INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)

ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XI November 2025

Page 2215

www.rsisinternational.org

Upper Primary Students’ Perceptions Towards the Use of Elsa Speak

in Improving English Pronunciation

*Grace Kho Hui Xuan, Melor Md Yunus, Hanita Hanim Ismail

Faculty of Education, Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor, Malaysia

*Corresponding Author

DOI:

https://dx.doi.org/10.47772/IJRISS.2025.91100176

Received: 19 November 2025; Accepted: 26 November 2025; Published: 04 December 2025

ABSTRACT

English pronunciation is a significant barrier for many Malaysian primary students, hindering their oral

participation and causing persistent low confidence and speaking anxiety. AI-powered tools, such as ELSA

Speak, offer personalised, real-time feedback that could address these limitations; however, relevant research

among young Malaysian learners is scarce. This mixed-methods study investigated the effectiveness of ELSA

Speak among 62 upper primary students (Year 5 and Year 6) in an SJKC school over four weeks. A questionnaire

measured students’ perceptions, confidence, motivation and speaking anxiety, while interviews with eight pupils

provided deeper insights. Quantitative analysis indicated generally positive perceptions and significant

improvements in students' reported pronunciation clarity, higher motivation, increased confidence and reduced

anxiety. Qualitative findings supported these outcomes, revealing that gamified features, colour-coded feedback

and private practice encouraged greater engagement and readiness to speak. Challenges identified included

speech-recognition inaccuracies, occasional technical issues and difficulty understanding linguistic terminology.

Overall, ELSA Speak demonstrates strong potential as a supplementary pronunciation tool, though

improvements in accuracy and child-friendly feedback are needed.

Keywords: Artificial Intelligence (AI), ELSA Speak, pronunciation training, Primary ESL Learners, perceptions

INTRODUCTION

Pronunciation is widely recognised as a critical component of English language proficiency, as it shapes learners’

intelligibility, communicative effectiveness and overall self-confidence (Derwing & Munro, 2015; Isaacs &

Trofimovich, 2017). In Malaysia, the importance of spoken English has intensified under the Common European

Framework of Reference (CEFR)-aligned curriculum, which foregrounds learners’ ability to communicate

confidently in real-world situations (Ministry of Education Malaysia, 2021). Despite these reforms, many

primary school learners continue to struggle with accurate pronunciation, resulting in hesitation during oral tasks

and restricted participation. These difficulties are commonly attributed to limited opportunities for individualised

practice, lack of corrective feedback, and the fear of making mistakes in front of peers (Ch’ng, Chan, & Saufi,

2025; Lau et al., 2025).

The emergence of Artificial Intelligence (AI)-enhanced language learning applications has created new

possibilities for addressing these longstanding issues. Among these tools, ELSA Speak has gained traction as a

popular mobile application that provides automated, instant feedback on pronunciation, using colour-coded

indicators and articulatory explanations to identify mispronounced sounds. By allowing learners to practise

independently at their own pace, AI pronunciation tools serve as valuable supplements to classroom instruction,

providing targeted and self-paced practice essential for improving pronunciation skills. Studies conducted

between 2024 and 2025 demonstrate that ELSA Speak significantly improves pronunciation accuracy, learner

confidence, and satisfaction across diverse contexts (Adawiah et al., 2024; Pham & Pham, 2025; Aulia & Santosa,

2025). These affordances are particularly relevant in the Malaysian primary context, where large class sizes and

limited instructional time pose challenges for teachers to provide targeted pronunciation support (Ag-Ahmad et

al., 2025; Mohammed et al., 2021).

INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)

ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XI November 2025

Page 2216

www.rsisinternational.org

Although studies have documented the benefits of AI-driven pronunciation tools among adult and tertiary

learners internationally, relatively little is known about their impact on younger ESL learners in Malaysia.

Primary pupils differ from older learners in terms of cognitive development, motivation and affective needs, and

their experiences with AI-based tools warrant dedicated investigation. Gaining insight into how AI tools

influence their pronunciation development, confidence, motivation and speaking anxiety is therefore crucial for

determining their suitability for integration into primary ESL classrooms.

To address this gap, the present study investigated the use of ELSA Speak among upper primary learners in an

SJKC school across four weeks. By employing a mixed-method design involving a student questionnaire and

semi-structured interviews, this research sought to provide a comprehensive understanding of pupils’ perceptions,

emotional responses and learning challenges when engaging with AI-driven pronunciation practice. The findings

contribute to the growing body of knowledge on AI-assisted language learning and offer insights into how

pronunciation training tools can be meaningfully integrated within the Malaysian primary English language

curriculum.

Research Problem and Its Importance

Despite sustained efforts to strengthen English language education in Malaysia, many primary school pupils

continue to struggle with pronunciation accuracy and oral fluency. These challenges are often reflected in

classroom observations, where pupils demonstrate hesitation when reading aloud, limited participation during

speaking tasks, and persistent reliance on teacher modelling. Such difficulties not only impede learners’ oral

intelligibility but also contribute to low self-confidence and heightened speaking anxiety—affective barriers that

significantly influence learners’ willingness to communicate. Current research indicates that Malaysian ESL

learners often struggle with speaking anxiety, inhibition and fear of negative evaluation, all of which adversely

affect their oral performance (Hashim, 2024; Aziz & Kashinathan, 2021).

A key issue is the limited access to personalised pronunciation practice in the primary school setting. Teachers

frequently struggle to provide personalised feedback due to time constraints, large class sizes and competing

curricular requirements. Prior studies highlight that large class sizes and curriculum pressures restrict teachers’

ability to provide targeted oral feedback, resulting in pupils receiving insufficient guidance on features such as

phoneme production, stress, and rhythm (Hazita, 2016; Mohammed, Raof, & Yusof, 2021). As a result, learners

have few opportunities to refine their pronunciation. Given these constraints, there is an increasing need to

explore alternative instructional tools that can complement classroom practices.

AI-driven applications, such as ELSA Speak, offer promising solutions by providing instant, precise, and

repeated feedback—key components for effective pronunciation improvement. However, despite the growing

use of such applications globally, there is insufficient evidence on how Malaysian primary learners perceive

these tools, whether they positively influence affective factors such as motivation and confidence, and what

challenges may hinder their effective use. This gap is especially critical given that many speech recognition

technologies are primarily designed for adult voices, which may pose additional challenges for younger users.

Hence, it is essential to investigate how upper primary students engage with AI pronunciation tools, how they

perceive their usefulness, and how these tools influence their emotional readiness to speak English. Insights into

these areas can inform pedagogical decisions, technology adoption and future curriculum planning within the

Malaysian ESL context. Addressing this research problem also aligns with broader national efforts to enhance

English language proficiency through technology-enhanced learning.

Research Objectives

The objectives of this research are:

1. To examine students’ perceptions of ELSA Speak as a pronunciation learning tool

2. To determine whether ELSA Speak improves students’ confidence, motivation and reduces anxiety in

speaking English

INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)

ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XI November 2025

Page 2217

www.rsisinternational.org

3. To find out the challenges and limitations experienced by the students when using ELSA Speak

LITERATURE REVIEW

Advancements and Impacts of AI on Pronunciation Training

Recent developments in AI have progressively reshaped pronunciation pedagogy by offering learners

personalised, real-time feedback that was once difficult to implement in traditional classroom settings. AI

systems equipped with automatic speech recognition (ASR) technology can analyse learners’ spoken output in

detail, identify errors and provide corrective guidance almost instantly, which is particularly beneficial for

younger learners who thrive on timely correction during oral practice (Prakash & Kausalya, 2025; Sungkur &

Shibdeen, 2025).

AI-based pronunciation training has been consistently shown to strengthen learners’ segmental accuracy, fluency

and intelligibility. Studies indicate that ASR-supported tools enable learners to repeatedly practise problematic

phonemes in an adaptive, learner-centred environment (Shak, Lee, & Loo, 2025). Moreover, research in ESL

contexts reveals that AI pronunciation tools enhance learners’ opportunities for speaking practice, particularly in

educational settings where teachers may lack sufficient time to provide one-on-one feedback (Thomas, 2025;

Vincent, Yunus, & Said, 2025). As such, AI functions as a complementary support system, enabling learners to

rehearse pronunciation independently and consistently.

Another major contribution of AI pronunciation tools lies in learner engagement. Many AI applications

incorporate gamified features, such as point systems, progress trackers and badges, which have been found to

significantly boost learner motivation and sustain interest in pronunciation tasks (Caiza, Villafuerte, & Guanuche,

2025). Moreover, the low-stress environment created by AI tools for oral practice reduces learners’ anxiety and

encourages a greater willingness to speak (Li, Mohamad, & You, 2025). Despite these advantages, challenges

remain, including accent misrecognition, inconsistent scoring and variations in feedback accuracy, particularly

when tools are used by younger learners with developing speech patterns (Lai & Chen, 2022). These limitations

underscore the need for further investigation into how primary pupils respond to AI pronunciation applications

in real classroom settings.

ELSA Speak as a Pronunciation Training Tool

ELSA Speak (English Language Speech Assistant) has emerged as a prominent AI tool designed specifically for

pronunciation training. The application employs advanced ASR and deep learning algorithms to analyse learners’

pronunciation at phoneme, word and sentence levels, while also assessing suprasegmental features such as stress

and intonation. Dhivya et al. (2023) note that the app’s colour-coded scoring system and step-by-step corrective

feedback allow learners to pinpoint articulatory issues more effectively than traditional drills.

ELSA Speak also supports learner autonomy. With intuitive features such as visual diagrams of mouth positions,

automatically generated lesson pathways and personalised progress tracking, the app encourages learners to take

charge of their own pronunciation practice (Dhivya, Hariharasudan, Ragmoun, & Alfalih, 2023). Learners can

monitor their progress, set goals and repeat exercises freely, which strengthens ownership of learning and

contributes to improved fluency. Research further underscores ELSA Speak’s alignment with Education 4.0

principles, highlighting its emphasis on learner-centred, technology‑enhanced, and self‑paced practice

environments (Dhivya et al., 2023).

However, most existing studies on ELSA Speak focus on adults or tertiary-level learners. For example, studies

with university students report notable gains in oral proficiency after sustained use of ELSA Speak but also

highlight mixed perceptions regarding feedback accuracy and phoneme scoring (Pham & Pham, 2025). Far fewer

studies have examined how younger learners interpret and benefit from the tool’s feedback systems. Primary-

aged pupils often require simpler instructions, clearer scaffolding and teacher mediation to comprehend

automated feedback (Kamaruddin, 2023; Zulkarnain, & Yunus, 2023). This gap in the literature suggests a need

to evaluate the suitability and effectiveness of ELSA Speak for younger ESL learners, particularly in the

Malaysian primary school context, where exposure to English varies widely.

INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)

ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XI November 2025

Page 2218

www.rsisinternational.org

Students’ Perceptions of AI Pronunciation Tools in English language learning

Learners’ perceptions play a central role in determining the success of AI-assisted pronunciation learning. When

learners feel supported, motivated or empowered by a tool, they tend to use it more actively; however, negative

perceptions may reduce engagement and limit learning gains (Amin, 2024). Research consistently indicates that

AI pronunciation tools can significantly enhance learners’ confidence, especially because they allow learners to

practise privately without the fear of being judged or criticised by peers (Sari & Kurniawan, 2025). Consistent

practice supported by clear, non-judgmental feedback contributes to improved self-belief in oral production tasks.

Motivation is another key factor influenced by AI-based tools. Studies show that AI applications with gamified

features, such as levels, achievements and interactive visual feedback, contribute to increased enjoyment and

commitment to practice (Yaccob et al., 2022; Liu, 2024). Learners are often encouraged by the incremental

progress and rewards provided by tools like ELSA Speak, which reinforce persistence and sustained effort. This

motivational impact is especially vital for primary school pupils, who often thrive in visually rich, reward-based

learning environments (Hung et al., 2018).

AI pronunciation tools have also been shown to reduce speaking anxiety. Speaking anxiety is a common obstacle

in ESL learning that often leads to avoidance of oral tasks. The private, self-paced nature of AI tools helps

learners practise without pressure, reducing the fear of being criticised for mispronunciations (Nadeem, Iftikhar,

& Rasheed, 2025). By offering a rehearsal platform before speaking publicly, these tools help learners feel more

prepared and less anxious during classroom speaking tasks. This reduction in anxiety is linked to increased

willingness to communicate, which is a key factor in successful oral language development.

However, despite these benefits, researchers also caution that technological challenges, such as inconsistent

speech recognition, misunderstanding of children’s voices and unclear scoring, can negatively affect learners’

confidence and motivation (Ngo, Chen, & Lai, 2024). For younger learners, misinterpreting feedback may lead

to confusion or frustration, particularly when scores appear inaccurate or seem difficult to improve. These

challenges suggest that teacher support and guided practice remain essential when integrating tools like ELSA

Speak into primary ESL classrooms.

METHODOLOGY

Research Design

This study adopted an explanatory sequential mixed-methods design, in which quantitative data were gathered

first, followed by qualitative data. The initial quantitative phase involved administering a questionnaire to

capture overall trends in pupils’ perceptions of ELSA Speak as a pronunciation training tool. The subsequent

qualitative phase consisted of semi-structured interviews, which were conducted to elaborate, clarify, and

provide deeper insight into the patterns identified in the survey responses. This design aligns with Creswell and

Plano Clark’s (2017) recommendation for studies seeking to explore the underlying explanations behind

quantitative patterns. Combining both statistical and narrative data enhanced the depth, validity and interpretive

strength of the findings.

Participants and Sampling

The participants in this study were 62 upper primary pupils from a Chinese national-type primary school (SJKC)

in Selangor, Malaysia. The sample comprised 31 Year 5 pupils and 31 Year 6 pupils, aged 11 and 12 years old.

All pupils were enrolled in English as a Second Language (ESL) classes and had learned English throughout

their primary schooling.

A convenience sampling method was used due to the school’s accessibility and willingness to participate. This

approach was appropriate given the structured school setting, the age of the pupils and the practical constraints

involved in conducting classroom-based research. All participants possessed basic digital literacy and had access

to a mobile device either personally or through shared family usage. Participation was voluntary, and written

informed consent was obtained from parents or guardians. To maintain confidentiality, pseudonyms were

INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)

ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XI November 2025

Page 2219

www.rsisinternational.org

assigned, and identifiable information was removed from all datasets.

Instruments

Two instruments were used in this study: a questionnaire and a semi-structured interview protocol. These

instruments were selected to capture both the quantitative patterns and qualitative depth necessary to understand

upper primary pupils’ experiences with ELSA Speak. The questionnaire was the main quantitative instrument

and consisted of 25 Likert-scale items designed to measure five constructs related to pupils’ use of the application,

namely perceived usefulness, pronunciation confidence, learning motivation, speaking anxiety and challenges

encountered during practice. All items were rated on a five-point scale ranging from Strongly Disagree to

Strongly Agree. The items were adapted from established instruments used in previous studies on AI-assisted

pronunciation learning, technology adoption and young learners’ speaking anxiety, but were modified to suit the

linguistic and developmental level of primary school pupils. To ensure clarity and content validity, the

questionnaire was reviewed by two TESL lecturers and an experienced ESL teacher who evaluated the suitability,

readability and alignment of each item with the target constructs. Their suggestions led to several refinements,

particularly in simplifying technical terms and improving the wording for young learners. In addition to expert

review, the questionnaire underwent a brief pilot check with a small group of pupils outside the main sample.

Feedback from the pilot confirmed that the instructions, item phrasing and response options were understandable

for upper primary pupils. The reliability of the questionnaire was subsequently tested using Cronbach’s alpha,

which produced a coefficient of 0.87, indicating strong internal consistency and confirming that the items were

appropriate for measuring the intended constructs.

The second instrument used in this study was a semi-structured interview protocol. The interview component

was designed to complement the questionnaire by exploring pupils’ perceptions, motivation, confidence and

speaking anxiety in greater depth. Eight pupils were purposively selected based on their questionnaire response

patterns to ensure variation in perspectives. The interview guide consisted of open-ended questions framed in

simple, age-appropriate language to help pupils express their thoughts comfortably. The semi-structured format

allowed the researcher to ask follow-up questions when necessary, enabling a more nuanced understanding of

pupils’ experiences with ELSA Speak. Interviews were conducted in a quiet classroom environment, and pupils

were given the option to respond in English, Mandarin or Bahasa Melayu to ensure full comprehension and

comfort. All interviews were audio-recorded with parental consent, transcribed verbatim and supplemented with

researcher field notes describing pupils’ non-verbal expressions and reactions. These qualitative data later played

an important role in enriching the interpretation of quantitative findings by highlighting pupils’ emotional

responses, preferred features of the application and the specific challenges they encountered during the

intervention period.

Data Collection and Data Analysis

Data collection took place over two sequential phases in accordance with the explanatory sequential mixed-

methods design. During the initial quantitative phase, pupils completed the questionnaire in their respective

classrooms under the supervision of the researcher and the English teacher. Clear verbal instructions were

provided, and where necessary, additional explanations were offered in Mandarin or Bahasa Melayu to ensure

comprehension without influencing responses. Pupils completed the questionnaire within approximately twenty

minutes, and all forms were collected immediately to maintain data accuracy and avoid missing responses.

Following the completion of the quantitative phase, the qualitative phase commenced with semi-structured

interviews involving eight purposively selected pupils. These pupils represented a diverse range of questionnaire

response patterns. Interviews were held in a quiet classroom space, audio-recorded with parental consent and

accompanied by researcher field notes that documented pupils’ non-verbal cues, hesitations, enthusiasm and

moments of uncertainty. Each interview lasted between ten and fifteen minutes and allowed pupils to express

their experiences and feelings regarding their use of ELSA Speak in greater depth.

Following data collection, the questionnaire responses were analysed using the Statistical Package for the Social

Sciences (SPSS) Version 27. Responses were coded numerically on a five-point Likert scale, enabling the

computation of descriptive statistics such as frequencies, percentages, means and standard deviations. These

INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)

ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XI November 2025

Page 2220

www.rsisinternational.org

descriptive measures provided a clear overview of pupils’ perceptions across the constructs of usefulness,

confidence, motivation, anxiety reduction and challenges. Reliability analysis using Cronbach’s alpha was

conducted within SPSS as part of the data screening process, confirming the strong internal consistency of the

questionnaire. Before analysis, the dataset was examined for missing values and inconsistencies; however, all

questionnaires were completed fully and no invalid entries were detected.

The qualitative interview data were analysed using Braun and Clarke’s (2006) six-phase framework for thematic

analysis. This process involved familiarising oneself with the transcripts, generating initial codes, identifying

and reviewing potential themes, refining theme definitions and integrating the findings into a coherent narrative.

Themes were generated inductively to allow pupils’ voices and personal experiences to emerge naturally from

the data. This analytic process provided richer contextual insights that explained, supported or expanded upon

the quantitative results. The integration of both datasets ultimately strengthened the interpretation of findings by

offering a more comprehensive understanding of pupils’ perceptions and experiences with ELSA Speak.

FINDINGS

This section presents the findings of the study, which investigated the perceptions of 62 upper primary pupils

(31 Year 5 and 31 Year 6) from a Malaysian SJKC towards the use of ELSA Speak for English pronunciation

learning. The results are organised into three main sections in accordance with the structure of the questionnaire:

(i) pupils’ perceptions of ELSA Speak as a pronunciation training tool, (ii) the influence of ELSA Speak on

pupils’ confidence, motivation, and speaking anxiety, and (iii) challenges and limitations encountered during the

use of the application. Qualitative findings from semi-structured interviews are also presented to complement

the quantitative data. The chapter concludes with a discussion integrating both sets of findings.

Survey

Pupils’ Perceptions of ELSA Speak

Table 1 presents pupils’ responses to Items 1 to 5, which examined their perceptions of ELSA Speak in terms of

usefulness, clarity of feedback, phoneme guidance, and overall contribution to pronunciation improvement.

Table 1 Pupils’ Perceptions of Using ELSA Speak

No.

Item

Mean

ELSA Speak helps me pronounce words correctly.

4.38

0.72

The feedback helps me improve my pronunciation.

4.32

0.74

I can learn correct mouth and tongue positions.

4.32

0.78

ELSA makes my speech clearer.

4.37

0.73

ELSA Speak is useful for learning pronunciation.

4.37

0.73

The results indicate overwhelmingly positive perceptions of ELSA Speak. More than 87% of pupils responded

with Agree or Strongly Agree across all perception items. The highest endorsement was for Item 1 (M = 4.38),

suggesting that pupils strongly believed the app helped them pronounce words correctly. Similarly, high means

for Items 2 to 5 (M = 4.32 to 4.37) reinforced the view that the app’s colour-coded feedback and articulatory

guidance were both accessible and useful. This aligns with ASR-supported pronunciation literature emphasising

the effectiveness of real-time phoneme-level feedback.

Influence on Confidence, Motivation, and Speaking Anxiety

Table 2 summarises Items 6 to 20, grouped under the constructs of pronunciation confidence, motivation, and

anxiety reduction.

INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)

ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XI November 2025

Page 2221

www.rsisinternational.org

Table 2 Influence of ELSA Speak on Confidence, Motivation, and Anxiety

Construct

Mean

Pronunciation Confidence (Items 6–10)

4.20

0.79

Learning Motivation (Items 11–15)

4.44

0.69

Speaking Anxiety Reduction (Items 16–20)

4.10

0.82

No.

Item

Mean

I feel more confident speaking English after using ELSA.

4.32

0.75

I am less afraid of making mistakes.

4.29

0.79

I am more willing to speak English in class.

4.26

0.81

I feel proud when I get a high score.

4.40

0.74

I can pronounce difficult words more confidently.

4.29

0.78

I enjoy using ELSA Speak.

4.43

0.72

I want to use ELSA even without teacher instruction.

4.35

0.78

I like the stars, coins, and scores.

4.45

0.71

ELSA makes me want to improve my pronunciation.

4.45

0.70

ELSA Speak is fun and interesting.

4.42

0.74

I feel less nervous speaking English.

4.23

0.85

I feel comfortable practising alone with ELSA.

4.23

0.82

I am less scared of being laughed at.

4.18

0.89

I prefer practising before speaking in class.

4.13

0.87

ELSA helps me overcome fear of speaking.

4.13

0.86

The construct with the highest mean score was motivation (M = 4.44, SD = 0.69). Items related to gamification

elements, such as enjoyment (Item 11), willingness to practise independently (Item 12), and appreciation of stars,

coins and scores (Item 13), recorded consistently high means (M = 4.35 to 4.45). These findings indicate that

the motivational design of ELSA Speak successfully sustained pupils’ interest and shaped positive learning

behaviours.

Pronunciation confidence was also recorded high endorsement (M = 4.20, SD = 0.79). Pupils reported feeling

more confident speaking English (Item 6: M = 4.32), less afraid of making mistakes (Item 7: M = 4.29) and more

willing to participate in class (Item 8: M = 4.26). The data suggest that regular practice with ELSA Speak helped

pupils rehearse pronunciation privately, increasing their readiness for classroom speaking tasks.

INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)

ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XI November 2025

Page 2222

www.rsisinternational.org

Speaking anxiety showed moderately high improvement (M = 4.10, SD = 0.82). Items such as “I feel less nervous

speaking English” (Item 16: M = 4.23) and “I am less scared of being laughed at” (Item 18: M = 4.18) reveal

that pupils experienced notable affective benefits. These findings imply that ELSA Speak functioned as a low-

pressure environment where pupils could practise without fear of judgment, improving their emotional comfort

during oral tasks.

Challenges and Limitations of Using ELSA Speak

Table 3 presents Items 21 to 25, which address potential difficulties encountered when using ELSA Speak.

Table 3 Challenges and Limitations of Using ELSA Speak

No.

Item

Mean

Sometimes ELSA Speak does not recognise my voice correctly.

3.50

1.03

I find some ELSA instructions or feedback difficult to

understand.

3.68

0.96

I feel frustrated when my score is low even when I try my best.

3.69

0.99

I need help to understand how to use ELSA Speak.

3.53

1.05

It is difficult to use ELSA when the Internet connection is slow.

3.74

0.99

The highest-rated challenge was related to internet connectivity (Item 25), with M = 3.74 (SD = 0.99), suggesting

that technical infrastructure played an important role in shaping pupils’ experiences. Difficulties understanding

feedback terminology (Item 22: M = 3.68) and frustration with low scores (Item 23: M = 3.69) also appeared

prominently. Challenges in voice recognition were noticeable but less severe (Item 21: M = 3.50), reflecting

occasional mismatches between pupils’ pronunciation and the app’s scoring algorithms.

These findings indicate that while ELSA Speak was generally well-received, certain technological and linguistic

barriers influenced pupils' overall satisfaction and ease of use. These challenges align with known limitations in

ASR-based tools, particularly when used with younger learners whose voices may not match adult-focused

training models.

Interview

The semi-structured interviews with eight upper primary pupils provided more in-depth insights into how ELSA

Speak supported their English pronunciation, influenced their confidence, motivation and speaking anxiety, and

what challenges they encountered during the four-week intervention. Thematic analysis yielded three

overarching themes: (1) pupils’ perceptions and experiences of using ELSA Speak, (2) the influence of ELSA

Speak on confidence, motivation and speaking anxiety, and (3) challenges and limitations of using the app.

Selected verbatim excerpts are presented to illustrate each theme.

Perceptions and Experiences with ELSA Speak

Across all eight participants, there was a consistently positive perception of ELSA Speak as a fun, game-like and

useful tool for pronunciation practice. Many contrasted the app with traditional classroom reading tasks,

suggesting that the interactive features made learning feel less like schoolwork and more like play.

For instance, one of the pupils described this contrast clearly, stating that:

“Sometimes English lessons can be a bit boring, but ELSA was like a game.” (P3)

INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)

ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XI November 2025

Page 2223

www.rsisinternational.org

Similarly, another pupil emphasised how the game-based design made pronunciation practice less intimidating

and more fun:

“It was fun because it’s not like normal English work. It’s more like playing a game but still learning.” (P5)

Beyond enjoyment, pupils valued ELSA Speak’s clarity in highlighting mispronounced sounds. Many

appreciated the colour-coded feedback and visual diagrams showing mouth and tongue positions. One of the

pupils described how the system guided self-correction:

“If I say wrongly, it underlines the part in red… red means very wrong and yellow means a bit wrong. So I try

again until it becomes green.” (P1)

Another pupil similarly valued the articulatory feedback:

“Before this, I didn’t know that for some sounds you need to put your tongue in a certain way. The picture helps

me understand better.” (P6)

Influence on Confidence, Motivation and Speaking Anxiety

A second major theme concerned how ELSA Speak influenced pupils’ confidence, motivation and speaking

anxiety in relation to English oral tasks. All eight participants indicated that their speaking confidence increased

over the four weeks. They attributed this to having opportunities to practise privately, receive feedback and

rehearse difficult words before reading in front of the class. One of the pupils described a noticeable shift in how

they read aloud in class:

“Last week teacher asked me to read one paragraph. Usually I read very softly, but this time I spoke louder.

Because I practised the difficult words the night before in ELSA.” (P3)

Another pupil reported a similar change in volume and self-assurance:

“Before this, I always read in a small voice. I scared people will hear my mistake. But now I speak louder.

Because I know I practised already.” (P4)

Motivation was another strong theme. Pupils discussed the satisfaction of earning streaks, stars and coins, which

encouraged them to practise regularly. One pupil described how gamification shaped their behaviour:

“The daily streak. I wanted to see how many days I can continue. And also the leaderboard. I see my friend’s

score and I want to beat it.” (P1)

Similarly, others described how friendly competition motivated more frequent use:

“The leaderboard! My friends and I compare scores. If they get a higher score, I want to beat them. It makes me

want to practise more.” (P7)

Beyond confidence and motivation, pupils consistently reported that ELSA helped reduce speaking anxiety.

Pupils expressed relief about being able to practise without fear of judgment. One pupil noted:

“When I practise with the app, I don’t feel shy because nobody is listening except the app. I can try many times.”

(P2)

Similarly, another pupil noted that the privacy of AI-mediated practice made them feel calmer:

“When I practise alone with ELSA, I don’t feel shy. I can try many times. Nobody is listening except the app.”

(P5)

INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)

ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XI November 2025

Page 2224

www.rsisinternational.org

Challenges and Limitations

Despite the positive experiences, pupils also faced several challenges while using ELSA Speak. One of the most

prominent concerns was inaccurate voice detection and inconsistent scoring, which at times reduced the

reliability of the feedback. Pupils expressed frustration when the app failed to recognise their pronunciations

correctly or assigned low scores despite repeated attempts. One remarked:

“Sometimes the app thinks I said it wrongly, even when I try very hard to say correctly… Sometimes I feel angry

because I already try so many times but still get red.” (P6)

Another pupil shared a similar sentiment regarding the inconsistency of the scoring system:

“Sometimes the mic doesn’t hear me properly. Even when I say correctly, it gives me red. Then I feel frustrated.”

(P4)

Another common limitation is related to technical issues, particularly slow internet and app freezing. One shared:

“Sometimes the internet is slow at home. When the Wi-Fi slow, the app loads very long. One time it suddenly

froze. I had to close and open again.” (P5)

In addition, many pupils found some of the feedback terms and instructions too difficult for their age. Words

such as intonation, stress, syllable and linking sounds were not easily understood without teacher or parent

support. As one explained:

“When I see words like ‘intonation’ or ‘stress’, I don’t understand. I needed to ask my father to explain.” (P2)

Another pupil also commented:

“Some instructions are difficult. Like ‘intonation’ or ‘stress patterns’. I don’t know what they mean. I had to ask

teacher to explain.” (P7)

A further challenge concerned sentence-level pronunciation tasks, where a single error could significantly lower

the overall score. One explained:

“The sentence ones. If I forget one word, it marks everything wrong. So I need to practise more times.” (P5)

Overall, the interview data indicate that while ELSA Speak was perceived as an enjoyable and beneficial tool

for pronunciation practice, refinements are needed in recognition accuracy, linguistic scaffolding and task

difficulty, especially when used with younger learners in primary school contexts.

DISCUSSION

Upper primary students’ perceptions of using ELSA Speak for English pronunciation practice

The quantitative findings from the perception items (Items 1–10) revealed that students generally held positive

perceptions towards ELSA Speak as a tool for pronunciation practice, with mean scores ranging from 4.20 to

4.40. These results indicate strong agreement that the application was enjoyable, user-friendly and beneficial for

improving pronunciation accuracy. Items related to perceived usefulness, such as “ELSA helps me speak more

clearly,” consistently received high ratings, suggesting that learners recognised the app’s capacity to support

accurate articulation and clearer speech.

The qualitative findings reinforce these findings. Interview participants consistently described ELSA Speak as

“fun,” “interesting,” and “like a game,” highlighting its gamified interface, such as stars, coins, streaks and

leaderboards, as core motivators. For example, P3 shared that “ELSA was like a game,” while P5 emphasised

that it “felt like playing but still learning.” These remarks resonate with the high mean values in the quantitative

data for items on engagement and enjoyment.

INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)

ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XI November 2025

Page 2225

www.rsisinternational.org

The clarity and usefulness of feedback also emerged as a strong theme in both data sets. Pupils valued colour-

coded scoring and articulatory guides, which helped them identify specific pronunciation errors. For instance,

P1 noted: “If I say wrongly, it underlines the part in red… so I try again until it becomes green.” This echoes the

high quantitative ratings for items related to correction and error identification.

Collectively, the quantitative and qualitative findings converge to indicate that pupils perceive ELSA Speak as

an effective, accessible and motivating tool for pronunciation practice. The combination of immediate feedback,

self-paced practice and interactive design appears to enhance their willingness to engage in pronunciation

activities, echoing findings from past research on AI-assisted oral practice.

ELSA Speak’s influence on students’ confidence, motivation and speaking anxiety

The quantitative findings for confidence, motivation and anxiety (Items 11–20) showed that students largely

agreed that ELSA Speak enhanced their confidence (M > 4.20), increased their motivation to practise English

pronunciation (M > 4.10), and reduced their speaking anxiety (M > 4.00). These positive outcomes were further

supported by qualitative interview findings, which provided rich explanations for the shifts.

Pupils frequently reported feeling “more ready,” “less afraid,” and “braver” after using ELSA Speak. Many

explained that the opportunity to practise pronunciation privately helped reduce the fear of making mistakes in

front of classmates. For example, P4 described a noticeable change in classroom reading: “Before this, I always

read in a small voice… but now I speak louder because I know I practised already.” Similarly, P3 recounted

speaking more confidently in class after rehearsing difficult words using ELSA. This aligns closely with the high

quantitative ratings for confidence-related items, indicating that repeated, individualised practice strengthened

pupils’ sense of preparedness and self-assurance.

Motivation also emerged as a strong theme across both data sets. Pupils described ELSA Speak’s gamified

features, such as stars, coins, streaks and leaderboards, as powerful motivators that encouraged them to practise

more frequently and persistently. Quantitative items related to enjoyment and willingness to practise achieved

some of the highest mean scores in the questionnaire. The interview excerpts illuminate this pattern. For instance,

P7 explained, “My friends and I compare scores. If they get higher, I want to beat them,” while P6 noted that

watching their score increase motivated continued practice. These comments reinforce how reward elements

played a significant role in sustaining interest, making pronunciation practice an enjoyable and self-driven

activity rather than a compulsory task.

Reductions in speaking anxiety were also evident. The quantitative results for anxiety-related items showed that

many pupils felt less nervous, less shy and more comfortable speaking English after using the app. The

qualitative data reveal that the private, non-judgmental nature of ELSA Speak was a key factor contributing to

this emotional improvement. P5 articulated this clearly: “When I practise alone with ELSA, I don’t feel shy. I

can try many times. Nobody is listening except the app.” Pupils also mentioned feeling calmer and more

emotionally steady after using the app, particularly before oral tasks. P3 described this shift: “Last time my heart

beat so fast when talking. But now I feel more steady because I practised already.”

Overall, the integration of quantitative and qualitative findings demonstrates that ELSA Speak plays a

meaningful role in enhancing upper primary students’ pronunciation confidence, sustaining their motivation and

reducing their speaking anxiety. The combination of instant feedback, gamified rewards and a private practice

environment creates favourable affective conditions that support oral language development. These findings

underscore the potential of AI-based tools to address both linguistic and emotional needs in primary ESL learning

contexts.

Challenges and limitations experienced by the students when using ELSA Speak

Although the overall findings revealed highly positive perceptions of ELSA Speak, both quantitative and

qualitative data indicated several challenges. These included inaccurate voice detection, inconsistent scoring,

technical disruptions, difficulty understanding linguistic feedback, and task demands that were sometimes

unsuitable for younger learners. While these limitations did not substantially diminish pupils’ overall enjoyment

INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)

ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XI November 2025

Page 2226

www.rsisinternational.org

or willingness to use the app, they highlight areas in which AI-based pronunciation tools require refinement

when used with younger learners.

One of the most frequently reported challenges involved the inaccuracy of ELSA Speak’s speech-recognition

system. This concern corresponded with comparatively lower SPSS mean scores for the related questionnaire

items and was further illuminated by the interview data. Several pupils reported frustration when the app marked

what they believed to be correct pronunciations as “wrong.” For example, P6 explained, “Sometimes the app

thinks I said it wrongly even when I try very hard… I feel angry because I already try so many times but still get

red.” P4 expressed a similar experience, noting that “Sometimes the mic doesn’t hear me properly. Even when I

say correctly, it gives me red.” These examples reflect known challenges in automatic speech-recognition

technology, which is often trained primarily on adult speech and may struggle with children’s developing

phonological patterns and pitch ranges.

Technical issues also emerged as a recurring limitation. Pupils described encountering slow internet connections,

long loading times and occasional app freezing, particularly when practising at home. These difficulties are

consistent with the higher SPSS mean score for the item related to connectivity problems. P3 shared, “When my

Wi-Fi slow, the app loads very long. One time the whole screen froze and I had to restart everything.” Such

interruptions disrupted the flow of practice and occasionally reduced learners’ motivation, suggesting that stable

digital infrastructure is essential for AI-mediated pronunciation activities to function effectively.

Another challenge related to pupils’ difficulty understanding some of the linguistic terminology used in ELSA

Speak’s feedback. Terms such as “intonation,” “stress patterns,” and “linking sounds” were unfamiliar to many

pupils and required teacher or parental explanation. P2 noted, “When I see words like ‘intonation’ or ‘stress’, I

don’t understand,” while P7 added that some feedback instructions were “too difficult.” This suggests that

although ELSA Speak offers detailed linguistic feedback, the complexity of its terminology may not be

developmentally appropriate for younger ESL learners, who may need simplified explanations or additional

scaffolding.

Sentence-level tasks posed another significant limitation. Pupils frequently explained that these tasks were

challenging because a single mispronunciation could result in a low overall score. P5 expressed this frustration,

stating, “If I forget one word, it marks everything wrong. So I need to practise more times.” This feedback

highlights how overly stringent scoring mechanisms may discourage young learners, particularly those who are

still developing foundational reading and pronunciation skills.

Overall, while these challenges did not deter pupils from using ELSA Speak, they did contribute to moments of

frustration and indicate important areas for improvement. Enhancing speech-recognition accuracy for children’s

voices, simplifying feedback terminology, adjusting scoring mechanisms for longer responses and ensuring

stable technological access may help improve the usability of ELSA Speak for younger learners. Addressing

these limitations is crucial for ensuring that AI-based pronunciation tools remain accessible, developmentally

appropriate and supportive of meaningful learning experiences in primary school contexts.

CONCLUSION

This study investigated upper primary students’ perceptions of ELSA Speak as an AI-driven pronunciation

training tool, focusing on its impact on their confidence, motivation, speaking anxiety and the challenges they

encountered during a four-week implementation period. Overall, the findings revealed that learners responded

positively to the app, perceiving it as enjoyable, accessible and effective for improving pronunciation. Features

such as instant colour-coded feedback, articulatory guides and gamified elements, including stars, streaks and

leaderboards, created an engaging learning environment that encouraged active participation and made

pronunciation practice less daunting for young ESL learners.

In addition to supporting linguistic development, ELSA Speak made a meaningful contribution to learners’

emotional readiness for speaking tasks. Many pupils reported feeling more courageous, better prepared and less

anxious about making mistakes after practising privately in a non-judgmental digital space. This improvement

in confidence and reduction in anxiety was consistently evident across both the quantitative scores and

INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)

ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XI November 2025

Page 2227

www.rsisinternational.org

qualitative remarks, suggesting that the app has the potential to strengthen not only pronunciation but also

learners’ self-belief and willingness to speak in class. These affective gains highlight the valuable role AI

pronunciation tools can play in creating a supportive and motivating environment for young ESL learners.

Despite these strengths, several limitations surfaced, particularly relating to technological constraints and the

developmental needs of younger users. Issues such as inaccurate speech recognition, inconsistent scoring, slow

internet connectivity and occasional app freezing disrupted learning and caused frustration. Some pupils

struggled to understand feedback terms like intonation or stress, highlighting the need for more age-appropriate

explanations and simplified scoring mechanisms.

In light of these challenges, several improvements are recommended. To facilitate better understanding of key

pronunciation concepts, learners would benefit from supplementary materials that simplify complex linguistic

terms, ideally incorporating visual representations, diagrams or child-friendly explanations. Furthermore,

ensuring stable internet access, updated devices or providing offline practice options would also help minimise

disruptions caused by freezing or delayed voice recognition. Finally, additional assessment opportunities beyond

the four weeks are encouraged to determine the long-term impact of ELSA Speak on sustained pronunciation

development and oral performance. While these refinements, tools like ELSA Speak hold strong potential for

effective integration into Malaysian primary ESL classrooms, supporting both linguistic growth and learners’

confidence in spoken English.

REFERENCES

1. Adawiah, R., Muliati, A., & Samtidar, S. (2024). The effect of ELSA Speak application on students’

English pronunciation development. Journal of Excellence in English Language Education, 3(1).

2. Ag-Ahmad, N., Syed Mohamed, A. T. F., & Majilang, D. B. (2025). Voices from ESL classrooms:

Overcoming challenges and enhancing English language teacher education in Malaysia. Studies in

English Language and Education, 12(1): 328-345 DOI:10.24815/siele.v12i1.39831

3. Amin, E. A.-R. (2024). EFL students’ perception of using AI text-to-speech apps in learning

pronunciation. Migration Letters, 21(3), 887–903.

4. Aulia, L. G. Y. M., & Santosa, M. H. (2025). AI in language education: Systematic review of ELSA

Speak’s impact on English. Jurnal Pendidikan Bahasa Inggris Indonesia, 13(1).

5. Aziz, A. A., & Kashinathan, S. (2021). ESL learners’ challenges in speaking English in Malaysian

classroom. International Journal of Academic Research in Progressive Education and Development,

10(2), 983–991. https://doi.org/10.6007/IJARPED/v10-i2/10355

6. Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in

Psychology, 3(2), 77–101. https://doi.org/10.1191/1478088706qp063oa

7. Caiza, G., Villafuerte, C., & Guanuche, A. (2025). Interactive application with virtual reality and artificial

intelligence for improving pronunciation in English learning. Applied Sciences, 15(17), 9270.

https://doi.org/10.3390/app15179270

8. Ch’ng, L. C., Chan, A. N., & Saufi, N. A. (2025). Exploring ESL public speaking anxiety among

undergraduates. Journal of Creative Practices in Language Learning and Teaching, 13(1), 146–169.

https://doi.org/10.24191/cplt.v13i1.4234

9. Creswell, J. W., & Plano Clark, V. L. (2017). Designing and conducting mixed methods research (3rd

ed.). Thousand Oaks, CA: SAGE Publications.

10. Derwing, T. M., & Munro, M. J. (2015). Pronunciation fundamentals: Evidence-based perspectives for

L2 teaching and research. John Benjamins.

11. Dhivya, D. S., Hariharasudan, A., Ragmoun, W., & Alfalih, A. A. (2023). ELSA as an Education 4.0 tool

for learning business English communication. Sustainability, 15(4), 3809.

https://doi.org/10.3390/su15043809

12. Hashim, H. (2024). Speaking anxiety among ESL learners in the Malaysian classroom. UiTM eBulletin

APB, 13, 22–28. Universiti Teknologi MARA.

13. Hazita, A. (2021). Implementation and challenges of English language education reform in Malaysian

primary schools. 3L The Southeast Asian Journal of English Language Studies, 22(3), 65-78.

DOI:10.17576/3L-2016-2203-05

INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)

ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XI November 2025

Page 2228

www.rsisinternational.org

14. Hung, H.‑T., Yang, J.‑C., Hwang, G.‑J., Chu, H.‑C., & Wang, C.‑C. (2018). A scoping review of research

on digital game‑based language learning. Computers & Education, 126, 1–13.

https://doi.org/10.1016/j.compedu.2018.06.001

15. Isaacs, T., & Trofimovich, P. (Eds.). (2017). Second Language Pronunciation Assessment:

Interdisciplinary Perspectives (Vol. 107). Multilingual Matters & Channel View Publications.

http://www.jstor.org/stable/10.21832/j.ctt1xp3wcc

16. Kholis, A. (2021). ELSA Speak app: Automatic speech recognition (ASR) for supplementing English

pronunciation skills. Pedagogy: Journal of English Language Teaching, 9(1), 1–14.

https://doi.org/10.32332/joelt.v9i1.2723

17. Lai, K. W. K., & Chen, H. J. H. (2022). An exploratory study on the accuracy of three speech recognition

software programs for young Taiwanese EFL learners. Interactive Learning Environments, 32(5), 1582–

1596. https://doi.org/10.1080/10494820.2022.2122511

18. Lau, K. Y. L., Adi Badiozaman, I. F., & Leong, H. J. Y. K. (2025). Predictors and moderators of ESL

classroom speaking anxiety: A structural equation modelling approach. Journal of Applied Research in

Higher Education. https://doi.org/10.1108/JARHE-05-2025-0380

19. Liu, L. (2024). Impact of AI gamification on EFL learning outcomes and nonlinear dynamic motivation:

Comparing adaptive learning paths, conversational agents, and storytelling. Education and Information

Technologies, 30, 11299–11338. https://doi.org/10.1007/s10639-024-13296-5

20. Li, W., Mohamad, M., & You, H. W. (2025). Exploring the effects of using automatic speech recognition

on EFL university students with high speaking anxiety. International Journal of Information and

Education Technology, 15(1), 187–194. https://doi.org/10.18178/ijiet.2025.15.1.2231

21. Ministry of Education Malaysia. (2021). English language education reform 2015–2025: Planning in

retrospect, implementation and prospective planning. English Language Teaching Centre Ministry of

Education Malaysia.

22. Mohammed, N. A., Raof, A. H. A., & Yusof, M. A. M. (2021). Challenges in implementing the CEFR in

English language classrooms. LSP International Journal, 8(2), 115–128.

https://doi.org/10.11113/lspi.v8.17977

23. Nadeem, S., Iftikhar, S., & Rasheed, A. (2025). Addressing the fear of speaking: The role of AI in

reducing speaking anxiety in ESL learners. Journal of Applied Linguistics and TESOL, 8(2), 825–833.

24. Ngo, T. T.-N., Chen, H. H.-J., & Lai, K. K.-W. (2024). The effectiveness of automatic speech recognition

in ESL/EFL pronunciation: A meta-analysis. ReCALL, 36(1), 4–21.

https://doi.org/10.1017/S0958344023000113

25. Pham, V. T. T., & Pham, A. T. (2025). English major students’ satisfaction with ELSA Speak in English

pronunciation courses. PLoS ONE 20(1): e0317378. https://doi.org/10.1371/journal.pone.0317378

26. Prakash, T., & Kausalya, S. (2025). Speech-to-text AI for improving English pronunciation in ESL

learners. International Journal of Innovative Science and Research Technology, 10(8), 1341–1349.

https://doi.org/10.38124/ijisrt/25aug1065

27. Vincent, J., Yunus, M. M., & Said, N. E. M. (2025). Using AI platforms to improve listening and speaking

skills in ESL primary students. International Journal of Academic Research in Progressive Education and

Development, 14(1), 1829–1843. http://dx.doi.org/10.6007/IJARPED/v14-i1/24848

28. Sari, F. K., & Kurniawan, P. Y. (2025). EFL students’ perceptions of AI speech recognition in

pronunciation practice: The case of Google Read Along. English Review: Journal of English Education,

13(1), 205–214. https://doi.org/10.25134/erjee.v13i1.10751

29. Sungkur, R. K., & Shibdeen, N. (2025). Pronunciation trainer for second language learning using

generative AI. International Journal of Educational Technology in Higher Education, 22(64).

https://doi.org/10.1186/s41239-025-00561-x

30. Yaccob, N. S., Rahman, S. F. A., Mohamad, S. N. A., Abdul Rahim, A. A., Rashid, K. K. A., Aldaba, A.

M. A., Yunus, M. M., & Hashim, H. (2022). Gamifying ESL classrooms through gamified teaching and

learning. Arab World English Journal (AWEJ) Special Issue on CALL, 8, 177–191.

https://doi.org/10.24093/awej/call8.12

31. Zulkarnain, N. S., & Yunus, M. M. (2023). Primary teachers’ perspectives on using artificial intelligence

technology in English as a second language teaching and learning: A systematic review. International

Journal of Academic Research in Progressive Education and Development, 12(2), 861–875.

https://doi.org/10.6007/IJARPED/v12-i2/17119