Integrating appreciative education with AI-assisted oral training for sustainable EFL learning: a study on speaking anxiety and oral proficiency

Background Foreign language speaking anxiety constitutes a persistent barrier to oral proficiency development among Chinese college students; however, prevailing technology-centered instructional approaches inadequately address learners' affective needs, thereby limiting their effectiveness in alleviating anxiety and fostering intrinsic motivation. This study proposed a blended pedagogical model—the “Praise-Speak” program that integrates offline appreciative education with AI-assisted oral training via the FIF platform. Methods A two-phase mixed-methods design was employed. In the first phase, large-scale surveys ( n = 1,044 for anxiety assessment; n = 209 for appreciative education needs) identified freshmen as the priority intervention cohort. In the second phase, a 16-week pilot intervention was conducted with 13 participants, utilizing a pretest-posttest design supplemented by reflective journal analysis. Results The findings demonstrated statistically significant improvements in IELTS speaking scores ( p = 0.005) and a marked reduction in speaking anxiety ( p < 0.001). Qualitative analysis revealed three core mechanisms underlying these outcomes: the establishment of psychological safety, positive receptivity to appreciative pedagogy, and technology-humanism synergy. Conclusion This study offers a replicable pedagogical framework that harmonizes technological empowerment with humanistic care, contributing to the realization of the United Nations Sustainable Development Goal 4.