Methodological association of foreign language teachers as a platform for collective mastering of neural network technologies: results of a longitudinal study

Автор: Волкова Юлия Владимировна

Организация: БелГУ, МОУ Стрелецкая СОШ

Населенный пункт: Белгородская область, г .Белгород

Abstract. The article presents the results of a longitudinal study devoted to the potential of the methodological association (MA) of foreign language teachers as a form of collective mastering of neural network technologies. The theoretical foundations of MA activity in the context of digital transformation are revealed, the necessity of systemic methodological support for teachers in the field of artificial intelligence is substantiated. The author's model of MA work is presented, including master classes, workshops, project seminars, prompt battles, summer school of digital literacy and a support stage. The results of the ascertaining (September 2026) and control (October 2026) stages, as well as data from a delayed observation a year after the start of the experiment, are presented. The dynamics of teachers' readiness to use neural networks is shown: the share of regular AI users increased from 25% to 75%, those owning prompt engineering – from 37.5% to 87.5%. Four groups of teachers were identified based on the results of the year: enthusiasts, pragmatists, skeptics and refuseniks. Conclusions are formulated about the need for differentiated support of teachers and continuation of MA work in the context of digital transformation of education.

Keywords: methodological association, neural networks, artificial intelligence, prompt engineering, intercultural communication, pedagogical experiment, digital transformation, professional development of teachers.

The current stage of development of the education system is characterized by active digital transformation covering all levels and components of the educational process. One of the most significant phenomena of recent years has been the rapid spread of neural network technologies and large language models. Artificial intelligence tools — ChatGPT, YandexGPT, Claude, DeepSeek, Twee, Grammarly, DALL-E, Synthesia, Magic School, and others — are increasingly penetrating pedagogical practice, offering teachers fundamentally new opportunities for preparing and conducting classes, developing educational materials, checking written work, and modeling communication situations.

The use of artificial intelligence is particularly relevant in teaching foreign languages, since English acts as the main means of intercultural communication in the global world. Neural networks open new horizons for the development of students’ intercultural competence: they make it possible to generate authentic dialogues and communication situations, synthesize speech close to that of native speakers, create visual supports and infographics, and organize language practice with chatbots in a psychologically safe environment.

However, there is a significant gap between the technological potential of neural networks and real pedagogical practice. Studies conducted in 2025 (Volkova Yu.V., 2025) showed that, with a high level of positive attitude toward AI (78%), only 23% of teachers regularly use neural networks to check work, and 65% rate their prompt-engineering skills as average or low. An objective contradiction arises between the rapid development of neural network technologies, their high didactic potential, and teachers’ insufficient methodological preparedness for their effective use.

Overcoming this gap is possible within the methodological association of foreign language teachers — a legally established form of organizing the pedagogical activity of a team that has an innovative function and can become a platform for the systematic mastery of neural network tools. However, the potential of the methodological association (MA) in this area remains insufficiently studied and implemented in practice, which determines the problem of this study.

The purpose of the study is to theoretically substantiate, develop, and experimentally test a model of activity for a methodological association of foreign language teachers aimed at mastering neural network technologies and using them for the development of intercultural communication.

The research hypothesis is that the activity of a methodological association of foreign language teachers will be effective in terms of mastering neural network technologies and integrating them into the educational process if the annual work plan of the MA includes a separate section, “Mastering and Implementing Neural Network Technologies”; if interactive forms of collective work are used (master classes, practicums, workshops, project seminars, prompt battles); if peer lesson visits are organized with a focus on the use of AI-generated materials and subsequent methodological analysis; if a bank of prompts and educational materials is created and regularly updated; and if teachers’ readiness is monitored and the program is adjusted on the basis of feedback data.

The research objectives are: to determine modern requirements for the competence of a foreign language teacher in the use of neural network technologies; to identify the functions and forms of MA activity that contribute to the effective mastery of digital technologies; to diagnose the initial level of teachers’ use of neural network tools; to develop and test a MA work program for introducing neural networks; to evaluate the effectiveness of the program and formulate practical recommendations.

The problems of methodological work at school are presented in the fundamental works of Yu.K. Babansky, M.M. Potashnik, T.I. Shamova, G.M. Kodzhaspirova, and P.I. Pidkasisty. These works reveal the functions of the methodological association, its forms of work, criteria of effectiveness, and its role in improving teachers’ qualifications. According to the definition given by T.I. Shamova and G.M. Kodzhaspirova, a methodological association is a structural subunit within the school system of advanced training that unites teachers of one or related disciplines for the collective solution of educational and methodological tasks, exchange of experience, and mutual professional development. Traditionally, the educational-methodological, diagnostic-analytical, informational, consultative, innovative, and educational functions of the MA are distinguished. In the context of digital transformation, the innovative function associated with the collective mastery of new technologies acquires particular importance.

The specific features of foreign language teachers’ activity have been studied by E.I. Passov, N.D. Galskova, and I.L. Bim. Scholars emphasize that a foreign language acts not only as an academic discipline but also as a means of intercultural communication, which requires the teacher to constantly update authentic materials, master current vocabulary and realities, and maintain his or her own foreign-language level. It is precisely these tasks that can be effectively addressed within the methodological association through the collective selection of materials, discussion of methods of working with them, and organization of language clubs and speaking practicums.

The application of artificial intelligence in education is the subject of works by V.B. Alferyeva-Termsikos, Ya. Goodfellow, I. Bengio, and A. Courville. Researchers consider prompt engineering as a strategy for developing students’ information culture, analyze the didactic potential of neural networks, and discuss the ethical aspects of their use. However, these works do not consider the methodological association as a form of collective mastery of AI tools by teachers.

Aspects of intercultural communication and existential self-determination of the individual in the digital era are revealed in the works of N.M. Lebedeva, T.G. Stefanenko, G.U. Soldatova, Yu.P. Zinchenko, L.A. Shaigerova, O.V. Khukhlaeva, E.R. Khakimov, and A.V. Reprintsev. Scholars emphasize that in a transitive society, the ability for intercultural dialogue and mastery of the global language of communication become important resources for personal development and adaptation.

Thus, theoretical analysis shows that, on the one hand, significant experience has been accumulated in the organization of methodological work and the application of AI in education; on the other hand, there is no comprehensive study examining the methodological association of foreign language teachers as a form of collective mastery of neural network technologies. This work fills this gap.

To address the research problem, a comparative analysis of existing neural network tools applicable to the work of a foreign language teacher was conducted. On the basis of the author’s classification (Volkova Yu.V., 2025), the following functional categories were identified: text generation (ChatGPT-4, Claude, Twee), image generation (DALL-E 3, Lexica.art, GigaChat), video generation (Synthesia, HeyGen, Pictory), work checking (Grammarly, LanguageTool, Writefull), presentation creation (Canva + Magic Studio, Gamma, Tome), lesson planning (Magic School, Eduaide.Ai, Twee), and interactive tasks (Genially, Twee, QuestionWell).

The analysis showed that for a foreign language teacher the most valuable tools are those capable of generating educational texts and dialogues (ChatGPT, Twee), creating exercises and worksheets (Twee, Magic School), checking written work (Grammarly, LanguageTool), and generating visual supports and videos (DALL-E 3, Synthesia). It was precisely these tools that were included in the program of the formative experiment.

The rationale for choosing the author’s technology of MA work is based on the principles of systematicity, practice orientation, a differentiated approach, continuity, and feedback. In contrast to traditional forms of methodological work (reports, presentations, literature reviews), the proposed model presupposes a predominance of interactive forms, active practice by each teacher, collective reflection, and the creation of a common product (a bank of prompts).

The experimental work was carried out on the basis of two educational organizations in Belgorod Oblast: the Municipal Educational Institution “Streletskaya Secondary General Education School” of Belgorod District and the Municipal Budgetary Educational Institution “Livenskaya Secondary General Education School No. 1” of Krasnogvardeysky District. The study involved 24 foreign language teachers, as well as heads of methodological associations and deputy directors for educational work. The participants’ ages ranged from 23 to 62, and their teaching experience ranged from 2 to 39 years. Among them, 6 people (25%) were young specialists with up to 3 years of experience, 9 people (37.5%) were teachers with 4 to 15 years of experience, and 9 people (37.5%) were teachers with more than 15 years of experience.

For diagnostics, the author-developed questionnaire “The Use of Neural Networks in the Work of a Foreign Language Teacher” was used, including seven blocks: attitude toward AI, frequency of use, tools used, tasks solved, self-assessment of prompt engineering, main difficulties, and need for methodological support. In addition, an observation map for MA meetings, a lesson analysis map for lessons with AI materials, a questionnaire for students, and an analysis of teachers’ activity products were used. The criteria for MA effectiveness were defined as quantitative indicators (the share of those proficient in prompt engineering, frequency of AI use, and the number of created materials) and qualitative indicators (satisfaction, changes in professional behavior, and activity at meetings).

The formative experiment was conducted from October 2025 to October 2026 and included five stages: preparatory (October 2025), training (November–December 2025), practical (January–March 2026), support and deepening (April–August 2026), and analytical (September–October 2026).

At the preparatory stage, an introductory MA meeting was held, responsible persons were identified, a working group for developing the bank of prompts was created, and mentors were appointed.

The training stage included a master class, “Neural Networks in the Work of an English Teacher: An Overview of Tools” (November 2025), and a practicum on prompt engineering (December 2025). Teachers mastered the prompt formula (role + context + task + format), created and tested their own requests.

The practical stage included a workshop on creating authentic dialogues and visual supports for the development of intercultural communication (January 2026), a round table on checking written work with Grammarly and LanguageTool (January 2026), a project seminar on creating a bank of prompts (February 2026), and peer lesson visits with a focus on AI materials (March 2026).

The support and deepening stage became a fundamentally new element of the program. In April 2026, a prompt battle was held — a competition for the best request to a neural network for solving a methodological task. In May 2026, a round table, “The Ethics of Using Artificial Intelligence in Education,” was held, at which issues of authorship, confidentiality, and academic honesty were discussed. In June 2026, a summer school of digital literacy was organized for teachers who had not been able to fully master neural network tools. In July–August 2026, teachers worked independently on creating original materials, and the bank of prompts was supplemented with another 20 requests.

The analytical stage included a delayed assessment (September 10–25, 2026) and final control diagnostics (the first half of October 2026).

The results of the ascertaining stage (September 2026) showed that noticeable changes had occurred over the year that had passed since the initial introduction to AI. A total of 83.3% of teachers (20 out of 24) had a positive attitude toward AI, 12.5% (3 people) were neutral, and 4.2% (1 person) were negative. A total of 45.8% (11 people) used neural networks regularly (daily or several times a week), 33.3% (8 people) used them episodically, and 20.8% (5 people) did not use them at all.

Among teachers who regularly used AI, 72.7% (8 people) noted that the time required for lesson preparation had decreased by 30–40%. A teacher with 12 years of experience said: “Previously, preparing one lesson with a worksheet, presentation, and dialogues took me two to three hours. Now, with the help of Twee and ChatGPT, I do it in 40–50 minutes. I spend the freed-up time on individual work with students.” However, among those who used AI episodically, 62.5% (5 people) indicated that preparation time had not decreased and sometimes had even increased because of the need for repeated correction.

An analysis of academic performance for the 2025/2026 academic year showed that in classes where teachers regularly used AI materials, the average grade in the subject increased by 0.4 points. In classes where teachers used AI episodically, the increase was 0.2 points. In classes where teachers did not use AI, the average grade remained at the same level. Teachers who regularly used AI noted an increase in student motivation (81.8%), an increase in the share of speaking activity (72.7%), improvement in the quality of visual materials (63.6%), and a decrease in anxiety when preparing for oral answers (54.5%).

Based on the results of the year, four groups of teachers were identified. Enthusiasts (25%, 6 people) actively used AI daily, had mastered prompt engineering at a high level, created original materials, and shared their experience. Pragmatists (20.8%, 5 people) used AI regularly but selectively, for specific tasks. Skeptics (33.3%, 8 people) used AI episodically, were often disappointed, and spent a great deal of time on correction. Refusers (20.8%, 5 people) did not use AI at all, citing lack of time, confidence in the effectiveness of traditional methods, or technical difficulties.

The results of the control stage (October 2026) showed significant positive dynamics compared with the ascertaining stage of 2025. The share of teachers regularly using AI grew from 25% to 75%. The share of those proficient in prompt engineering at an average and high level increased from 37.5% to 87.5%. The share of those using specialized educational neural networks (Twee, Magic School) grew from 12.5% to 62.5%. Satisfaction with MA support was 87.5%. Teachers created and tested more than 50 educational materials with the help of neural networks, and the bank of prompts was supplemented with 35 new requests and amounted to more than 65 units.

Qualitative analysis showed that the use of neural networks contributed to increased student motivation, an increase in the share of speaking activity in lessons, and a decrease in anxiety when preparing for oral answers. Heads of MAs noted that the work of the association had become more interactive and practice-oriented, attendance at meetings had increased, and teachers’ activity had grown.

At the same time, the delayed assessment revealed problem areas: 20.8% of teachers had not been able to integrate AI into their practice and had returned to traditional methods of work; 33.3% used neural networks episodically and needed additional support; difficulties remained related to unfamiliarity with specialized tools (54.2%), the need to correct results (50%), and ethical concerns (25%).

CONCLUSIONS

The conducted longitudinal study allows the following conclusions to be made.

The methodological association of foreign language teachers is an effective form of collective mastery of neural network technologies, provided that relevant activities are systematically included in the work plan and interactive forms are used (master classes, practicums, workshops, project seminars, prompt battles).

Over the year of implementing the formative experiment program, there was a significant increase in teachers’ readiness to use AI: the share of those regularly using neural networks grew from 25% to 75%, those proficient in prompt engineering from 37.5% to 87.5%, and those satisfied with MA support reached 87.5%.

The time spent on lesson preparation decreased by 30–40% among teachers who had mastered AI at a sufficient level. In classes of regular users, positive dynamics in academic performance and student motivation are observed.

Stable groups of teachers formed: enthusiasts (25%), pragmatists (20.8%), skeptics (33.3%), and refusers (20.8%). This indicates the need for a differentiated approach to methodological support.

Significant difficulties remain: unfamiliarity with specialized tools, the need to correct results, and ethical concerns. The need for methodological support remains high (79.2%), but requests are becoming more specific and practice-oriented.

A one-time introduction to AI is insufficient for the sustainable integration of technologies into pedagogical practice. Long-term, systematic work of the methodological association is needed, including a stage of support and deepening.

Prospects for further research are associated with the development of a long-term model of methodological support for teachers in the field of AI, the creation of an online course for heads of MAs on prompt engineering, and assessment of the influence of neural network use on students’ intercultural competence in the delayed period.

References

1. Alferyeva-Termsikos V.B. Prompt engineering as a strategy for developing students’ information culture // International Journal of Humanities and Natural Sciences. – 2024. – No. 9-1(96). – Pp. 10–15.
2. Babansky Yu.K. Selected Pedagogical Works. – Moscow: Pedagogika, 2017. – 560 p.
3. Bim I.L. Theory and Practice of Teaching German. – Moscow: Prosveshchenie, 2019. – 256 p.
4. Bordovsky G.A. Managing the Quality of the Educational Process. – St. Petersburg: Publishing House of Herzen State Pedagogical University of Russia, 2019. – 359 p.
5. Volkova Yu.V. Digital technologies and intercultural communication: the potential of neural networks in teaching English // (in press). 2025.
6. Galskova N.D. Modern Methods of Teaching Foreign Languages. – Moscow: ARKTI, 2018. – 192 p.
7. Goodfellow Ya., Bengio I., Courville A. Deep Learning. – Moscow: DMK Press, 2018. – 652 p.
8. Zinchenko Yu.P., Shaigerova L.A. Methodological problems of studying ethnocultural identity // Ethnocultural Identity as a Factor of Social Stability in Contemporary Russia / Ed. by Yu.P. Zinchenko, L.A. Shaigerova. Vol. 1. – Moscow, 2016. – Pp. 12–28.
9. Kodzhaspirova G.M. Pedagogy: Textbook for Universities. – Moscow: Yurait, 2021. – 719 p.
10. Lebedeva N.M., Tatarko A.N. Methods of Ethnic and Cross-Cultural Psychology. – Moscow: HSE Publishing House, 2011. – 239 p.
11. Passov E.I. The Communicative Method of Teaching Foreign Language Speaking. – Moscow: Prosveshchenie, 2019. – 208 p.
12. Advanced Training for Teachers “Artificial Intelligence: A Start into the Future” // Center for the Development of IT Education at MIPT. – URL: https://edu.mipt.ru/ai/ (accessed: 01.10.2026).
13. Potashnik M.M. Managing Teacher Professional Development. – Moscow: Pedagogical Society of Russia, 2020. – 448 p.
14. Reprintsev A.V. Cultural and educational environment in the formation of the personality of a Russian person // Pedagogika. – 2015. – No. 1. – Pp. 88–96.
15. Soldatova G.U. Psychology of Interethnic Tension. – Moscow: Smysl, 1998. – 389 p.
16. Stefanenko T.G. Ethnopsychology: Textbook for Universities. – Moscow: Aspekt Press, 2014. – 352 p.
17. Khukhlaeva O.V., Khakimov E.R., Khukhlaev O.E. Multicultural Education. – Moscow: Yurait, 2019. – 283 p.
18. Shamova T.I. Management of Educational Systems. – Moscow: Akademiya, 2018. – 384 p.
19. Shamionov R.M., Golovanova A.A. Social Psychology of Personality. – Moscow: INFRA-M, 2017. – 300 p.
20. Deborah Rivas-Drake, Carol Markstrom, Seth J. Schwartz. Ethnic and Racial Identity in Adolescence: Implications for Psychosocial, Academic, and Health Outcomes // Child Development. – 2014. – Vol. 85, No. 1. – P. 40–57.