• Şaziye Kanyılmaz

  Author

A new technology called Generative AI is a promising solution to boost personal language learning; however, there is still insufficient empirical evidence about its impact on improving measurable English language skills using interactive storytelling. In this research, an attempt was made to test and develop a generative AI-based framework for creating interactive stories for intermediate English as a second language (ESL) learner. A one-group pretest-post-test quasi-experiment design was utilized with the involvement of 60 intermediate ESL learners who took part in the six-week intervention and completed two 40-minute interactive storytelling sessions per week. Speaking fluency, vocabulary use, narrative comprehension, and learner confidence were measured before and after the intervention by means of the paired-sample t-test and the calculation of effect sizes (Cohen's d). A survey of engagement was also performed after the intervention to assess students' attitudes toward the proposed framework. Significant gains in all skills were observed – speaking fluency increased from 58.4% to 74.6%, vocabulary use increased from 61.2% to 78.3%, narrative comprehension increased from 60.5% to 77.1%, and confidence increased from 55.8% to 73.9%. All improvements turned out to be highly significant (p < .001) and very large (Cohen's d = 2.59–2.95).  Also, 86% of students stated that the storytelling tasks were fun, 81% were ready to continue using the AI-powered tool, and 79% noted their increased self-confidence in speaking English. Thus, the preliminary results of a one-group pilot experiment show that generative AI-based interactive storytelling can be considered a promising methodological approach that is learner-centered and effective in helping students learn English; however, due to the lack of a control group in the experiment, further research needs to be done to verify these results.

• PDF