Page 605 - ISC PROCEEDINGS 21.4

exercises. This personalized learning experience enabled them to learn at a pace they
preferred and be offered the specific support needed hence they achieved positive
vocabulary acquisition outcomes. Also, AI learning tools enhanced the level of
engagement and motivation among learners. It was also reported that creativity of AI-
based activities helped many of the participants to enjoy vocabulary learning, unlike
classroom exercises. The results are in line with the existing research pointing to the
efficiency of AI-based learning systems in enhancing the engagement of learners and
boosting their vocabulary acquisition (Godwin-Jones, 2019).
Students made great improvement in vocabulary scores where the students scored
M = 56.40 (SD = 8.21) in pre-test, and M = 72.85 (SD = 7.95) in the post-test. Paired-
sample t-test revealed that the difference was statistically significant, t (49) = 9.84, p <
0.001, and had a large effect size (d = 1.39). This finding suggests that the vocabulary
acquisition of students was affected significantly by AI-based learning instruments
positively.
Table 1. Vocabulary Pre-Test and Post-Test Results
Test N Mean Score Standard Deviation Mean Difference
Pre-Test 50 56.40 8.21 ------
Post-Test 50 72.85 7.95 +16.45
Source: Author
Indicates the comparison of post-test and pretest scores of vocabulary. The findings
reveal that there is a significant increase in the vocabulary knowledge of students who
utilize AI-supported learning too. The average score was raised to 72.85 and 56.40, which
indicates that AI-related vocabulary tools have a positive impact on vocabulary learning.
6.2. Pronunciation development
The findings of the pronunciation tests also indicated that the pronunciation
accuracy of the learners had improved significantly after using the speech recognition
tools of AI. Before the intervention, a large number of the participants exhibited a
problem with the production of some sounds in English, stress, and intonation. The
causes of such pronunciation challenges are familiar to all ESL learners because of
discrepancies between the phonological systems of their first language and English
(Derwing and Munro, 2015). Following the introduction of AI-assisted pronunciation
practice, the learners demonstrated significant improvement in several aspects such as
the production of phonemes, word stress, and speech intelligibility. Automatic speech
recognition (ASR) technology was very instrumental in this enhancement. The AI devices
enabled students to save their speech and get instant feedback on the speech patterns of
the errors. Such feedback allowed the learners to determine the areas of error and repeat
the pronunciation of the correct words. One more significant benefit of AI-based
pronunciation devices was the possibility to practice on their own. In the conventional
classrooms, there is usually a constraint in teaching pronunciation because of time and
the size of the classroom. The use of AI offered learners the opportunity to rehearse and
speak and repeat without having to present to the audience. This independent learning
space was successful in alleviating anxiety on the learners and making them more
confident when speaking the English language. The results are consistent with the
previous studies, which reveal that speech recognition technologies can be used to
improve pronunciation learning considerably, through customized feedback and the
ability of learners to learn at their own pace (Liakin, Cardoso, and Liakina, 2015).

604