The Influence of Artificial Intelligence-mediated Speaking on EFL Learners’ Performance in Terms of Writing Accuracy and Fluency

The objective of the present study was to investigate the influence of artificial intelligence-mediated speaking on EFL learners’ performance in terms of writing accuracy and fluency. A total of 60 participants were selected from the population of intermediate EFL learners from a private language institute in Shiraz, Iran. The first language of all of the participants was Persian, and English was considered a foreign language for them. Intermediate learners were selected based on their scores on OPT. The participants whose scores were from 40 to 50 were selected as the intermediate learners. Then, the study was run in two classes: the Control class (n = 30) and the experimental class (n = 30). In order to evaluate the writing proficiency of the EFL learners, two 50-minute timed-writing activities were given to the learners before and after the treatment as the pre- and post-tests, respectively. The writing pre-test of both groups was run before the treatment. The experimental group learners obtained sixteen 60-minute experiment sessions with the English Learner Speech Analyzer, while the control group obtained no treatment (following the conventional methods of instruction). All learners in the control and experimental groups were post-tested at the end of the research. The post-test scores of two groups were contrasted to find out the influence of artificial intelligence-mediated speaking on EFL learners’ performance in terms of writing accuracy and fluency. To do this, in addition to descriptive statistics, an Independent-Samples t -test was run. The findings of the study indicated that learners’ writing in the experimental group was improved and learners indicated improvements in terms of accuracy and fluency.