Formation of lexical and grammatical skills of technical university students based on phrasal verbs of the English language through practice with a chatbot

Importance. The current stage of digital transformation of education, supported by global initiatives from UNESCO and the national development strategies of the Russian Federation until 2030, is characterized by the integration of innovative technologies such as artificial intelligence (AI) technologies into the higher education system. This area is particularly relevant in technical universities, where the formation of professional foreign language communicative competence is the main goal of foreign language education. Phrasal verbs, widely used in professional communication and technical documentation, remain one of the most difficult lexical and grammatical categories of the English language for students of future engineers. However, in practice, their study is complicated by the lack of a systematic approach in textbooks, a shortage of classroom hours, and the low effectiveness of traditional methods of teaching vocabulary and grammar. To achieve the main purpose of the study, chatbots are used as a tool for extracurricular speech practice aimed at automating the skills of using phrasal verbs. The purpose of this study is to determine the effectiveness of the methodology for the formation of lexical and grammatical skills of technical university students based on phrasal verbs of the English language through practice with a chatbot.

Research Methods. To conduct this study, the following groups of methods were used: a) theoretical: study and analysis of domestic and foreign pedagogical, methodological, and psychological scientific literature on the research problem; analysis and generalization of experience on the problem under study, modeling; b) empirical: observation, questioning, interview, testing; c) statistical: quantitative and qualitative analysis results obtained, mathematical data processing; d) formative: experimental work was carried out under controlled conditions to verify the author's teaching methodology; the data obtained were used to analyze and identify cause-and-effect relationships between variables. The effectiveness of the methodology for the formation of lexical and grammatical skills of technical university students based on phrasal verbs through practice with a chatbot was carried out by comparing the test results at the ascertaining and control stages. IBM SPSS Statistics 21 software was used to analyze the data, and the averages were compared according to the parametric statistical method, the Student’s t-test. The object of control was 15 lexical and grammatical skills (9 receptive and 6 productive skills).

Literature Review. During the analysis of scientific and educational literature on the research topic, the results of scientific papers on the use of chatbots for the formation of foreign language communicative competence were summarized, and the main research directions were identified.

Results and Discussion. The developed methodology for the formation of lexical and grammatical skills of technical university students based on phrasal verbs of the English language through practice with a chatbot was tested during experimental study and proved its effectiveness in comparison with traditional teaching methods in all controlled parameters. The conducted experimental study revealed that at the initial stage of training, both groups of control group (CG) (N = 24) and experimental group (EG) (N = 24) have an equal level of language training, since there was no statistical significance between the groups (p > 0.05 for most of the controlled parameters). The indicators in the EG have a high statistical significance (p < 0.001), which indicates the pronounced effectiveness of the applied teaching methodology. In CG, most indicators are also significant, but the significance level varies between p < 0.001 (parameters 2–8, 10–11, 13–15) and p < 0.05 (parameters 1, 9, 12). The least pronounced and having weak statistical significance in CG is the controlled parameter No. 1 “to correlate the sound form of a word with its meaning” (*p = 0.037), while in EG, despite the fact that it was better formed, it nevertheless has a low t = 4.87 (p < 0.001) compared to with other skills. The difference in effectiveness between СG and EG is obvious due to the fact that the Student’s t-test scores in EG are significantly higher than in СG, for example, for parameter No. 6 “to differentiate phrasal verbs and similar monolex verbs” t = 4.67 (СG) versus t = 7.89 (EG). The largest gap in the Student’s t-criterion is observed in productive skills, for example, for parameter No. 9 “to predict grammatical constructions with phrasal verbs” t = 4.12 (CG) versus t = 8.12 (EG).

Conclusion. Experimental study has revealed that chatbots ChatGPT and DeepSeek do not have the technical capability to develop listening skills. At the same time, some chatbots (for example, Replika AI) provide more natural communication, TalkPal chatbots and Praktika.ai are more effective for correcting errors. Speech practice with a chatbot aroused the interest of students, which helped to increase their motivation to learn English. The interactive learning format and the ability to receive instant feedback have made the process more fun and accessible. Using several different chatbots at the same time allowed students to familiarize themselves with the possibilities provided by chatbots and choose the one that best suited the needs of a particular student. The results of the research can be used in further study to identify the linguistic and didactic potential of AI technologies and its application in the framework of foreign language training for students of nonlinguistic universities, the formation of lexical and grammatical skills of students through other AI technologies, as well as in the methodology of teaching a foreign language.