Smart Virtual Assistant: Generating Pedagogical Activities to Enhance Perceived Self-Efficacy and Learning Performance

Abstract

The swift assimilation of artificial intelligence in the field of education has led to the emergence of intelligent virtual assistants who can generate adaptive, learner-centered pedagogical activities. The present study aimed to shed light on the effects that Smart Virtual Assistant (SVA)-generated educational activities, implemented through two different practice patterns — individual guidance and small-group collaboration — could have on undergraduate students’ academic performance and perceived self-efficacy in an English language learning context. Thus, fifty nursing students enrolled in an English course were randomly assigned to either an individual or a small group SVA-supported learning condition within the framework of a quasi-experimental post-test only design. Over eight weeks of intervention, during which structured adaptive activities generated and scaffolded by SVA were logically integrated into their lessons, data were collected via Logical Thinking Test (in English) results, as well as figures for multidimensional Perceived Self-Efficacy Scale results, which assessed perception across three dimensions: academic, social, and emotional. The above results, therefore, support the proposition that when SVAs are capable of providing adaptive guidance and feedback, along with mastery-oriented activities within lessons, the inherent pedagogical affordances of the assistant itself may outweigh the influence of activity format. The study advances AI-enhanced TEFL research by conceptualizing SVAs as generators of pedagogical activities and demonstrating their capacity to support the achievement of learning outcomes and self-efficacy under various instructional arrangements.