Using generative artificial intelligence to develop English speaking skills in large classes at a non-linguistic university
TL;DR
The study confirmed the hypothesis that the teacher’s use of GenAI for the synchronous creation of learning materials in a large student group significantly optimizes the lesson, as evaluated against five criteria: instructional time density, student speaking practice coverage, the intensity of speech practice, student engagement, and the number of completed tasks.
Abstract
The aim of this study is to develop and pilot a methodology for the classroom use of generative artificial intelligence (GenAI) by an English teacher for the synchronous generation of educational materials designed to enhance speaking skills (discussion prompts, translation tasks, monologue talking points, and group discussion scenarios) in large groups of economics students (comprising 30-35 people) at non-linguistic universities. The article substantiates the relevance of using generative artificial intelligence under conditions of large class sizes and classroom time constraints; analyzes theoretical approaches to integrating artificial intelligence into foreign language teaching; proposes a proprietary methodology for the synchronous generation of four types of oral practice tasks using prompts; describes the course and results of experimental instruction, which established quantitative performance indicators; and provides practical recommendations for implementing the proposed approach. The scientific novelty of the work lies in the substantiation of a new type of pedagogical activity – the synchronous, real-time generation of educational materials in the classroom; the development and piloting of a methodology for using generative AI as an operational tool for managing speaking practice in large groups; the quantitative verification of efficiency criteria for this methodology; and the identification of the limits of the proposed approach’s applicability, which are associated with the mandatory presence of a teacher for substantive verification and correction of the generated content. As a result, the study confirmed the hypothesis that the teacher’s use of GenAI for the synchronous creation of learning materials in a large student group significantly optimizes the lesson, as evaluated against five criteria: instructional time density, student speaking practice coverage, the intensity of speech practice, student engagement, and the number of completed tasks.