Методика преподавания языка | Филологический аспект: Методика преподавания языка и литературы Методика преподавания языка и литературы №02 (25) Март 2024 - Апрель 2024

Возможности использования чат-ботов на базе искусственного интеллекта в преподавании английского языка как иностранного для формирования коммуникативных умений студентов

Падалецкий Евгений Алексеевич
магистрант института международного образования, Московский педагогический государственный университет РФ, г. Москва
Смирнова Наталья Владимировна
канд. филол. наук, доцент кафедры английского языка и цифровых образовательных технологий, Московский педагогический государственный университет, РФ, г. Москва, Natalia22L2016@yandex.ru

Аннотация: В статье рассматривается исследование возможности применения чат-ботов на базе искусственного интеллекта для формирования коммуникативных умений у студентов, изучающих английский язык как иностранный. Статья содержит краткую историю чат-ботов. Авторы рассматривают коммуникативные умения и основные проблемы, возникающие при их формировании. В статье рассматриваются возможные стратегии применения чат-ботов для формирования коммуникативных умений студентов. Статья содержит описание результатов эксперимента по использованию чат-ботов как инструмента формирования коммуникативных умений студентов.
Ключевые слова: чат-боты на базе искусственного интеллекта, коммуникативные умения, английский язык как иностранный

Padaletskii Evgenii Alekseevich
Master student of Institute of International Education, Moscow State Pedagogical University, Russia, Moscow
Smirnova Natalia Vladimirovna
Cand. Sci. (Philology), assistant professor of English language and IT technologies department, Moscow State Pedagogical University, Russia, Moscow

Abstract: The article considers a study on the opportunities of the use of AI-powered chatbots to develop communication skills among EFL students. The article contains a brief history of chatbots. The authors consider communication skills and the main problems that arise during their development. The article discusses possible strategies for the use of chatbots to develop students' communication skills. The article contains a description of the results of an experiment on the use of chatbots as a tool for the development of EFL students' communication skills.
Keywords: AI-powered chatbots, communication skills, EFL

In an era of rapid technological advancement, the incorporation of artificial intelligence has brought novel approaches to teaching. The increasing use of AI-powered chatbots is one remarkable development. AI-powered chatbots are computer programs that are able to simulate human-like conversations with users. These chatbots have become useful educational resources, particularly for language learning. They can do more than just simple automation; they are capable of offering customized and interactive learning. The modern world makes it more and more necessary for people to have access to contemporary technical tools that allow them to practice their language skills at any time, anywhere, and on any schedule. In order to gain essential knowledge for the changing educational landscape, we explore the influence of chatbots on improving communication skills in students learning English as a foreign language.

The history of chatbots dates back to 1950 when Alan Turing introduced the concept of a computer program capable of having conversations with people without revealing its artificial nature, known as the Turing test. (Turing, 1950) The first chatbot, ELIZA, was developed in 1966 at MIT by Joseph Weizenbaum. ELIZA could engage in basic natural conversations by analyzing input sentences and generating human-like responses. (Neff & Nagy, 2016) In 1972, the chatbot PARRY was created, simulating a patient with schizophrenia and showing more advanced emotional responses compared to ELIZA. (Güzeldere & Franchi, 1995) Jabberwacky, the first AI-powered chatbot, emerged in 1988, utilizing contextual pattern matching to generate responses based on previous conversations. (Wallace, 1995) The term "chatterbot" appeared in 1991 within the online game TINYMUD, where a chatbot interacted with players, proving to be more engaging than human players. This early success paved the way for further advancements in conversational AI technology. In 1995, ALICE was introduced as an Artificial Linguistic Internet Computer Entity, focusing on simulating human-like conversations through pattern matching and rule-based responses. ALICE demonstrated advanced conversational skills and participated in the Loebner Prize competition, contributing significantly to the field of artificial intelligence. An important moment in chatbot technology happened in 2001 with the creation of SmarterChild, which provided real-time information from databases and was integrated into messaging platforms like AOL and Microsoft, enhancing human-machine interaction. (Hoffer et al., 2001) The evolution of chatbots progressed with the introduction of personal voice assistants in smartphones and smart speakers, offering quick and meaningful responses to oral commands for various tasks like home device monitoring, internet searches, calls, and weather updates. Finally, in November 2022, OpenAI released ChatGPT, a groundbreaking AI language model based on GPT-3 architecture, revolutionizing natural language processing and conversational AI by generating human-like text responses across different communication contexts, displaying continuous development and commitment to enhancing communication opportunities for users. (Aljanabi et al., 2023)

In the field of TEFL, communicative language teaching has gained widespread acceptance globally as an approach that prioritizes meaningful communication, fluency, and real-life language use over grammar accuracy. (Pemberton, 2024) Originating in the 1960s and 1970s, CLT was developed to address the limitations of traditional language teaching methods like grammar-translation and the audio-lingual method. Influenced by scholars like Noam Chomsky, CLT emphasizes the functional and communicative aspects of language, encouraging learners to use language authentically and creatively. The main principles of CLT focus on communication as the primary goal, authentic language use, student-centered learning, integrating personal experience into classroom activities, and connecting in-class tasks with outside activities. (Nunan, 1991)

The literature review suggests that the core principles of CLT can be effectively implemented in EFL classrooms through AI-powered chatbots. These chatbots enable real-time communication in the target language, simulate authentic language scenarios, adapt to individual learner needs, incorporate personal experiences into interactive tasks, and bridge in-class and out-class activities by facilitating social interaction in the target language. (Ruan et al., 2021)

AI-powered chatbots offer a range of benefits that complement traditional learning methods. While they cannot fully replace interactions with native speakers due to their artificial nature, these computer systems bring unique advantages to the EFL classroom. AI chatbots analyze learner’s messages and provide personalized feedback connected with students’ individual strengths and weaknesses. (Jiang, 2022) For example, platforms like TalkPal can analyze voice or text messages, offering corrections and explanations promptly. This feedback is often more consistent and unbiased than human feedback, as it follows predetermined criteria and algorithms.

One of the key advantages of AI-powered chatbots is their ability to individualize learning experiences. These chatbots can adjust to students' specific needs and learning pace, ensuring they are appropriately challenged without feeling overwhelmed. (Sun & Zhou, 2023)

Moreover, the constant availability of AI-powered chatbots is a significant benefit for learners. Unlike human teachers or classmates, these chatbots are accessible at any time, allowing students to engage in language practice and seek clarification whenever they need it. This on-demand assistance promotes continuous learning and enables students to progress at their own pace. (Grudin & Jacques, 2019)

Furthermore, AI-powered chatbots enhance interaction in the language learning process. They provide opportunities for students to practice speaking, writing, listening, and reading skills in communicative contexts. Through interactive dialogues and role-play scenarios, students can immerse themselves in authentic communication experiences, improving their language proficiency and confidence in communication in English. (Perez et al., 2020)

Nevertheless, AI-powered chatbots have limitations that hinder their ability to fully replace human interaction. Among these limitations there is a lack of human touch crucial for interpersonal skills, challenges in understanding language nuances, potential biases from training data, and difficulties with idioms and colloquial phrases. (Roos & Lochan, 2022)

AI-powered chatbots have made significant steps in the realm of language learning, with such platforms like Duolingo, Mondly, and Andy. For instance, Duolingo integrated chatbots into its iOS app back in 2016 to facilitate real-life conversation practice across multiple languages. These chatbots guide users by suggesting words and phrases, connecting interactions to individual proficiency levels. However, they primarily focus on structured dialogues rather than open-ended speech interactions. Similarly, Mondly's chatbots, available on Android and iOS, immerse users in scenarios like ordering at a restaurant to reinforce basic language skills effectively. On the other hand, Andy acts as a virtual English teacher, aiding users in grammar practice, vocabulary expansion, casual conversations, and language games. While not the most advanced, Andy proves particularly beneficial for English beginners, offering a supportive learning environment for both Android and iOS users. (Floris et al., 2021)

AI-powered chatbots have a promising future in language education, offering personalized, engaging, and efficient learning experiences compared to traditional methods. These chatbots can adapt to various learning styles, and provide 24/7 availability for practice. They create immersive environments, enhance student engagement, offer instant feedback, and personalized support, ultimately improving learning outcomes.

Communication skills are crucial in TEFL for successful language learning, learner engagement, and classroom effectiveness. As TEFL evolves, educators increasingly acknowledge the pivotal role of communication skills in facilitating language acquisition and creating a positive learning environment. Defined as "the ability to convey information and ideas effectively," (Collins Dictionary) communication skills are essential for effective language communication and learning.

Communication skills encompass both verbal and non-verbal aspects. Verbal communication involves listening, reading, writing, and speaking, requiring proficiency in vocabulary, grammar, pronunciation, and fluency to express thoughts clearly. (Hamlin et al., 2024) Active listening and contextual communication are essential components of verbal skills. On the other hand, non-verbal communication includes gestures, facial expressions, and body language, adding depth to messages and enhancing interpersonal connections. Written communication involves reading and writing with a focus on clarity, coherence, grammar, and organization. Oral communication comprises listening and speaking, emphasizing effective spoken interactions through clear articulation, understanding non-verbal cues, adapting to the audience, and active listening by summarizing and confirming understanding. (Stephen, 2024) The main information about communication skills is presented in the table below:

Table 1. Types of communication

The multifaceted nature of communication skills requires integrative methodologies that are able to develop the language skills and advanced aspects of communication skills presented in the table above. The literature review has shown that the following integrative methodologies ensure the development of communication skills:

• Communicative language teaching that focuses on interactive tasks, role-plays, simulations, and communicative games that allow learners to use language for its communicative purposes. (Pemberton, 2024)
• Task-based learning and project-based learning, which are parts of CLT. TBL presupposes that learners deal with real-world tasks that require the use of language in context. In PBL activities learners need to communicate using the language in order to create a project, presentation, conduct a research, etc. (Welling, 2024)
• Content-based instruction presupposes language instruction with relevant material form other academic courses. Students are involved in projects, conversations, and activities that presuppose the use of language to understand and transmit information. (Snow, 2023)
• Content and language integrated learning that presupposes teaching of academic subjects in the target language. (Dalton-Puffer, 2023)

In order to develop communication skills effectively, we also need to take into account the barriers that appear when learners need to communicate in English. These problems can be of three types: psychological, linguistic, and environmental.  Among psychological problems scholars single out exam anxiety, communicative worry, and fear of getting poor marks. Linguistic problems are mainly represented by limited vocabulary and grammar knowledge that prevents learners from being accurate and fluent when communicating in English. Besides, we can single out problems with intonation and pronunciation. Finally, when learners do not have enough exposure to authentic foreign language use, they face environmental problems. (Xu, 2023)

The literature review has revealed that AI-powered chatbots can be implemented in TEFL in two ways: daily conversations and role-playing scenarios. Daily conversations presuppose planned or unplanned communication on daily topics, while role-playing scenarios means intentional imitation of one of the communicative acts, such as checking-in at the hotel. The importance of role-playing scenarios can be explained by several factors, such as opportunities to apply English in authentic contexts; meaningful communication; increase in cultural awareness and empathy; encouragement of creativity and critical thinking. Additionally, AI-powered chatbots can be used for ordinary communication with the aim of overall development of communication skills in oral and written form. (Yang et al., 2022) Thus, AI-powered chatbots can help solve psychological, linguistic, and environmental problems that learners face during learning English by reducing learners' anxiety and overall stress, providing new lexical and grammatical units within the communication process, and making a significant impact on learners' immersion and fluency. Nevertheless, it is essential to maintain a balance between the use of AI-powered chatbots and human interaction in the EFL classroom, as human connection is essential for promoting social and emotional growth, cultural awareness, and fluency in everyday communication situations. (Thadphoothon, 2020)

The study adjusted the experimental approach aimed at proving the efficiency of AI-powered chatbots use in developing communication skills of EFL students. The researcher selected four existing groups of his private students enrolled in B2 general English course. The two groups (experimental and control) were formed from the four groups. The experimental group had 15 participants and the control group consisted of 12 learners. The average age of participants was 17-18. The research involved independent (AI-powered chatbots and traditional communication in class) and dependent (communication skills) variables. The duration of the experiment was 16 weeks including 32 classes (48 contact hours, 3 hours per week). The participants had never tried the use of modern technologies for learning.

Both groups were offered pre- and post-tests aimed at identifying any significant changes in variables. The tests were identical in structure, however different in content.

During the experiment, both qualitative (questionnaires before and after the intervention) and quantitative (pre- and post-tests) data were gathered.

The intervention consisted of three stages. The pre-test was administered to both groups during the first phase. The test had four tasks: oral monologue, oral dialogue, written monologue, written dialogue, and its aim was to measure the participants’ level of communication skills development. In order to assess the pre- and post-tests the official scales of the Cambridge exams were used. Additionally, the experimental group was offered a questionnaire in an effort to learn more about the students' attitudes toward their English communication abilities. Throughout the while-intervention phase, there were 30 lessons (45 contact hours). During this phase, both groups had regular general English lessons using a curriculum based on the Macmillan Education textbook Gateway. The experimental group also got daily assignments designed to encourage the usage of AI-powered chatbots outside of the classroom. Connected to the subjects taught in contact classrooms, the assignments gave students more opportunity to practice speaking and writing in dialogues. For the daily tasks we chose two chatbots: Replika for written assignments, and TalkPal for oral tasks. Each task was in the form of a question that was a starting point for a conversation with the chatbot. In this article we provide an example of one of such tasks:

While observing the groups, the researcher collected data that was important to the study. A comparison of the pre- and post-test findings was used in the study's final phase to discover changes in the dependent variable. Additionally, following the experiment, questions on the experimental group's opinions regarding the employment of AI-powered chatbots were posed to the students. After the intervention, the investigator examined and evaluated the collected data, interpreted the results, and made the findings public.

The results of the pre-test showed that both groups had the average level of communication skills development. The mean score of the experimental group was 63 points out of 80 possible, while the control group had 61 points. Thus, the difference in the development of communication skills between the experimental and the control groups was 2.5 points.

Table 2. Mean Score of the Pre-test

After the intervention, the participants were offered the post-test that was structurally identical to the pre-test, however the content was different and adapted to the students’ acquired level of overall English proficiency. According to the results of the post-test, the students of the experimental groups had significantly increased their level of communication skills development. The experimental group had 69 points out of 80, and the control group got 63 points. The gap between the groups was 6 points. Thus, we can conclude that the implementation of AI-powered chatbots into the EFL classroom has a positive impact on students’ communication skills development.

Table 3. Mean Score of the Post-test

The experimental group was also offered questionnaires before and after the experiment. The aim of the questionnaire before the experiment was to understand the students’ own perception of their communication skills in English. The questionnaire after the experiment was aimed at identifying the students’ impression on the use of AI-powered chatbots and their opinion considering the development of their communication skills. The results of the questionnaire offered before the intervention, displayed that the majority of the participants (53.3%) had an average level (“neutral” in a Likert scale) of confidence about their level of English, while 46.7% responded that they “agreed” that they had confidence about their English level. Then, 100% of the participants chose “neutral” for the statement “It is easy for me to communicate in English”. As for the difficulties arising in communication in English, 60% of the learners ticked limited vocabulary, 40% - limited grammar, 26.7% - pronunciation difficulties and difficulties with understanding informal language, 33.3% - difficulties with understanding spoken language, 46.7% - fear of making mistakes, and 53.3% of the respondents claimed the lack of practice as a difficulty that prevents them from the effective communication in the English language. Additionally, the majority of the learners (66.7%) claimed that improving English is important for them. Notably, 86.7% and 80% of the participants responded that it is significant for them to practice English in real life and have an access to interactive learning materials. Finally, the most of the learners (73.3% and 60% respectively) do not practice English outside the classroom and do not use any technological and/or online resources for learning English.

Furthermore, the questionnaire administered after the intervention showed that 73% of the participants noticed that it became easier for them to communicate in English. Besides, 80% of the learners stated the increase in their vocabulary, 46.7% - of grammar. More than a half (66.7% and 60% respectively) of the participants admitted that the chatbots helped them to achieve a desirable level of English practice and reduce the level of fear of making mistakes. As the most useful features of the chatbots the participants chose their interactivity and constant availability. Besides, most of the learners (86.7%) agreed that started feel more confident about their English abilities. Moreover, 53.3% and 26.7% of the respondents were satisfied or strongly satisfied with the experience of the use of the chatbots. Additionally, 46.7% of the learners considered AI-powered chatbots to be a valuable tool for improving communication skills in English, and 73.3% of the participants claimed that they would continue the use of chatbots for practicing English.

In conclusion, the implementation of AI-powered chatbots into the EFL classroom leads to significant increase in students’ communication skills as well as their confidence and motivation to improve their knowledge of the English language.

1. Aljanabi, Mohammad & G. Yaseen, Mohanad & Ali, Ahmed & Abed, Saad & Chatgpt,. (2023). ChatGpt: Open Possibilities. 4. 10.52866/20ijcsm.2023.01.01.0018.
2. Dalton-Puffer, Christiane. (2023). CLIL as a Pedagogical Approach. 10.1002/9781405198431.wbeal20465.
3. Floris, Flora & Suseno, Eunike & Ananda, Ellara & Hia, Irene. (2021). Chatting with Andy.
4. Grudin, J.T., & Jacques, R.D. (2019). Chatbots, Humbots, and the Quest for Artificial General Intelligence. Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems.
5. Güzeldere, G., & Franchi, S. (1995). Dialogues with colorful “personalities” of early AI.
6. Hamlin, Emily & McGloin, Rory & Bridgemohan, Alex. (2024). Communication skills training: a quantitative systematic review. Development and Learning in Organizations: An International Journal. 10.1108/DLO-08-2023-0188.
7. Neff, G. (2016). Talking to Bots: Symbiotic Agency and the Case of Tay. International Journal of Communication.
8. Nunan, D. (1991). Communicative Tasks and the Language Curriculum. TESOL Quarterly, 25(2), 279-295. http://dx.doi.org/10.2307/3587464
9. Pemberton, Ian. (2024). Communicative Language Teaching. 10.1007/978-3-031-53414-0_5.
10. Perez, E., Perez, M., & Guerrero, R. (2020). The impact of AI-powered chatbots on language learning: A systematic literature review. International Journal of Artificial Intelligence in Education, 30(3), 1-35. https://doi.org/10.1007/s40593-020-00215-1
11. Roos, R., & Lochan, R. (2022). Chatbots in Education: A Systematic Review. Journal of Educational Technology Development and Exchange, 15(2), 1-25. https://doi.org/10.1186/s41239-022-00250-1
12. Ruan, Sherry; Jiang, Liwei; Xu, Qianyao; Liu, Zhiyuan; Davis, Glenn M; Brunskill, Emma; Landay, James A. (2021). EnglishBot: An AI-Powered Conversational System for Second Language Learning. In Proceedings of the 26th International Conference on Intelligent User Interfaces (IUI '21). Association for Computing Machinery, New York, NY, USA, 434–444. https://doi.org/10.1145/3397481.3450648
13. Snow, Marguerite. (2023). Content‐Based Instruction as a Pedagogical Framework. 10.1002/9781405198431.wbeal0192.pub3.
14. Stephen, Jacqueline. (2024). Skills and Strategies for Communication. 10.1007/978-3-031-54439-2_11.
15. Thadphoothon, Janpha. (2020). Chatbots as Conversational Partners in the EFL/ESL Context. 10.13140/RG.2.2.26837.37604.
16. Turing, Alan M. (1950). Computing machinery and intelligence. Mind 59 (October):433-60.
17. Welling, Joanne. (2024). PPP vs TBL. 10.1017/9781009024808.013.
18. Xu, Cheng. (2023). College students’ communication disorder and anxiety: self-efficacy regulation analysis. CNS Spectrums. 28. S119-S119. 10.1017/S1092852923005539.
19. Yang, Hyejin & Kim, Heyoung & Lee, Jang Ho & Shin, Dongkwang. (2022). Implementation of an AI chatbot as an English conversation partner in EFL speaking classes. ReCALL. 34. 1-17. 10.1017/S0958344022000039.