The Impact of Digital and Artificial Intelligence Technologies on the Improvement of Foreign Language Listening, Speaking, Reading and Writing Skills: A Meta‐Analysis

Abstract

With the continuous advancement of educational technology, the application of digital and artificial intelligence (AI) technologies in foreign language learning has gradually become an effective means of enhancing language skills. Utilising personalised feedback, intelligent learning support and instant interaction, these technologies have been widely adopted for cultivating foreign language skills such as listening, speaking, reading and writing. However, existing meta‐analysis primarily focuses on the effects of single technologies or individual language skills, lacking a systematic assessment of the comprehensive impact of various technologies on language skill improvement. This study aims to systematically evaluate the overall effect of digital and AI technologies on the improvements in foreign language listening, speaking, reading and writing skills through a meta‐analysis, particularly examining their moderating effects across different technology types, educational levels and intervention conditions. The research objectives are to reveal the overall impact of these technologies on the improvements in foreign language listening, speaking, reading and writing skills and to explore the associated disparities and key influencing factors. This study employed a meta‐analytic approach, synthesising 40 relevant empirical studies published between 2014 and 2025, encompassing 3367 participants. The standardised mean difference (SMD) was calculated, and a random‐effects model was used to estimate the overall effect size. Additionally, the influence of moderating variables—such as sample stratification, learning outcomes, research design and technology type—on the learning effects was analysed. The results indicate that digital and AI technologies have a significant positive effect on foreign language proficiency, with an overall effect size of 0.962, representing a large effect. Significant between‐group differences were identified for research design, intervention duration and the number of technologies used. In contrast, no significant between‐group differences were found across language skill types, educational levels, sample sizes, learning environments or technology types. Although positive effects were observed across the listening, speaking, reading and writing subgroups, the differences among these skill categories were not statistically significant. Similarly, both AI and general digital technologies showed significant positive effects, but their between‐group difference was not significant. The effects associated with intervention duration and the number of technologies used should be interpreted cautiously, as the patterns were not linear and some subgroups contained only a small number of studies. The findings demonstrate that digital and AI technologies can effectively support the development of foreign language listening, speaking, reading and writing skills. However, their effectiveness should not be attributed simply to a particular technology type, language skill category or learning environment. Instead, the results suggest that study design, intervention duration and the way multiple technologies are combined may influence estimates of intervention effectiveness. In particular, longer intervention duration or the use of more technologies does not necessarily lead to better learning outcomes. These findings highlight the importance of aligning technological interventions with instructional objectives, learner characteristics and specific implementation conditions. This study provides empirical evidence for the application of digital and intelligent technologies in foreign language education and offers guidance for the design and evaluation of future technology‐supported language learning interventions. The main methodological contribution of this study lies in its integration, via meta‐analysis, of empirical evidence regarding the enhancement of foreign language skills through digital and AI technologies. Furthermore, by examining moderating variables such as language skill type, technology type, learning environment, intervention duration and study design, the study provides a basis for explaining heterogeneity across studies. The findings suggest that the effectiveness of technology‐supported foreign language learning should be understood in the context of specific teaching objectives, learner characteristics and implementation conditions, rather than being attributed simply to a particular category of technology or the tool itself. Future research should further strengthen comparisons of different technological mechanisms, language skill types and intervention conditions, whilst enhancing the rigour of research designs to establish a more robust evidence base.