Effects of AI-based educational applications on EFL students’ speaking and writing performance: A pre-experimental study

Resumen

In the Ecuadorian educational context, particularly at the secondary school level, traditional methodologies centered on memorization and grammatical instruction of the English language still predominate, which has constrained the development of communicative competencies, especially oral and written expression. In response to this issue, the study was conducted in a school setting where low levels of performance were observed in pronunciation, fluency, spelling, and punctuation, as well as a limited use of artificial intelligence–based educational applications. The aim of the study was to determine the impact of implementing practical exercises supported by ELSA Speak and Writing & Improve on the enhancement of integrated oral and written production skills in secondary school students. The research adopted a quantitative approach with a pre-experimental pretest–posttest design. A validated structured questionnaire was administered to a sample of 60 students. The data were recoded into comparative categories and converted into proportions for inferential statistical analysis. Subsequently, a Shapiro–Wilk normality test and a paired-samples t-test were conducted using SPSS software. The results demonstrated a significant improvement across all assessed skills following the didactic intervention. Statistical analysis confirmed significant differences between the pretest and posttest ( < 0.05). The findings indicated that the pedagogically contextualized integration of tools such as ELSA Speak and Writing & Improve promotes more active, autonomous, and meaningful learning. Nevertheless, the study could be strengthened by including a control group and extending the intervention period, thereby opening the possibility for future research.