The particular dialect or language that a person chooses to use on any occasion is called a code 2270 Copyright © 2024 The Author IDEAS is licensed under CC-BY-SA 4.0 License Issued by English study program of IAIN Palopo IDEAS Journal of Language Teaching and Learning, Linguistics and Literature ISSN 2338-4778 (Print) ISSN 2548-4192 (Online) Volume 13, Number 1, June 2025 pp. 2270 - 2284 The Use of Yoodli as Media to Improve Speaking Skills in English Education Students Muammar Suharso1*, Ahmad Amin Dalimunthe2 1.2Universitas Islam Negeri Sumatera Utara Corresponding-Email: mu’ammar0304213114@uinsu.ac.id Received: 2025-06-04 Accepted: 2025-06-25 DOI: 10.24256/ideas. v13i1.5364 Abstract This research aims to determine whether there is a substantial relationship between the speaking Ability of second-year students at the Public Islamic High School in Medan and the speaking Ability of second-year students, who are assessed through an experimental and control group methodology. The present study's sample comprised 68 students from class XI Sains A 1 & XI Sains A 2. The data were collected by implementing two distinct evaluation tools: a speaking test utilizing the Yoodli platform in class A1 and a test without using Yoodli in class A2. The findings of this study indicated that second-year students at Public Islamic High School in Medan demonstrated higher academic performance in their classes that utilized the Yoodli system than those who did not. The study's findings revealed a substantial improvement in the speaking proficiency of the institution in question. The coefficient correlation had a Rxy distribution of 0.306. This signifies a significant disparity between the two groups. The use of Yoodli significantly impacts the improvement of verbal proficiency. Keywords: Significant, Yoodli, Speaking Ability Introduction To acquire proficiency in English, the acquisition of four language skills, listening, reading, speaking, and writing, is essential. According to (Haycraft, 1978), speaking and writing are productive skills, whereas listening and reading are classified as receptive skills. Vocabulary is the most fundamental aspect of language learning, particularly in spoken language proficiency. Vocabulary is a pivotal component of oral communication. Acquiring a rich vocabulary is conducive to developing language skills in multiple languages. http://u.lipi.go.id/1457703302 IDEAS, Vol. 13, No. 1, June 2025 ISSN 2338-4778 (Print) ISSN 2548-4192 (Online) 2271 Lado (1961) posits that vocabulary comprises words that function as components of grammatical structures. One of the issues is the issue of verbal communication, which is a necessary ability within the productive skill group. Suryani, L. and Argawati, N. O. (2023) posit that effective communication is essential for success. To communicate successfully, it is imperative to acknowledge the pivotal role of speech in this regard. A substantial body of previous research supports the hypothesis that acquiring language skills is complex. A word may be defined as a linguistic unit that can function as an independent utterance. In this regard, it cannot be disaggregated into two or more components that can be categorized based on their similarities. One artificial intelligence that can be used to teach speaking is Yoodli. Yoodli is one of the newest AIs used in ELT for speaking. It is a complimentary artificial intelligence communication coaching tool that assists users in enhancing their communication skills within a personalized and confidential setting, accessible at app.yoodli.ai. Yoodli is an AI-powered coach and a modern communication training medium designed to help its users become better English speakers. So far, few journal articles have discussed the usefulness of noodles. According to Maureen (2024), Yoodli users can choose the focus to be trained, such as filler words, repetition, phrases used, weak words, eye contact, talk time, centring, etc. This focus is very influential in showing one's fluency, confidence and Ability in English. After recording the speech, Yoodli will play it back with a transcript and send a right and wrong analysis of the user's words. So users get immediate feedback on their learning in private, without the pressure and worry of others finding out. Yoodli helps users be more confident and capable of communicating in English. The speaking skill is critical, given that many students elect to learn English to achieve communicative proficiency. Using English as a communication medium with individuals from diverse international backgrounds renders it comprehensible. As Aini et al. (2022) indicate, speaking is the capacity to articulate oneself within a given context, convey actions or conditions, or engage in discursive reasoning and articulate a sequence of thoughts with minimal anxiety. The capacity to claim an individual as proficient in English is contingent upon the language being spoken fluently. Consequently, the Ability to speak is of significant importance. According to Violita & Cholsy (2022), Numerous AI studies have been conducted to determine how instructors and students feel about using AI-based educational tools. According to Sumakul et al. (2022), every instructor and student mentioned the advantages of using AI in the classroom, demonstrating a Ela Carmelia Mukti Sambas, Emeliyah Sukma Dara Exploring English Vocabulary Mastery in Student with ASD At Upper Primary School Level: Case Study of Multisensory Approach Implementation 2272 favourable opinion of AI-based applications. However, university students during the 2018–2019 school year had a poor perception of AI-based applications (Keles & Aydin, 2021). Generational disparities and differing perspectives are believed to affect the success of AI-based solutions in education. Everyone has a unique perspective on the world, especially in unfamiliar surroundings. Therefore, it could be different. This research examined the use of Yoodli AI by seeing the impact of using Yoodli on students' perception in the speaking classroom. Following the 2013 curriculum, students must develop oral and written communication competencies. Various literary genres, such as narrative, recount, and descriptive text, are recommended for students to study in the 2013 curriculum. Following the initial observation at the Public Islamic School in Medan, the author engaged fully in the educational process, contributing to discussions, describing individuals or locations, and engaging in dialogue. During speaking activities, students experience challenges articulating their thoughts, attributable to the crucial function of vocabulary in communicating ideas or expressing their intended messages. Words from English with an Indonesian counterpart and vice versa often raise the question of their meaning. The author proposed that their restricted vocabulary hindered the students' verbal proficiency. Numerous studies have shown that technology significantly impacts children's speaking abilities. An investigation was implemented to investigate the perspectives of English language students regarding implementing Yoodli AI in speaking classes (Pricilia & Rahmansyah 2024). According to the findings of their study, students in the Institute of Tapanuli Selatan who struggle with speaking are significantly helped by utilizing the Yoodli. In her study, Anggraini et al. (2023) also claimed that digital tools may help students speak more fluently; in this study, the author used the Yoodli English web. Research by Nunan (1997) claims that teaching pupils to use digital tools has improved their English language proficiency, which also supports this era. According to the relevant research that has already been cited, using technology to teach and acquire vocabulary can increase student interaction and engagement. In light of prior findings, the current study tried to look at how Yoodli is used to teach speaking skills to young learners. Yoodli is a user-friendly software with a ton of features like what Pricilia & Rahmansyah (2024) did for their study about Yoodli, including choosing the focus to be trained such as filler words, repetition, phrases used, weak words, eye contact, talk time, centring and so on. This focus is very influential in showing one's fluency, confidence and Ability in English. Furthermore, there is still little talk about research on the use IDEAS, Vol. 13, No. 1, June 2025 ISSN 2338-4778 (Print) ISSN 2548-4192 (Online) 2273 and gain of noodles to teach students speaking skills. The researcher chose the Yoodli web AI since an English teacher has struggled to teach a student with a kind of character to speak english. Additionally, every instructor and student mentioned the advantages of using AI in the classroom, demonstrating a favourable opinion of AI-based applications. However, university students during the 2018–2019 school year had a poor perception of AI-based applications (Keles & Aydin, 2021). It is believed that generational disparities and these varying perspectives impact the efficacy of AI-based solutions in education. Everyone has a unique perspective on the world, especially in unfamiliar surroundings. Therefore, it could be different. Additionally, it is handy to have, portable and has many functions explained in detail. All of these things may be taught via this program. The primary research question guides this inquiry: Does Yoodli usage significantly impact students' speaking test scores? Method This study employed a quasi-experimental methodology, focusing on the designs of non-equivalent control groups. The data set was obtained through a sampling of clusters. The experimental and control groups were the two classes used in this methodology. The audience under consideration consists of second- grade students from 14 classes at Public Islamic School for the 2024/2025 academic year. The author identified XI A1 as the experimental group and XI A2 as the control group. The class sizes were 36 and 32 students, respectively, yielding 68. The students were selected for participation because of their inadequate speaking skills. The experimental group utilized Yoodli, whereas the control group used a lecture-based approach. The tools utilized by the author in the research comprised an oral evaluation and a written test—the oral examination aimed to evaluate the students' speaking proficiency as a communication tool. Additionally, the author administered a quiz to assess participants' perceptions of Yoodli's efficacy as an instructional tool. The survey consisted of 20 questions. The ideas articulated in the essay were structured according to the Indonesian approach of integrating English to enhance students' understanding. The author assessed students' speaking abilities through a rubric that evaluates fluency, accuracy, and comprehensibility. All items were modified based on Widdowson's work (1990). The author assessed the outcomes of the essay through a written document and enquiries. The author employed SPSS 28.0 to compute the mean score, standard deviation, t-test, and oral assessment frequency and percentage rates. Ela Carmelia Mukti Sambas, Emeliyah Sukma Dara Exploring English Vocabulary Mastery in Student with ASD At Upper Primary School Level: Case Study of Multisensory Approach Implementation 2274 Scoring Classification of the Students’ Pre-Test and Post-Test in the Experimental Group Table 1 presents the pre-test results for the experimental group. Any student did not achieve excellent scores. 12.50% of students achieved a perfect score. 34.38% achieved good scores, 43.75% attained fairly good scores, while 9.38% received poor scores. None of the experimental group received a poor or abysmal score on the post-test; 31.25% received a perfect score, 53.12% a good score, and 15.62% a pretty good score. Table 1. Percentage and Frequency of Students’ Scores of Pre-Test and Post-Test of Speaking Ability Experimental Group Score Classification Pre-Test Percentage Post-Test Percentage 96 – 100 Excellent 0.00% 0.00% 86 – 95 Very Good 12.50% 31.25% 76 – 85 Good 34.38% 53.12% 66 – 75 Fairly Good 43.75% 15.62% 56 – 65 Fair 9.38% 0.00% 46 – 55 Poor 0.00% 0.00% Under 45 Very Poor 0.00% 0.00% TOTAL 32 100% 100% Table 2 indicates a significant improvement in the student's scores in each speaking component. Most students received the "inferior" category on the pre- test's fluency component, while they attained the "poor" classification in the accuracy and comprehensibility components. The experimental group's post-test results demonstrated that the students' comprehensibility was exceptional, while their fluency and accuracy were average following the treatment. IDEAS, Vol. 13, No. 1, June 2025 ISSN 2338-4778 (Print) ISSN 2548-4192 (Online) 2275 Table 2. The result of each Speaking Components Experimental Group F = Fluency A = Accuracy C = Comprehensibility Figure 1. Comparison Result of Experimental Group The Students’ Pre-Test and Post-Test in the Control Group Scoring Classification Table 3 indicates that no students achieved excellent scores in the control group. Specifically, 8.33% attained perfect scores, 52.78% received good scores, 36.11% earned fairly good scores, and 2.78% obtained fair scores. In the control group, the post-test results indicated that no students achieved excellent scores. Specifically, 19.44% attained perfect scores, 44.44% received good scores, 30.56% Criteria Pre-Test (F) Pre-Test (A) Pre-Test (C) Post-Test (F) Post-Test (A) Post-Test (C) Very Poor 9 5 - - - - Poor 6 8 10 5 4 3 Average 7 9 8 10 9 7 Good 5 6 8 7 8 9 Excellent 5 4 6 10 11 13 Total 32 32 32 32 32 32 Ela Carmelia Mukti Sambas, Emeliyah Sukma Dara Exploring English Vocabulary Mastery in Student with ASD At Upper Primary School Level: Case Study of Multisensory Approach Implementation 2276 earned fairly good scores, and 5.56% obtained fair scores. Notably, no students scored poor or very poor. Table 3. The result of each Speaking Components Controlled group The outcomes of the speaking element for the controlled group from the pre- test and post-test are shown in Table 4. Students' fluency was in the poor category in the pre-test. The student's accuracy fell into the "average" range, while their understanding was "good." Based on the students' post-test findings, Table 4 presents the results of the speaking component for the controlled group, comparing pre-test and post-test outcomes. Students demonstrated poor fluency in the pre-test assessment. The student's accuracy was "average," whereas their understanding was deemed "good." The post-test findings indicate that students achieved a "good" rating in fluency and accuracy, while their comprehensibility component received ratings ranging from "average" to "good" and "excellent." Table 4. The Speaking Result of Each Component Control Group Score Classification Pre-Test Percentage Post-Test Percentage 96 – 100 Excellent 0.00% 0.00% 86 – 95 Very Good 8.33% 19.44% 76 – 85 Good 52.78% 44.44% 66 – 75 Fairly Good 36.11% 30.56% 56 – 65 Fair 2.78% 5..56% 46 – 55 Poor 0.00% 0.00% Under 45 Very Poor 0.00% 0.00% TOTAL 36 100% 100% Criteria Pre-Test (F) Pre-Test (A) Pre-Test (C) Post-Test (F) Post-Test (A) Post-Test (C) Very Poor - - - 8 4 Poor 5 4 3 6 9 10 Average 10 9 7 7 8 6 Good 7 8 9 5 6 8 Excellent 14 15 17 6 5 8 Total 36 36 36 36 36 36 IDEAS, Vol. 13, No. 1, June 2025 ISSN 2338-4778 (Print) ISSN 2548-4192 (Online) 2277 Figure 2. Comparison Result of Control Group The Student's Comparison Between Pre-Test and Post-Test in Control Groups and Experimental The control group had an average score of 77.58 and a standard deviation 6.76, while the experimental group recorded an average score of 82.47 with a standard deviation 5.97. The author argued that students' speaking performance in both the experimental and control groups was comparable, as they were classified within the same category. The experimental and control groups showed different average scores after the treatments. The experimental group's average score (82.47) surpasses that of the control group (77.58), with the experimental group demonstrating a standard deviation of 5.97, in contrast to the control group's 6.76. After the intervention, the experimental group demonstrated significantly better outcomes than the control group. Figure 1. Comparison of Average score Ela Carmelia Mukti Sambas, Emeliyah Sukma Dara Exploring English Vocabulary Mastery in Student with ASD At Upper Primary School Level: Case Study of Multisensory Approach Implementation 2278 Paired Sample T-test Analysis 1. Pearson’s r = 0.306 → Shows a weak to moderate positive correlation between Yoodli application and examination results. 2. Sig. (2-tailed) = 0.011 → Since p < 0.05, the correlation is statistically significant, indicating that Yoodli usage is significantly associated with higher scores. The subsequent table presents the outcomes of the analysis: Table 5. The Analysis Correlation of Experimental & Control Group Independent Dependent **. Correlation is significant at the 0.01 level (2-tailed). The correlation coefficient between the English-speaking proficiency of second- year students at the public Islamic High School in Medan and their speaking ability is 0.830. The connection between the two variables is positive. According to the analysis by Aini et al. (2022) referenced in the preceding chapter, it can be concluded that this correlation value indicates a significant positive relationship. There was a significant Relation between the two variables, as indicated by the significance value of 0.011 (less than 0.05). Consequently, this research posits a significant disparity between the Experimental and Control groups of second- year students at the Public Islamic High School in Medan. Coefficient Determination The extent to which the independent variable (X) influenced or contributed to the dependent variable (Y) was assessed using the coefficient of determination test (R square). The researcher utilized SPSS Statistics 28 to compute the coefficient of determination test. The subsequent table presented value as a result: Independent Pearson Correlation 1 .306** Sig. (2-tailed) .011 N 68 68 Dependent Pearson Correlation .306** 1 Sig. (2-tailed) .011 N 68 68 IDEAS, Vol. 13, No. 1, June 2025 ISSN 2338-4778 (Print) ISSN 2548-4192 (Online) 2279 Table 6. Determination Coefficient Test Mod le R R Square Adjusted R Square Std. Error of the Estimate 1 .306a .936 .080 5.975 a. Predictors: (Constant), Speaking The coefficient determination Value (R square) is 0.0936, corresponding to 9.36%. In conclusion, the experiment utilizing noodles as the independent variable, while the control group did not employ noodles, demonstrates a 9.36% contribution to the speaking Ability of second-year students at Public Islamic High School in Medan. 90.64% of the contribution to students speaking capability is derived from additional factors. Result This research examined two variables: Students utilizing Yoodli serve as the independent variable (x), while students not utilizing Yoodli function as the dependent variable (y). The study displayed the information carried out at Medan's Public Islamic School. This section presents a comprehensive Results segment accompanied by a chart illustrating the principal findings from the study regarding Yoodli's influence on students' Ability to speak. This study investigated the impact of Yoodli AI as a resource for enhancing students' English-speaking abilities. The results derive from the pre-test and post-test scores of two groups: - Experimental Group (students using Yoodli) - Control Group (students using traditional methods) 1. Improvement in Speaking Performance The experimental group (Yoodli users) demonstrated a superior speaking skill enhancement compared to the control group. The pre-test and post-test scores indicated that. The number of students achieving scores in the "very good" and "good" categories rose following the utilization of Yoodli. The experimental group exhibited greater fluency, accuracy, and comprehensibility enhancements than the control group. Ela Carmelia Mukti Sambas, Emeliyah Sukma Dara Exploring English Vocabulary Mastery in Student with ASD At Upper Primary School Level: Case Study of Multisensory Approach Implementation 2280 2. Statistical Analysis of the Impact of Yoodli Statistical Measure Experimental Group Control Group Average Pre-Test Score 77.58 76.40 Average Post-Test Score 82.47 78.20 Standard Deviation 5.97 6.76 Pearson Correlation (r) 0.306 (Moderate Positive Correlation) - Significance (p-value) 0.011 (Statistically Significant, p < 0.05) - R-Square Value 0.0936 (Yoodli contributed 9.36% to improvement) - The p-value (0.011) indicates that the enhancement observed in the experimental group was statistically significant, confirming that Yoodli positively impacted students' speaking skills. 3. Performance by Speaking Component Each speaking component (fluency, accuracy, comprehensibility) showed improvement in both groups, but the experimental group had a greater increase. Speaking Component Pre-Test (Experimental) Post-Test (Experimental) Pre-Test (Control) Post-Test (Control) Fluency Poor/Average Good/Very Good Poor/Average Slight Improvement Accuracy Poor Good/Very Good Average Some Improvement Comprehensibility Fairly Good/Poor Excellent Fairly Good Some Improvement IDEAS, Vol. 13, No. 1, June 2025 ISSN 2338-4778 (Print) ISSN 2548-4192 (Online) 2281 4. Chart: Pre-Test and Post-Test Comparison Scores Figure5. The Chart Result Discussion This study's results address the research question regarding a typical investigation involving two groups: an experimental group of students utilizing Yoodli and a control group of students not utilizing Yoodli at Public Islamic School in Medan. In this study, the author examined student performance by administering a test to each variable and employing a quasi-experimental design with a two-group research framework to analyze the outcomes between both variables. Comparable outcomes are also demonstrated by several prior studies concerning students' speaking performance. The research corroborates earlier findings (Keles & Aydin, 2021) that AI-driven tools enhance language acquisition and elevate student motivation. Artificial intelligence in education has been shown to improve learning during emergencies (Masruddin & Nasriandi, 2022), rendering it an invaluable resource for remote and self-directed learning. Conversational AI tools mitigate foreign language anxiety (Ji et al., 2023) and provide tailored coaching (Altynbekova & Zhussupova, 2020). Additionally, examining English language students' perceptions regarding using Yoodli AI for English speaking (Aini et al., 2022). This research corroborated the findings of Adawiyah & Dalimunthe (2024), which indicated that students' speaking skills and motivation can be enhanced by using the TikTok application as a medium for developing English speaking proficiency. TikTok as a medium is more efficacious than traditional methods in teaching speaking, particularly in online classes. Hongsa et al. (2023) TikTok is an online social network with brief footage. It has been employed as an educational resource to assist students in enhancing their speaking and communication Ela Carmelia Mukti Sambas, Emeliyah Sukma Dara Exploring English Vocabulary Mastery in Student with ASD At Upper Primary School Level: Case Study of Multisensory Approach Implementation 2282 abilities. The TikTok application allows participants to produce and share various vertical short videos available to everyone at any time (Dewi & Sari, 2022). Simultaneously, TikTok learning presents a methodology that emphasizes student interaction and communication (Adawiyah & Dalimunthe, 2024). TikTok is regarded as a suitable Social Network for improving speaking Ability due to its acknowledged importance in this domain (Nasichah, 2023). The implementation of TikTok as a medium in the domains of teaching and learning speaking skills yields several advantageous outcomes. The TikTok application enhances students' creativity in responding to assignments by creating short videos, utilizing templates provided by the application, which students subsequently modify according to their inventive inclinations (Ocan a- Ferna ndez et al., 2019). Establishes an ideal environment for English speaking and actively engages students in project-based learning while facilitating contextual and significant learning. This also enhances their interest, inspiration, participation, and satisfaction. The results of this research illustrated the advantages of Yoodli learning, emphasizing its benefits for students. This research corroborated the findings of Harahap & Dalimunthe (2024), who discovered that students' speaking skills improved through an application designed to enhance these skills, namely SpeechAce. The interactive functionalities of SpeechAce, including speech detection and detailed suggestions, fostered enthusiastic student participation. Students' proficiency and confidence in spoken English were enhanced through continuous practice and constructive feedback. This research corroborates the findings of Permatasari & Lubis (2024), which utilized ELSA to augment speaking skills, while SpeechAce provided more comprehensive and immediate feedback, demonstrating efficacy in improving learners' speaking proficiency. Considering the previously mentioned findings, the author has determined that Yoodli significantly impacts students' speaking skills. However, numerous factors can influence students' proficiency in speaking English. Not only influenced by verbal proficiency but also impacted by challenges inherent in the English language, particularly in oral communication. Students' speaking scores can be affected by various factors, including inadequate self-confidence, unfamiliarity with topics, and insufficient knowledge, which may hinder their willingness to speak in class. Concerning the pronunciation of English words, they fear committing grammatical errors. Conclusion Because artificial intelligence and other forms of technology have vastly improved speaking exercises, lecturers will have the opportunity to question students about their benefits and drawbacks to assist students in enhancing their English language proficiency. The study concludes that Yoodli AI is a valuable tool for enhancing English fluency, accuracy, and comprehensibility. When incorporated IDEAS, Vol. 13, No. 1, June 2025 ISSN 2338-4778 (Print) ISSN 2548-4192 (Online) 2283 into educational settings, AI-based tools can improve language learning, as demonstrated by the experimental group's notable improvement over the control group (students using Yoodli). Similarly, in post-tests, students who used Yoodli did better, scoring higher on accuracy and fluency. According to statistical analysis, the improvement was significant and not coincidental (p-value = 0.011). Because Yoodli offered immediate feedback, students could self-correct their errors and gain confidence. Among the challenges were technical problems, a lack of experience with AI tools, and applying AI- generated corrections. Therefore, even though Yoodli improves speaking abilities, it should be used in conjunction with conventional teaching techniques for best results. Future research should expand the study's scope, explore long-term effects, and compare AI learning with traditional methods. Researchers can develop stronger, data-driven recommendations for using AI tools like Yoodli, ELSA speak, Duolingo or Google AI tools in English language education by focusing on larger samples, personalization, and teacher integration. References Adawiyah, R., & Dalimunthe, A. A. (2024). Implementing tiktok application as a medium for developing students' english speaking skills. Edukasi: Jurnal Pendidikan Dan Pengajaran, 11(01), 84–100. Aini, N., Amalia, F., & Ningrum, A. S. B. (2022). Improving students' speaking skills using the Hello English application as a home learning medium. Improving Students’ Speaking Skill Using Hello English Application as a Medium of Learning from Home, 10(1), 730–745. Altynbekova, G., & Zhussupova, R. (2020). Mobile application Fluentu for public speaking skills development. SHS Web of Conferences, 88, 2008. Dewi, P., & Sari, D. E. (2022). Perception of Digital Storytelling in Overcoming Fear for Speaking English through Interdisciplinary Project of Gender Issues. IDEAS: Journal on English Language Teaching and Learning, Linguistics and Literature, 10(2), 1635-a . Harahap, A. P., & Dalimunthe, A. A. (2024). Improving Junior High School Students' English-Speaking Skills Using the Speechace Application. IDEAS: Journal on English Language Teaching and Learning, Linguistics and Literature, 12(2). Haycraft, J. (1978). An introduction to English language teaching. (No Title). Hongsa, N., Wathawatthana, P., & Yonwilad, W. (2023). The Effects of TikTok Application on Improving EFL Students‟ English-Speaking Skills. World Journal of English Language, 13(7). Husnaini, H., Yahya, A., & Putri, N. I. W. (2023). The Efficacy of the Presentation, Practice, and Production (PPP) Method on the Speaking Skill of the English Learners Community (ELC) Students. EDULANGUE, 6(1), 45-61. Husnaini, H., & Sompa, S. R. (2025, June). Using Speechway Application to Increase Ela Carmelia Mukti Sambas, Emeliyah Sukma Dara Exploring English Vocabulary Mastery in Student with ASD At Upper Primary School Level: Case Study of Multisensory Approach Implementation 2284 Students'speaking Skills in English Education Department at Iain Palopo. In Proceedings of The Third International Conference on English Language Education. Ji, H., Han, I., & Ko, Y. (2023). A systematic review of conversational AI in language education focuses on collaboration with human teachers. Journal of Research on Technology in Education, 55(1), 48–63. Keles, P. U., & Aydin, S. (2021). University Students’ Perceptions about Artificial Intelligence. Shanlax International Journal of Education, 9, 212–220. Lado, R. (1961). Language Testing: The Construction and Use of Foreign Language Tests. A Teacher’s Book. Masruddin, M., & Nasriandi, N. (2022). Lexical and Syntactical Errors Performed by Junior High School Students in Writing Descriptive Text. IDEAS: Journal on English Language Teaching and Learning, Linguistics and Literature, 10(1), 1094–1100. Nasriandi, N., & Masruddin, M. (2021). The Use of British Parliamentary Debate Style in Teaching Speaking Skill. IDEAS: Journal on English Language Teaching and Learning, Linguistics and Literature, 9(1). Nasichah, A. H. (2023). Analyzing students' perception about using the TikTok application to enhance students' speaking skills at UIN Prof. KH Saifuddin Zuhri Purwokerto. ELITE JOURNAL, 5(1), 101–110. Nunan, D. (1997). Language Teaching and Research. Ocan a-Ferna ndez, Y., Valenzuela-Ferna ndez, L. A., & Garro-Aburto, L. L. (2019). Artificial Intelligence and Its Implications in Higher Education. Journal of Educational Psychology-Propositos y Representaciones, 7(2), 553–568. Permatasari, S., & Lubis, Y. (2024). ENHANCING PRONUNCIATION SKILLS IN EFL STUDENTS THROUGH THE ELSA SPEAK APPLICATION. Indonesian EFL Journal, 10(2), 205–212. Pricilia, G. M., & Rahmansyah, H. (2024). EXPLORING ENGLISH LANGUAGE STUDENTS’PERCEPTIONS ON THE USE OF YOODLI AI IN SPEAKING CLASS. UNNES-TEFLIN National Conference, 6, 410–415. Violita, V., & Cholsy, H. (2022). Speech Acts Equivalence of Audiovisual Translation on Enola Holmes Netflix Movie Subtitle. IDEAS: Journal on English Language Teaching and Learning, Linguistics and Literature, 10(1), 209–225. Widdowson, H. G. (1990). Aspects of language teaching. Oxford University Press.