, and highlight the necessity of acquiring critical AI literacy, transcending mere instrumental competence, for effective decision-making processes. The analysed policy documents emphasize ethical oversight, academic integrity, and critical artificial intelligence literacy as faculty member responsibilities, which does not seem feasible in real-life applications. Therefore, the street-level bureaucrats approach offers a strong analytical frame to grasp how faculty members fill the institutional gaps. Document and literature reviews have highlighted that academic integrity stands out as the main ethical issue regarding GenAI. Faculty members have developed a range of strategies, including redesigning assignments and exams, setting clear guidelines regarding the use of generative AI, establishing more visible communication about the ethical issues, and restructuring assessment processes. It is anticipated that the data to be gathered across six disciplines will reveal some differences specific to the disciplines and some common points in the approaches of the faculty members towards the ethical use of GenAI. While the current study is limited to views and practices of faculty members in one university only, it does pinpoint the determining role of faculty members in ethical GenAI integration and has the potential to demonstrate that individual practices may form a strong basis to guide policy. References Ahmed, R. (2024). Exploring ChatGPT usage in higher education: Patterns, perceptions, and ethical implications among university students. Journal of Digital Learning and Distance Education, 3(6), 1122–1131. https://doi.org/10.56778/jdlde.v3i6.363 Bilgin, S., & Güngören, Ö. C. (2025). A detailed examination of faculty acceptance of artificial intelligence: Insights from key variables. Journal of Pedagogical Sociology and Psychology, 7(3), 64–77. https://doi.org/10.33902/jpsp.202534245 Bozkurt, A., Karadeniz, A., Baneres, D., Guerrero-Roldán, A. E., & Rodríguez, M. E. (2021). Artificial intelligence and reflections from educational landscape: A review of AI studies in half a century. Sustainability, 13, 800. https://doi.org/10.3390/su13020800 Braun, V., & Clarke, V. (2021). One size fits all? What counts as quality practice in (reflexive) thematic analysis? Qualitative Research in Psychology, 18(3), 328–352. https://doi.org/10.1080/14780887.2020.1769238 Council of Higher Education. (2024). Yükseköğretim kurumları bilimsel araştırma ve yayın faaliyetlerinde üretken yapay zekâ kullanımına dair etik rehber [Ethical guide on generative AI usage in scientific research and publication activities]. https://proje.yok.gov.tr/documentFiles/17539645334.Y%C3%BCksek%C3%B6%C4%9Fretimde%20%C3%BCretken%20yapay%20zeka%20kullan%C4%B1m%C4%B1-tr.pdf European Parliament and Council of the European Union. (2024). Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonised rules on artificial intelligence (Artificial Intelligence Act). Official Journal of the European Union, L, 2024/1689. http://data.europa.eu/eli/reg/2024/1689/oj Guest, G., Bunce, A., & Johnson, L. (2006). How many interviews are enough? An experiment with data saturation and variability. Lipsky, M. (1980). Street Level Bureaucracy: Dilemmas of the Individual in Public Services. Russell Sage Foundation. http://www.jstor.org/stable/10.7758/9781610447713 Marín, Y. R., Caro, O. C., Rituay, A. M., Llanos, K. A., Perez, D. T., Bardales, E. S., Tuesta, J. N., & Santos, R. C. (2025). Ethical challenges associated with the use of Artificial Intelligence in university education. Journal of Academic Ethics, 23(4), 2443–2467. https://doi.org/10.1007/s10805-025-09660-w OECD (2026), OECD Digital Education Outlook 2026: Exploring Effective Uses of Generative AI in Education, OECD Publishing, Paris, https://doi.org/10.1787/062a7394-en. Patton, M. (2014) Qualitative Research and Evaluation Methods. 4th Edition, Sage, Thousand Oaks. Scientific and Technological Research Council of Türkiye. (2024). Destek süreçlerinde üretken yapay zekânın (ÜYZ) sorumlu ve güvenilir kullanımı rehberi [Guide on responsible and reliable use of generative AI in support processes].https://tubitak.gov.tr/sites/default/files/2025-10/UYZ_Rehberi_v03_TR.pdf Temper, M., Tjoa, S., & David, L. (2025). Higher education act for AI (HEAT-AI): A framework to regulate the usage of AI in Higher Education Institutions. Frontiers in Education, 10, 1505370. https://doi.org/10.3389/feduc.2025.1505370 Yin, R. K. (2018). Case Study Research and Applications: Design and Methods (6th ed.). Thousand Oaks, CA: Sage. 16. ICT in Education and Training
Paper Teachers' Use of Digital Resources in Education: A Step Towards Agency, Inclusive Education and Transformation? 1: Department of Education, University of Oslo, Norway; 2: Unit for Digitalization and Educational Quality, University of South-Eastern Norway, Norway Presenting Author:The integration and use of technology in education have been at the forefront and highly discussed for more than two decades. Oftentimes, these discussions are led by several stakeholders (e.g., policymakers, researchers, teachers, students, parents) holding strong opinions – yet on a diverse array. Moreover, their diverse experiences, attitudes and knowledge have led to a public debate echoing two main sides – those supporting increased access and use of technology in education and those against it (Siddiq & Scherer, 2025). However, most actors working in and with schools agree that a more balanced approach is needed when considering what role technology should have in education – emphasizing teachers as critical in this endeavor (Siddiq & Scherer, 2025). Further, across time several aspects have changed and keep changing, such as the technology itself and the access to it (Erstad & Siddiq, 2023), educational policy, curricula, and teachers’ attitudes, knowledge and competence to use technology for teaching and learning (Howard et al., 2020). Also, the focus on teachers’ digital literacy and professional development programs to support and develop such competences has received widespread attention. In Norway and other northern countries this is evident through for example the massive governmental programs for teacher education. Along these lines, access to open educational resources (OER) has increased, and been encouraged as a useful tool in providing inclusivity in education. OER are defined as “learning, teaching and research materials in any format and medium that reside in the public domain or are under copyright that have been released under an open license, that permit no-cost access, re-use, re-purpose, adaptation and redistribution by others” (UNESCO, 2019). Thus, OER come with an open license and grants others the rights to access, re-use, adapt, modify, repurpose, and redistribute. Research has shown that OER can support the diverse needs of individual learners, promote gender equality and incentivize innovative pedagogical, didactical and methodological approaches (Bossu et al., 2012; Farrow et al., 2024). However, previous studies show that teachers lack knowledge of OER and simply confuse OER with digital resources in general (Baas et al., 2019). The EQui-T (European quality development system for inclusive education and teacher training) project, supported under the Erasmus+ teacher academies program aims at enhancing high quality teaching in an inclusive European context. This research employ data from EQuiT-project and investigate teachers’ understanding of OER, and the reasoning behind why they use OER and/or digital learning resources, addressing the research questions: RQ.1 How do primary and secondary school teachers understand the concept of OER? RQ.2 What reasons do primary and secondary school teachers provide for using digital resources for teaching and learning? We use the OER definition as point of departure along with the SAMR-framework (Romrell et al., 2014). The SAMR model presents strategies for classroom technology implementation at four stages including: substitution, augmentation, modification, and redefinition. While the first two levels, substitution and augmentation (technology acts as a direct substitute, with functional improvement) present a simple use of technology in pedagogical practice, e.g., simply replacing traditional materials with digital ones, the next two stages add an advancement in terms of technology integration. These are modification (technology allows for significant task redesign) and redefinition (technology allows for the creation of new tasks, previously inconceivable). Classrooms in which this kind of mastery is embedded are more novel and show immersive uses for technology, e.g., they create and publish own work across multiple forms of media and technology. We use the OER definition and the SAMR model as our underlying theoretical framework to investigate to which extent the teachers’ explanations and reflections in the interviews align with these respectively. Methodology, Methods, Research Instruments or Sources Used An interview guide was developed, reviewed and refined by the EQui-T project team. Semi-structured interviews were used to give place for new themes to arise or further development on the selected topics (Brinkmann & Kvale, 2015 p.156). Interview data from five European countries (Austria, Estonia, Italia, Norway and Spain) were collected, comprising a total of 38 interviews – counting 6-10 interviews in each country. The interviews were conducted in the native language of each country, transcribed and translated to English by the local project teams. The authors of this paper coded and analysed all the data to reduce bias. We employed thematic analysis to identify different threads of meaning in the data (Braun & Clarke, 2006; 2019). This method was specifically selected to ensure the researchers’ interpretations were at the centre of the analytical process (Braune & Clarke, 2019). Indeed, a dialogic approach to data analysis was used, where the two main researchers analysed and generated patterns of meaning in the data together, through discussion. Moreover, the other project partner countries reviewed the results and engaged in critical and reflective discussions to ensure understanding, translation-related issues, and could explain local context if confusing or confounding issues arised. The analytical steps were as following. In a first step of familiarization, the authors read through all the transcripts to get acquainted with the data. In a second step, the authors started coding part of the data inductively and individually. These initial codes were collected and discussed in a first analysis workshop. The rest of the data was then coded following the codes developed during the first workshop. During a second analysis workshop, the authors reviewed the codes used to analyse the whole dataset and generated themes from these codes. These themes were reviewed during the last analysis step of writing the article and discussed in line with the different categories of the SAMR framework (Rohrell et la., 2014) and the OER definition (UNESCO, 2019). Conclusions, Expected Outcomes or Findings Our preliminary results show that overall the teachers’ lack understanding of the concept of Open Educational Resources (OER). Even though teachers declared understanding the definition of OER (UNESCO, 2019), the analysis of the interview data showed that they could not connect this definition to concrete examples of OER, or only focused on some specific aspects of OER (e.g., free to use). Also, somewhat confusion around who authors and shares open resources was displayed. Moreover, they seldomly connected the concept to licences and privacy rights. These findings are aligned with previous research (Baas et al., 2019). However, they are surprising as the teachers report a widespread use of digital resources in their pedagogical practice. This study also shows that teachers, when explaining the reasons for using digital learning resources, are placing their students’ needs at the forefront (e.g., academic level, interests, physical barriers). Teachers are selecting, adapting and using digital resources to make their lessons more inclusive and adapted to all their students. A second main reason for using digital resources, presented by the teachers in our data, is student engagement. Teachers use digital resources to motivate and engage their students in the subject. Variation and flexibility are two main factors emphasised in the interviews. These two findings are important as they show teachers’ focus on students in how they plan their teaching. Both of these findings can also be linked to recommendations from the Universal Design Guidelines (Meyer & Rose, 2024) to increase inclusivity in the classroom. Our results are further discussed grounded in the underlying framework, and implications for research, policy and practice will be highglighted. References Baas, M., Admiraal, W. and van den Berg, E. (2019) ‘Teachers’ Adoption of Open Educational Resources in Higher Education’, Journal of Interactive Media in Education, 2019(1), p. 9. Bossu, C., Bull, D., & Brown, M. (2012). Opening up down under: the role of open educational resources in promoting social inclusion in Australia. Distance Education, 33(2), 151–164. https://doi.org/10.1080/01587919.2012.692050 Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), 77–101. Braun, V., & Clarke, V. (2019). Reflecting on reflexive thematic analysis. Qualitative Research in Sport, Exercise and Health, 11(4), 589–597. https://doi.org/10.1080/2159676X.2019.1628806 Brinkmann, S., & Kvale, S. (2015). InterViews: Learning the craft of qualitative research interviewing (3rd ed.). Sage. Farrow, R., Iniesto, F., Pitt, R., Weller, M., & Bossu, C. (2024). Innovation with Open Educational Resources: An integrative review of drivers, barriers and enablers. Journal of Open, Distance, and Digital Education, 1(2) pp. 1–41. https://doi.org/10.25619/y3yte265. Erstad, O., & Siddiq, F. (2023). Educational assessment of 21st century skills—novel initiatives, yet a lack of systemic transformation, Editor(s): Robert J Tierney, Fazal Rizvi, Kadriye Erkican, International Encyclopedia of Education (Fourth Edition), Elsevier, 2023, Pages 245-255, ISBN 9780128186299, https://doi.org/10.1016/B978-0-12-818630-5.09038-2. Howard, K. S., Tondeur, J., Siddiq, F., & Scherer, R. (2020). Ready, set, go! Profiling teachers’ readiness for online teaching in secondary education, Technology, Pedagogy and Education. Doi:10.1080/1475939X.2020.1839543 Meyer, A., & Rose, D. H. (2024). Universal Design for Learning: Principles, Framework, and Practice (D. Gordon, Ed.; 2nd ed.). Cast Professional Publishing. Romrell, D., Kidder, L. C. & Wood, E. (2014). The SAMR Model as a Framework for Evaluating mLearning. Journal of Asynchronous Learning Networks, 18 (2). https://eric.ed.gov/?id=eJ1036281 Siddiq, F. & Scherer, R. (2025). Learner’s Screen Use. Time to Reload the Debate. Nordic Journal of Digital Literacy, 20 (1), p. 35–39. DOI: https://doi.org/10.18261/njdl.20.1.6 UNESCO (2019). Recommendation on Open Educational Resources (OER). Retrieved from the United Nations Educational, Scientific, and Cultural Organization website: https://unesdoc.unesco.org/ark:/48223/pf0000373755/PDF/373755eng.pdf.multi.page=3. 16. ICT in Education and Training
Paper When Perceived Readiness is Not Enough: Interactional Strategies in Student Teachers’ AI Use University of Eastern Finland, Finland Presenting Author:Theoretical framework The rapid rise of generative Artificial Intelligence (GenAI) is reshaping (teacher) education and raising urgent questions about student teacher readiness for it and teacher training programs that can support it. Studies such as Bui et al. (2025) show that student teachers often display low adoption and sporadic use of GenAI, highlighting both a foundational experience gap and a skill gap in areas like prompt engineering, evaluative judgment, and ethical considerations. While AI is increasingly seen as a collaborative thinking partner rather than a source of perfect answers (Bui et al., 2025), some student teachers still lack AI literacy skills, such as technical understanding, critical evaluation, and practical application (Laupichler et al., 2023), to critically assess or strategically integrate its outputs (Laru et al., 2025). Teacher education programs in international contexts face additional challenges, since it adds layers of cultural and institutional complexity and digital gap issues, requiring careful design to ensure knowledge transfer across contexts (Assefa et al., 2025). Teacher education in Finland, including its international master’s degree programs in education, is known for its research orientation, where student teachers learn how to implement research skills into their everyday teaching (Krokfors, 2011). Research-based teaching is characterized by having research as a basis for all teaching, the activities are based in a way that students adopt a research attitude when inquiring about and solving pedagogical problems, and students formally learn research skills (Krokfors, 2011). However, the research skills of international students coming to Finland are very diverse and depending on their respective teacher education research-related culture. While empirical evidence on GenAI’s direct impact on research strategy development is limited, studies suggest that GenAI-assisted teaching, especially when combined with teacher supervision, can enhance learning satisfaction and engagement (Tang et al., 2025), which may contribute to more effective research-based teaching practices by student teachers. Therefore, GenAI can function as a research assistant for student teachers who are learning about education research skills by offering interactive, personalized, and reflective tools that support their research practice about education research. However, the effective use of GenAI in this role depends not only on access to such tools but also on student teachers’ perceptions of AI as a research assistant and on their interactional competencies, including how they prompt and respond to the system. Accordingly, examining how student teachers perceive AI’s role as a research assistant—through the lenses of pedagogical and linguistic interactional analyses (Bui et al., 2025; Eguchi et al., 2025)—becomes essential. Such analysis can illuminate how students interact with and benefit from individualized AI platforms, and whether these tools meaningfully support their reflections on using AI in research contexts. Recent reviews emphasize that AI literacy involves not only technical understanding but also competencies in managing, questioning, and repairing interactions with AI systems (Biagini, 2025; Yang et al., 2025). Controlling, skipping, or resisting the bot demonstrates students’ ability to assert agency, evaluate system behavior, and regulate conversational flow—skills increasingly recognized as central to AI literacy in education.
Research Aim and Questions RQ1: What are student teachers’ perceptions of readiness in using GenAI tools as research assistants? RQ2: What kind of interaction patterns occur during an AI-driven conversation session about using GenAI tools as research assistants? Methodology, Methods, Research Instruments or Sources Used This study employed a mixed-methods exploratory research design to investigate student teachers’ readiness to use GenAI as a research assistant and their interactional practices when engaging with an AI‑supported research system. In international master’s degree programs in Finland, research-methods training is compulsory (Filippou et al., 2017). Quantitative research courses typically emphasize basic and intermediate statistical analyses using tools like SPSS and jamovi, whereas qualitative courses focus on philosophical foundations, research paradigms, extensive literature engagement, and hands-on work with data collection and analysis. Because qualitative inquiry requires iterative reading and writing work across all stages of the research process, the qualitative course provided an ideal setting to examine how international students adopt GenAI as a research assistant, from literature review to data collection and analysis, until reporting. Participants were 21 student teachers enrolled in the Qualitative Research Methods course in 2025 within two international teacher education master’s programs at a Finnish university. This cohort was selected to capture the wide cultural and educational diversity characteristic of the international master’s student population in education at the institution. Half of the students had no previous teaching experience, and the other half had 2-5 years of experience. Completion of the task was obligatory as part of the course activities, but participation in the research was voluntary, with data protection safeguards managed by the platform. Data were collected using an adapted Generative AI Readiness Scale (Bui et al., 2025) in 6-point Likert scale and with additional open-ended questions. The data collection was administered through the SchoolAI platform (https://openai.com/index/schoolai/). It is powered by GPT‑4.1 and supported by image generation and text‑to‑speech capabilities. It provides a controlled and observable AI infrastructure for educational research. The system prompt with 1000 words, presented students with (a) an overview of the task, (b) a consent section, (c) the scale items with optional follow‑up questions, (d) an open‑ended reflection questions, and (e) instructions specifying the expected tone and quality of AI feedback. Quantitative data were analyzed through descriptive statistics to examine students’ readiness and perceptions of using AI as a research assistant. Qualitative analysis followed the general guidelines of conversational analysis (CA) (Sidnell & Stivers, 2012). The chat transcripts were analyzed through core CA lenses to observe how the interactions between the participants and the bot unfolded, revealing conversational aspects such as turn-taking and design, sequences, repair, resisting and redirecting. By comparing the cases, we identified recurring interaction practices. Conclusions, Expected Outcomes or Findings Students showed high perceptions of AI’s usefulness (M = 4.81, SD = 0.60), strong Behavioural intention to use it for research (M = 4.83, SD = 0.60), realistic understanding of GenAI accuracy (M = 4.79, SD = 0.41), and high GenAI readiness (M = 4.62, SD = 0.75); indicating that participants felt well prepared to use AI tools in research tasks. Across all AI led inquiries, four interaction patterns shaped how participants navigated the dialogues. In Elaborating, students followed the bot’s sequence, accepted follow ups, tolerated minor errors, and provided extended, reflective turns. In Skipping, users minimized interaction, producing short answers and using pace control moves such as “skip” or “next,” adapting the bot to suppress follow ups and minimize its responses. Controlling behavior involved restarting, reordering, or correcting the bot; often triggered by sequencing confusion or perceived repetition. Users occasionally asked to bypass the sequential protocol (“show all items”), requiring the bot to re ground the interaction to align with the user’s preferences. In Resisting, users expressed frustration, rejected the conversational format, or set boundaries (“just finish”), prompting the bot to reduce elaboration toward task completion. Our results demonstrate students’ emerging AI readiness and their ability to regulate, repair, and strategically manage real‑time dialogue with an AI bot. However, problematic interaction modes of the bot, such as unwanted repetition and unnecessary follow ups challenged students’ conversational strategies and exposed their difficulties in steering the system toward the next task efficiently. While effective repairs consisted of quick state acknowledgment (“already recorded”) or suppressing follow ups after repeated refusals; some participants responded to these disruptions by repeating their previous inputs and, eventually, abandoning the interaction altogether. These patterns illustrate that even high self-perceived AI readiness does not necessarily translate into smooth or effective interaction with AI tools when they present unexpected outputs. References Assefa, Y., Melaku, M. G., Bekalu, T. M., Shouket, A. T., & Yibeltal, A. A. (2025). Rethinking the digital divide and associated educational in(equity) in higher education in the context of developing countries: the social justice perspective. The International Journal of Information and Learning Technology, 42(1), 15-32. https://doi.org/10.1108/IJILT-03-2024-0058 Biagini, G. (2025). Towards an AI literate future: A systematic literature review exploring education, ethics, and applications. International Journal of Artificial Intelligence in Education, 35, 2616–2666. https://doi.org/10.1007/s40593-025-00466-w Bui, P., Korhonen, T., Kontkanen, S., Karme, S., Piispa-Hakala, S., & Veermans, M. (2025). Exploring pre-service teachers’ generative AI readiness and behavioral intentions: A pilot study . LUMAT: International Journal on Math, Science and Technology Education, 13(1), 8. https://doi.org/10.31129/LUMAT.13.1.2755 Eguchi, M., Takizawa, K., Saeki, M., Kurata, F., Suzuki, S., Matsuyama, Y., & Sawaki, Y. (2025). Human‐versus artificial intelligence‐delivered roleplay tasks for assessing interactional competence: An applied conversation analytic study. Tesol Quarterly, 59, S183-S219. Filippou, K., Kallo, J., & Mikkilä-Erdmann, M. (2017). Students’ views on thesis supervision in international master’s degree programmes in Finnish universities. Intercultural Education, 28(3), 334–352. https://doi-org.ezproxy.uef.fi:2443/10.1080/14675986.2017.1333713 Krokfors, L., Kynäslahti, H., Stenberg, K., Toom, A., Maaranen, K., Jyrhämä, R., … Kansanen, P. (2011). Investigating Finnish teacher educators’ views on research‐based teacher education. Teaching Education, 22(1), 1–13. https://doi-org.ezproxy.uef.fi:2443/10.1080/10476210.2010.542559 Laupichler, M. C., A. Aster, N. Haverkamp, and T. Raupach. (2023). Development of the “Scale for the Assessment of Non-experts’ AI literacy”–An Exploratory Factor Analysis. Computers in Human Behavior Reports 12:100338. https://doi.org/10.1016/j.chbr.2023.100338. Laru, J., Celik, I., Jokela, I. and Mäkitalo, K. (2025): The antecedents of pre-service teachers’ AI literacy: perceptions about own AI driven applications, attitude towards AI and knowledge in machine learning, European Journal of Teacher Education. 10.1080/02619768.2025.2535623. Sidnell, J., & Stivers, T. (Eds.). (2012). The handbook of conversation analysis. John Wiley & Sons. Tang, Q., Deng, W., Huang, Y., Wang, S. and Zhang, H. (2025), Can Generative Artificial Intelligence be a Good Teaching Assistant?—An Empirical Analysis Based on Generative AI-Assisted Teaching. J Comput Assist Learn, 41: e70027. https://doi-org.ezproxy.uef.fi:2443/10.1111/jcal.70027 Yang, Y., Zhang, Y., Sun, D., He, W., & Wei, Y. (2025). Navigating the landscape of AI literacy education: Insights from a decade of research (2014–2024). Humanities and Social Sciences Communications, 12, Article 45. https://doi.org/10.1057/s41599-025-04583-8 | ||