INTRODUCTION
Statement of the problem
In recent years, Vietnam has made significant investments in language education, particularly in enhancing the quality of English teaching and learning, as English is recognized as a global language (Vo et al., 2018).
Speaking is the most crucial skill to master when learning a language, as it serves as a key driver of social connection, status, career growth, and business success (Bygate, 1987, as cited in Baleghizadeh & Nasrollahi Shahri, 2014) Similarly, Ur (1996) emphasizes the importance of speaking in language acquisition.
Effective language proficiency is crucial for learners to demonstrate their knowledge, as speaking encompasses all forms of understanding Without the ability to articulate their expertise in their specialized fields, learners risk having their potential overlooked As a result, researchers in English language education are actively exploring diverse strategies to enhance both the effectiveness and motivation of language learners, particularly those who are not majoring in English.
As English continues to gain global popularity, many learners are pursuing language competence certificates to enhance their communication skills This trend has led to increased interest in internationally recognized exams like TOEIC, TOEFL, and IELTS, which are essential for meeting graduation and recruitment standards (Nguyen, 2015) Among these, the TOEIC four skills test—comprising TOEIC Listening & Reading and TOEIC Speaking & Writing—is becoming increasingly favored due to its lower examination costs and relevance to business language contexts.
With nearly all university students in developed countries owning mobile devices, particularly smartphones, mobile-assisted language learning (MALL) applications have emerged as valuable tools for enhancing classroom learning and self-study One notable example is Cake – Learn English for Free, a user-friendly app that has gained popularity on Android and iOS Leveraging YouTube videos featuring native speakers from reputable channels like BBC and TED Talks, Cake offers learners authentic resources for language practice anytime and anywhere The app also facilitates speaking practice through common communication patterns, utilizing artificial intelligence for voice recognition and providing immediate feedback These features are designed to motivate learners and are particularly beneficial for mastering the TOEIC Speaking test, which demands quick reflexes in a timed setting.
A learner-centered approach is essential for effective knowledge acquisition, particularly in English language learning Many people mistakenly think that obtaining English certificates from language centers is merely about memorizing information to achieve high scores.
To address concerns about test preparation, Pan (2010) advocates for a shift away from solely focusing on test-taking strategies Instead, educators should implement methods that empower students to excel independently on assessments Consequently, the researcher aims to improve outcomes by integrating TOEIC preparation instruction with the Cake app.
Research aim
In line with problems mentioned in the previous part, this study intends to investigate the impact of Cake app on enhancing TOEIC speaking competence for EFL
This study evaluates the TOEIC speaking results of three learners at a private English center in Ho Chi Minh City after completing a course that utilized the Cake app Additionally, it explores the learners' perceptions of the app in relation to improving their TOEIC speaking skills.
Research questions
To successfully reach for the research aim, the current study attempts to answer two research questions:
(1) To what extent does Cake app impact on enhancing TOEIC speaking competence of EFL learners at IMAP English center?
(2) How do EFL learners at IMAP English center perceive Cake app in their TOEIC speaking practice?
The significance of the research
The researcher has decided to implement the Cake app in the teaching and learning process of speaking skills to enhance and motivate students in the future The findings of this research may be advantageous to various stakeholders.
This application offers significant benefits to students by promoting active learning and effective speaking skills, particularly for TOEIC speaking tests Furthermore, it plays an increasingly important role in helping learners develop essential language competencies for their future endeavors.
This article aims to raise teachers' awareness of utilizing the app for enhancing their teaching practices It offers practical suggestions to create an optimal learning environment for students Importantly, the app encourages a shift from traditional, teacher-centered methods to more student-centered approaches by increasing student talk time in speaking classes, aligning with contemporary educational principles.
The researcher gains significant benefits from this study, enhancing her knowledge and skills in developing MALL applications The research findings will reveal important implications aimed at maximizing the efficiency of this educational tool.
LITERATURE REVIEW
Speaking competence in TOEIC speaking test
2.1.1 The concept of speaking skill
According to Bailey (2005), language learning encompasses four essential skills: listening, speaking, reading, and writing, which can be categorized as either productive or receptive Productive language involves the generation of speech or writing by learners, while receptive language pertains to the input received during reading or listening Additionally, the medium of communication, whether oral or written, plays a significant role in understanding speaking skills Consequently, speaking is defined as a productive oral skill that involves creating structured verbal utterances to effectively convey information.
Effective communication involves understanding the nuances of when, why, and how to engage in conversation, which requires advanced skills in producing and managing interactions, such as asking questions or taking turns (Schmitt, 2010) Speaking is often considered the most challenging of the four language skills due to two main factors identified by Nunan (2003): it occurs in real-time, leaving no opportunity for revision, and the immediacy of the interaction means that the listener is waiting for a response without delay.
According to Ur (1996), oral skills are essential for language mastery, as speaking encompasses all other forms of knowledge, making it crucial for learners to demonstrate their understanding of the language Brown (2000) emphasizes that speaking, alongside listening, serves as an effective tool for communication, enabling learners to quickly express ideas and comprehend others, which aligns with the primary goal of language teaching: developing communicative competence Richard (2008) notes that learners often assess their language learning success based on their perceived progress in speaking Furthermore, Leong and Ahmadi (2017) highlight that speaking integrates various language skills, allowing learners to utilize vocabulary, phrases, and grammar acquired from reading and listening, while simultaneously enhancing their reading abilities and writing output Additionally, practicing intonation and pronunciation helps learners adapt to different accents and variations in the language.
The article highlights the significance of speaking skills in language learning and outlines the challenges teachers encounter when organizing speaking activities to enhance these skills for their students.
2.1.2 The problems with speaking activities
According to Ur (1996), teachers often face four key challenges during speaking activities: inhibition, lack of ideas, uneven participation, and reliance on the mother tongue Brown (2000) highlights that inhibition stems from learners' anxiety about producing responses that may be perceived as "wrong, stupid, or incomprehensible" (p 269), leading to fears of making mistakes and losing face in front of peers.
Students often feel judged by their teachers and peers (Nguyen & Tran, 2015) Pham et al (2020) highlight that learners frequently struggle to express themselves due to a lack of topical expertise, a notion supported by Bailey (ed Nunan, 2003), who notes that students often repeat what the teacher says without developing their own language skills Additionally, the absence of motivation to become proficient speakers is evident, as students recognize that their English Language Teaching (ELT) curriculum fails to adequately assess their speaking competence (Pham et al., 2020) In large classes, this issue is exacerbated, with only one dominant student speaking at a time, leaving others with minimal opportunities to participate (Leong).
Harmer (1991, as cited in Vo et al., 2018) highlights a key challenge in English language learning, noting that learners who share the same native language often resort to using it during English classes This tendency occurs because it is easier for them to articulate their thoughts, particularly when engaging in discussions on unfamiliar topics.
In English classes in Vietnam, learners often resort to using Vietnamese during group discussions, leading to a situation where only a few confident students participate while others hesitate due to fear of making mistakes It is essential for both language learners and teachers to understand that effective speaking encompasses more than just grammar, pronunciation, or vocabulary; it involves multiple concurrent processes, including cognitive, physical, and sociocultural factors This dynamic interaction of a speaker's knowledge and skills, which must be activated rapidly in real time, is known as speaking competence.
8 competence is focused on developing holistically, the above problems will be expected to be somewhat alleviated (Dr Akhyak & Anik, 2013)
Speaking competence encompasses a broader scope than speaking performance, as learners' academic achievements are influenced by their beliefs about their capabilities (Meera & Jumana, 2015) Hardan (2013) emphasizes that performance is shaped by cognitive, motivational, and attitudinal traits that influence learners' behaviors This study defines performance as a collection of psychological and mental behaviors that students exhibit while completing academic tasks, ultimately leading to their final scores.
Speaking competence is a complex skill that involves multiple tasks to generate meaningful and appropriate statements (Johnson, 1996, as cited in Burns, 2012) According to Burns (2012), it encompasses knowledge of language and discourse, core speaking skills, and communication strategies Mastery of language includes grammar, vocabulary, pronunciation, and pragmatic connections Core speaking skills involve rapid speech processing to enhance fluency through pacing, chunking, pausing, and effective conversation management Additionally, communication strategies help learners address language gaps by utilizing paraphrasing, gestures, fillers, and seeking clarification.
Figure 2.1 Components of second language speaking competence
(Source: Goh & Burns, 2012, p 53, as cited in Burns, 2012, p 167)
Along with that mindset, Brown (2004) identifies 16 skills encompassing 11 micro- and 5 macro skills that a person with speaking competence must possess, as tabulated in Table 1 below:
Table 2.1 Micro- and macroskills of speaking competence
1 generate distinctions between English phonemes and allophonic variations
2 create language pieces of various lengths
3 produce stress patterns, stressed and unstressed syllables, rhythmic structure, and intonation contours in English
4 use reduced forms of words and phrases
5 utilize a suitable amount of lexical units to achieve pragmatic goals
6 produce fluent speech at a variety of delivery speeds
7 monitor one's own oral production and employ a variety of strategic tactics such as fillers, pauses, self-corrections to improve the message's clarification
8 use grammatical word classes (e.g nouns, verbs), and systems (e.g tenses, agreement), word order, rules, etc
9 produce speech in natural components, such as phrases, pause groups, breath groups, and sentence constituents
10 express a specific idea in many linguistic forms
11 employ cohesive devices in spoken discourse
12 perform communicative functions in a way that is appropriate for the situations, participants, and goals
13 employ appropriate styles, pragmatic rules and other sociolinguistic elements in face-to-face conversations
14 convey connections between internal and external relationships of an event, generalize and illustrate information
15 communicate with nonverbal indicators such as facial features and body language
16 develop and employ a variety of speaking techniques, such as highlighting keywords, offering context for interpreting word meanings, and measuring how well your interlocutor understands you
Speaking competence is a complex and multifaceted skill that demands learners to develop various aspects, including language knowledge, psychological readiness, and additional factors.
To master social influences, it is essential to understand the abstract nature of the speaking competence scale, which has led to the development of various foreign language proficiency tests, including standardized examinations like TOEIC, TOEFL, and IELTS These tests evaluate learners' English language proficiency levels (Vu et al., 2016) As noted by Brown (2004), the components of speaking competence are meticulously applied in the scoring criteria of these proficiency tests This article will specifically focus on the TOEIC Speaking test, which is the subject of the current research.
The TOEIC, or Test of English for International Communication, was established in 1979 by ETS at the request of the Japanese Ministry of Foreign Trade and Industry Its primary purpose is to assess the English proficiency levels of nonnative speakers in global contexts In recent years, the TOEIC test has gained significant attention in Vietnam's educational landscape, with many universities adopting it as part of the National Foreign Languages 2020 Project, as noted by Nguyen (2015) This initiative aims to enhance language programs and improve learning outcomes for students, as highlighted by Lai and Nguyen (2018).
Le et al (2020) highlight that specific universities in Vietnam, including Ho Chi Minh City University of Natural Resources and Environment and Tay Do University, are adopting standardized English proficiency assessments Tannenbaum et al (2015) further note that institutions like the National Economics University, Diplomatic Academy, Ho Chi Minh City University of Science, University of Economics and Law, and University of Social Sciences and Humanities are implementing TOEIC four skills as part of their admission and graduation criteria Additionally, Doan (2012) reports that over 300 small and medium enterprises, encompassing both private and multinational firms in sectors such as banking and tourism, are involved in this educational trend.
12 services, information technology, telecommunications, etc in Vietnam, require valid
TOEIC certificates as one of the initial conditions of recruitment
The TOEIC test primarily includes two types: TOEIC Listening and Reading, and TOEIC Speaking and Writing, which can be taken together or separately The TOEIC Speaking test, as outlined by ETS (2012), is a computer-delivered assessment designed to evaluate spoken English communication skills in everyday and business contexts Comprising five parts with eleven tasks, the test lasts approximately twenty minutes and accommodates a wide range of English proficiency levels, from highly skilled speakers to those with limited abilities Its design aims to provide insights into a speaker's language capabilities across different proficiency degrees.
(1) “The test taker can generate language intelligible to native and proficient nonnative English speakers
Mobile-assisted language learning (MALL)
2.2.1 Definition of Mobile learning (m-learning)
Mobile learning (m-learning) is often poorly defined, as various groups interpret it based on their unique objectives and experiences One key aspect of m-learning is learner mobility, allowing individuals to access educational resources in one location and apply them in another This flexibility enables learners to revisit and expand upon previously acquired knowledge, facilitating a broader understanding as they navigate through different topics rather than adhering to a singular path.
Mobile learning (m-learning) is defined as an educational approach that primarily utilizes handheld or portable devices, such as mobile phones, smartphones, personal digital assistants (PDAs), and tablet PCs, while excluding desktops and laptops that restrict mobility (Traxler, 2005; Traxler & Kukulska-Hulme, 2007) This form of learning facilitates education in on-the-go situations by leveraging the capabilities of portable, Internet-enabled devices (Crompton, 2013, as cited in Crompton & Burke, 2018).
In an effort to come up with the most comprehensive and relevant notion, Nguyen
M-learning, as highlighted in a 2016 study, represents a significant advancement in educational technologies, enabling learning to occur anytime and anywhere This is largely due to the widespread availability of handheld, palmtop, and wireless devices, which facilitate seamless access to educational resources.
2.2.2 Definition of Mobile-assisted language learning (MALL)
M-learning works well in settings where access to mobile technology is more common and embraced, or where lifestyles require more accessible training and learning options (Dudeney & Hockly, 2008) Moreover, language learning has become more popular as a result of new social challenges such as multicultural and multilingual communities due to the growth in human mobility all over the world, which leads to the demand of communicating appropriately and meaningfully (Kukulska-Hulme, 2010) Therefore, with the widespread use of mobile devices for training and learning, a new digital means (i.e., mobile phones) has merged m-learning with language learning; this approach is known as Mobile-Assisted Language Learning (MALL) (Çakmak, 2019)
According to Gilgen (2005, as cited in Kukulska-Hulme, 2010), providing students with tangible experiences enhances the excitement and relevance of language learning This insight led to the development of Mobile Assisted Language Learning (MALL) to support and enrich the language acquisition process.
As a result, MALL has a parallel development process and shares a lot of similarities with language learning This will be discussed in further depth in the following part
Historically, language teaching techniques focused on the analysis of language forms and structures, aiming for optimal instructional sequences (Schmitt, 2010) However, there has been a significant shift towards a learner-centered approach, which prioritizes the needs of learners and emphasizes the development of language skills, engagement in communicative processes, and the construction of language knowledge (Burns & Richards, 2012).
MALL has evolved from a focus on teacher-driven and formal activities to increasingly prioritize learner-driven and informal activities, as noted by Kukulska-Hulme (2010) This shift mirrors broader trends in language learning and technology integration, highlighting the importance of learner autonomy in mobile-assisted language learning.
In the realm of Mobile Assisted Language Learning (MALL), two significant trends have emerged: content-based and design-based approaches Content-based MALL research emphasizes the delivery of instructional materials through methods such as text messaging, quizzes, and surveys, aligning with traditional educational paradigms (Peterson & Divitini, 2004, as cited in Kukulska-Hulme & Shield, 2008) Conversely, the design-based approach focuses on empowering learners to take control of their education by creating fragmented activities, like short learning modules, that cater to their mobile learning needs (Trivanova & Ronchetti, 2003, p 1796, as cited in Kukulska-Hulme & Shield, 2008) This approach ensures that learners can access resources conveniently, enhancing their overall learning experience.
22 to authentic and immersive content that is updated on a regular basis (Fallahkhair et al.,
2007, p 312, as cited in Kukulska-Hulme & Shield, 2008)
Guo (2014) introduces the concept of MapALL (Mobile apps-Assisted Language Learning), highlighting how mobile applications facilitate individualized and learner-centered language learning by providing access to resources anytime and anywhere Almadhady et al (2021) further emphasize that modern mobile applications for language learning, particularly in English, enhance personal and autonomous learning environments, offering flexible and dependable access to educational materials.
Learner-centered approach in MALL
Mobile-Assisted Language Learning (MALL) emphasizes the importance of empowering learners to take control of their education, as mobile devices provide constant access and flexibility for learning This learner-centered approach, highlighted by Nguyen (2016), contrasts with traditional classrooms where students are passive recipients of knowledge Instead, learners are encouraged to actively master their own understanding and share it with peers and teachers, enhancing their knowledge through collaboration When effectively implemented in a MALL environment, this approach fosters behaviors aligned with modern language teaching and learning trends.
Figure 2.2 Learners’ behaviors in a MALL environment
Figure 2.2 illustrates the three key interactions in self-directed learning: learners engaging with themselves, their mobile devices, and their social connections Nguyen (2016) highlights the significance of learners identifying their personal learning goals and setting a pace that aligns with their abilities Similarly, Klimova (2019) provides examples, noting that students in Tourism Management prefer to learn language skills relevant to their field from the outset of their English courses, while upper-intermediate learners aim to avoid basic language patterns In essence, it is crucial for learners to understand what and how to learn in a way that is most effective for them.
After determining their needs, learners choose what they want to learn “based on a wide selection of course materials” with their own familiar devices (Nguyen, 2016, p
According to Walsh (2019), learners respond favorably to Mobile-Assisted Language Learning (MALL) because it enhances the portability and usability of learning materials, allowing for effective learning both inside and outside the classroom, thanks to the widespread availability of mobile devices.
To enhance self-directed learning, learners must engage in social interactions through mobile devices, allowing them to connect with peers and instructors Research by Song and Fox (2008, as cited in Kukulska-Hulme, 2010) highlights that undergraduate English learners actively use mobile technology to support their learning by collaborating with fellow students to address their vocabulary needs Moreover, instructors play a crucial role by guiding students in selecting appropriate materials and knowledge acquisition, transitioning from traditional knowledge presenters to facilitators of discussion and opinion (Nguyen, 2016, p 18).
According to Koole (2009), effective pedagogical strategies involve selecting and organizing learning activities based on real-life situations, encouraging learners to explore their physical and cultural environments Teachers play a crucial role in identifying tools that meet these needs, ensuring that learning activities are grounded in authentic experiences while promoting self-discovery among learners This study aims to enhance the learning process for the TOEIC Speaking test by integrating a mobile application called Cake, fostering a more positive teaching-to-the-test environment through a learner-centered approach.
Introduction of Cake – Learn English for Free
Cake, short for "Learn English for Free," is a new language-learning app designed for Android (minimum version 4.4) and iOS (minimum version 10.0) This innovative application utilizes YouTube-based videos to help learners discover authentic English expressions curated from popular channels such as BBC, TED Talks, The Ellen Show, and MasterChef By subscribing to these channels, users can enhance their vocabulary, learn various expression styles, and gain insights into cultural aspects related to their interests Cake offers free versions in 18 languages, including Vietnamese, making it user-friendly and accessible for Vietnamese learners.
Video lectures are updated daily, featuring concise videos that allow learners to grasp the context of expressions, which are repeated three times for better retention Afterward, target expressions are isolated for speaking practice Additionally, learners can explore similar phrases in other videos by watching a brief advertisement The app also provides conversation simulations with native speakers, allowing users to practice their speaking skills in a role-play format Following their practice, learners can assess their pronunciation using AI speech recognition, receiving immediate feedback rated from A to D, enabling them to identify and correct mistakes effectively.
Users can customize their learning experience by adjusting video speeds from 0.75x to 1x, allowing them to keep pace with the content The platform offers various levels, including introductory, elementary, intermediate, and advanced, to cater to individual learner needs After completing activities, learners earn stars for motivation and can compete on a global leaderboard, adding a gamified element to their education Additionally, the application tracks learning history, sending daily notifications to help users review previous lessons regularly.
Learner-centered approach and Cake app
As highlighted in II.4, some similarities between the Cake application and the learner-centered approach by Nguyen (2016) in II.3 were revealed
Table 2.4 Similarities between Cake app and learner-centered approach
Learners’ behaviors in a MALL environment
Learners’ behaviors in using Cake app
Intrapersonal interaction Subscribing favorite channels
Choosing interesting content Learning authentic materials anytime and anywhere
Device interaction Adjusting speed of videos
To enhance language learning, it's essential to select appropriate difficulty levels, utilize AI voice recognition for constructive feedback, and set up notifications for lesson reviews Engaging in social interactions and practicing conversations in a virtual environment can significantly improve fluency Additionally, watching videos of native speakers in various contexts provides valuable insights into real-life language use.
The Cake app aligns well with the learner-centered approach in English teaching and learning, as shown in Table 2.4 However, enhancing social interaction in real-life situations would significantly improve its effectiveness, relying on the coordination and organization of teachers.
Previous related studies and research gap
Cake is a smartphone application that has gained popularity over the past three years, particularly among students at Anbar University in Iraq, who appreciate its ability to facilitate English language learning anytime and anywhere for free Research by Almadhady et al (2021) and Fitria et al (2021) highlights that users experience fewer errors and engage in longer conversations, leading to improved fluency and confidence in speaking English publicly Luu et al (2021) note that Cake, along with other English learning apps, enhances speaking competence and revitalizes traditional English courses Anis (2020) praises the app for enabling students in Indonesia to practice English through movies and videos, while also making learning more efficient and enjoyable Additionally, Yanthi (2020) points out that Cake provides immediate feedback, encouraging active participation without the fear of making mistakes However, most existing studies rely on surveys and interviews to assess the app's effectiveness, with a notable lack of research linking Cake to proficiency tests like the TOEIC Speaking test.
IMAP Vietnam, an English center system where the researcher is working, is currently well known for offering TOEIC preparation courses with the brand Anh Ngu
At the English center, Ms Hoa oversees a structured curriculum for TOEIC® Listening, Reading, Speaking, and Writing, primarily catering to non-English major students and workers who need to establish a basic foundation in English Due to the necessity of teaching in Vietnamese, students have limited exposure to English, which contributes to their anxiety during speaking activities and TOEIC speaking tests To address this issue, the center has introduced Cake – Learn English for Free, providing students with additional English exposure and immersive listening and speaking materials that enhance their understanding and adaptability for the TOEIC speaking test.
This study aims to address the research gap regarding the effectiveness of the Cake mobile application in enhancing TOEIC speaking skills, as there are limited studies on this topic With the increasing number of TOEIC speaking preparation courses anticipated in the future, understanding the impact of the Cake app on learners' speaking abilities is crucial.
METHODOLOGY
Research design
A quasi-experimental design was employed to assess how Cake enhances TOEIC speaking competence among learners at an IMAP English center As defined by Creswell (2012), this design involves two groups—a treatment group and a comparison group—to establish potential cause-and-effect relationships between variables Fraenkel and Wallen (2009) further emphasize that this approach allows researchers to evaluate the impact of the treatment The study specifically aimed to identify the advantages of using Cake for improving speaking skills for the TOEIC test The choice of a quasi-experimental design was influenced by the inability to artificially create groups, as participant availability dictated the structure of the classes, which are organized by the academic department based on the Academic Officer's decisions.
Research site and participants
This study took place at an IMAP English center branch in Thu Duc City, Ho Chi Minh City, involving 32 mixed-aged learners from various academic backgrounds and both genders The participants exhibited diverse educational experiences.
All participants, having studied English for several years at various schools, successfully passed the final test for the PRE level at IMAP English Center and registered for level A+ This indicates they achieved a score of approximately 350/990 on the TOEIC Listening and Reading sections Eager to enhance their skills, these learners also enrolled in additional TOEIC Speaking and Writing preparation courses In summary, A+ level students are engaged in a comprehensive TOEIC preparation program aimed at achieving higher band scores, targeting 450-500/900 for Reading and Listening, and 70-110/200 for Speaking and Writing.
Figure 3.1 Two alternatives after finishing level PRE at IMAP English center system
Instruments
3.3.1 Pre-test and post-test: TOEIC Speaking test
To gather quantitative data, TOEIC Speaking pre-tests and post-tests were conducted with two groups Creswell (2012) emphasizes the importance of using pre-tests to confirm that participants share similar characteristics prior to intervention, while post-tests are essential for evaluating outcomes after the intervention has been implemented.
33 interventions to determine whether the suggested treatment is effective with the experimental group or not
The TOEIC Speaking test consists of five parts and eleven questions, lasting around 20 minutes In part 1, test takers read two texts aloud, while part 2 involves describing two pictures Part 3 requires responses to three short questions on a specific topic, and part 4 asks for answers to three additional questions based on provided information Finally, in part 5, candidates express their opinions on a particular topic All questions must be completed within the allotted time and are rated accordingly.
The TOEIC Speaking test was complemented by a questionnaire developed to enrich the quantitative data collected in this study According to Wilkinson and Birmingham (2003), a well-designed questionnaire can provide valuable insights that are easy to analyze, making it a less resource-intensive research tool for gathering opinions from a large audience The questionnaire utilized a five-point Likert scale, ranging from strongly disagree to strongly agree, allowing for convenient analysis using SPSS version 26 Delivered in paper format to participants in the experimental group during the final course session, this questionnaire closely followed the framework established by Metruk (2021) It comprised two main sections: the first focused on the aspects of the Cake app that enhance TOEIC speaking competence, while the second explored learners' perceptions of the app's impact on their preparation for the TOEIC Speaking test, grounded in the learner-centered approach discussed in section II.5.
In addition to the quantitative data collection methods mentioned, a semi-structured interview featuring five open-ended questions was conducted to gather qualitative insights According to Wilkinson and Birmingham (2003), interviews provide a deeper understanding of the significance behind the data, unlike other instruments that focus solely on surface elements Furthermore, Dürnyei (2007) emphasizes that semi-structured interviews allow researchers to prepare guiding questions while also enabling participants to elaborate on their responses.
These 5 interview questions of this study inquired about (1) progress comparison in TOEIC Speaking competence before and after using Cake app, (2) benefits and (3) difficulties when using this app for learning TOEIC Speaking, (4) how this app affect their attitudes and motivation towards learning English speaking in general as well as TOEIC speaking in particular, and (5) suggestions for using this app during TOEIC speaking test preparation courses.
Procedures
The researcher obtained permission from the Academic Officer to conduct four A+ classes across two campuses in Thu Duc City, with sessions scheduled for both first and second shifts starting in February 2021 Following this, the researcher collaborated with another English teacher to evaluate the pre-test and post-test results Two of the four classes were randomly selected as the experimental group, consisting of 15 and 17 students, while the control group comprised 16 students in each class.
35 test occured in class on the second session of both groups, after the first one for introduction and reminder
The A+ course spanned 31 sessions over 2.5 months, with each session lasting 1.5 hours It included a mid-term test, a final-term test, a final-term test correction, and 28 lessons The course followed a consistent schedule throughout its duration.
Table 3.1 Fixed schedule of the course employed in the whole system
Session 31 Final-term Test Correction
(*) Tests from the academic department
Modified schedule of the course for research purposes will be presented as follow:
Table 3.2 Modified schedule of the course for research purposes
Session 1 Introduction of the course + Lesson 1 +
Reminders for the session 2’s test
Session 2 TOEIC Speaking pre-test
Session 32 TOEIC Speaking post-test
Session 33 Final-term Test Correction +
(*) Tests from the academic department
The research utilized pre-test and post-test materials sourced from "Gwen Tv," a reputable YouTube channel recognized for its authentic TOEIC Speaking test simulations The selected questions included "Read a text aloud" for questions 1 and 2, and "Respond to questions" for questions 5, 6, and 7, effectively mirroring the real TOEIC Speaking test format provided by ETS.
To achieve the highest possible score of 26, candidates must express their opinions within a structured time frame that aligns with typical classroom settings, ensuring full engagement from all students The test focuses on general communication skills rather than specialized TOEIC knowledge During the assessment, test videos were shown on a laptop to maintain proper timing, with two raters scoring independently without disclosing their findings until the test concluded.
Following the pre-test, the Cake app was introduced to enhance TOEIC speaking skills in two experimental classes The researcher guided students in installing the app and demonstrated its features, motivating them to learn a minimum of one sentence daily To ensure accountability, students were required to share screenshots of their progress in a dedicated Zalo group created for daily teacher-student interaction.
Figure 3.2 A screenshot from learners sent to Zalo group
In the initial 30 minutes of each lesson from session 3, the class focused on practicing sentence patterns shared on the Zalo group, which were further detailed in a specific lesson plan.
The first 30 minutes of each lesson
1 T divides Ss into groups of 4
2 T asks Ss to show other members the sentence they learned on Cake, explaining the meaning and context of the sentence
3 All of members in group have to know 4 sentences of the day clearly
4 T asks Ss to create a conversation (or more) with the topic related to TOEIC Speaking
(meetings, schedules, work procedures, reservations, negotiations, etc.) using these 4 sentences of the group
5 T appoints random Ss to present created conversation(s) in front of the class
6 At the same time with (5), T appoints another member to type these sentences on
Word and show on TV, then T saves and sends the file to class every day
7 T exchanges members of groups every lesson
Sharing new sentences to each other
Using these sentences in context related to TOEIC topics
Avoiding dominant Ss (every learner has to be responsible for the overall results of the group)
Creating new groups (their subscribed channels may vary)
Figure 3.3 An example for communication activities using Cake in class
From four random sentences taken from Cake above, students have to think of a dialogue related to TOEIC Speaking The following example simulates the questions 8, 9, 10
"Respond to questions by using provided information":
During a phone call, a customer inquires about the museum's themes for the day and the ticket pricing from the ticket salesperson The salesperson provides information on the current exhibits and their corresponding themes, along with the cost of admission This exchange helps the customer make an informed decision before purchasing a ticket.
S: Hello How can I help you?
C: Hi I would like to buy a ticket for a visit to your museum today I just come to this city for a conference and it is more than one hour until starting time Any recommendations?
S: Could you hold the line, please?
S: …We are going to have a stone collection from a famous archeologist in 10 minutes Are you interested?
C: Hmmm… In all honesty, I don’t like it That sounds quite boring for me S: Well, the tickets are 10% off; moreover, the archeologist himself will lead you to each stone sample and tell you about the story behind it Why not give it a try?
For the comparison group with the other two classes, the researcher provided an ebook called Everyday Conversations: Learning American English (ISBN 978-1-625-
To ensure equal treatment of the two groups, audio files were provided as supplementary material Additionally, the researcher dedicated the first 30 minutes of each lesson from session 3 to dividing the groups for an in-depth exploration of the conversations presented in the book.
Figure 3.4 Contents of the book Everyday Conversations: Learning American
Figure 3.5 An example for communication activities using the extra material in control group
In the final extra session of the course, the researcher administered a TOEIC Speaking test to both groups, following the same procedures as the pre-test Two raters evaluated the post-test scores, allowing the researcher to compare results and assess the groups' progress Additionally, a questionnaire was distributed to all participants in the experimental group during the final-term test correction session After gathering the responses, five random learners were chosen for a 20-minute face-to-face interview.
Data collection
Quantitative data were collected and analyzed using IBM SPSS Statistics 26.0 An independent samples t-test was performed to determine if there was a statistically significant difference between the pre-test and post-test results.
The study involved two groups evaluated before and after the treatment, utilizing Cronbach’s Alpha test to assess the reliability of the questionnaire by examining the relationship between its items and the main construct Additionally, data gathered from semi-structured interviews conducted in Vietnamese were translated into English, coded, and categorized into specific themes for analysis.
Reliability and validity
Reliability, as defined by Fraenkel and Wallen (2009) and Creswell (2012), refers to the consistency of test results over time when participants respond in the same manner Validity, on the other hand, involves creating an instrument that accurately measures what it is intended to assess This study employs the standardized TOEIC Speaking test to evaluate treatment effectiveness, with results assessed by two raters to ensure inter-rater reliability, a process that involves multiple evaluators to enhance objectivity The TOEIC scoring is based on an established marking scheme from ETS, minimizing subjective bias Additionally, the questionnaire was translated into Vietnamese to prevent misunderstandings and piloted with a small group to refine any irrelevant items The study also utilized triangulation, which combines quantitative data from pre-tests, post-tests, and questionnaires with qualitative data from semi-structured interviews, as suggested by Creswell (2012) This approach enhances the research by leveraging the strengths of various instruments while addressing their weaknesses.
FINDINGS AND DISCUSSION
Data analysis
The pretests of both CG and EG were marked by two teachers (rater 1 and rater
2) to ensure the inter-rater reliability of the pretest scores for comparison The correlation of pretest scores by the two raters were examined by Pearson Correlation Coefficient in SPSS The results of this correlation examination would be presented in the following tables: Table 4.1 showed the correlation of the pretest scores by rater 1 and rater 2 in CG and Table 4.2 provided the correlation of the pretest scores by rater 1 and rater 2 in EG
Table 4.1 Correlation of pretest scores of the CG by two raters
Table 4.2 Correlation of pretest scores of the EG by two raters
As can be indicated from the data in Table 4.1, Sig (2-tailed) = 000 < 05, this means there was a statistically significant correlation between scores of pretest by rater
The Pearson Correlation Coefficient indicated a strong correlation between rater 1 and rater 2 for CG pretest scores (r = 0.968) Similarly, a robust correlation was found for EG pretest scores, with a Pearson Correlation Coefficient of r = 0.974 and a significance level of Sig (2-tailed) = 000, which is less than 05 These findings establish and confirm the inter-rater reliability of pretests for both groups Consequently, the scores from rater 1 were selected for further analysis.
The pre-test scores for both the Control Group (CG) and the Experimental Group (EG) were analyzed using a Normal Q-Q Plot to assess the normality of the underlying population distribution As shown in Figure 4.1, the scores from both groups aligned closely along a straight line, suggesting that the distribution of scores was normal.
Figure 4.1 Normal Q-Q plots for TOEIC speaking pretest results
With the reliability of the speaking pretest scores validated, the mean scores of
The pretest scores for the Control Group (CG) and Experimental Group (EG) were analyzed, revealing that the CG had a mean score of 9.41 (SD=5.01, N=2) while the EG had a mean score of 9.34 (SD=4.84, N=2), indicating only a slight difference between the two groups An independent samples T-test was conducted to assess the significance of this difference Levene’s test showed a significance value of 498, confirming equal variances, and the T-test yielded a 2-tailed significance value of 960, which is greater than 05 Consequently, it can be concluded that there is no statistically significant difference in speaking performance between the CG and EG prior to the intervention, suggesting their performances were comparable.
Table 4.3 Descriptive statistics of pretest scores
Table 4.4 Independent Sample T-Test of pretest results
The posttests of both CG and EG were marked by two teachers (rater 1 and rater
2) to ensure the inter-rater reliability of the posttest scores for comparison The correlation of posttest scores by the two raters were examined by Pearson Correlation Coefficient in SPSS The results of this correlation examination would be presented in the following tables: Table 4.5 showed the correlation of the posttest scores by rater 1 and rater 2 in CG and Table 4.6 provided the correlation of the posttest scores by rater 1 and rater 2 in EG
Table 4.5 Correlation of posttest scores of the CG by two raters
Table 4.6 Correlation of posttest scores of the EG by two raters
As can be indicated from the data in Table 4.5, Sig (2-tailed) = 000 < 05, this means there was a statistically significant correlation between scores of posttest by rater
The analysis revealed a strong correlation between the posttest scores of rater 1 and rater 2, with a Pearson Correlation Coefficient of 0.800 for CG and 0.956 for EG, both indicating significant inter-rater reliability The significance level for these correlations was confirmed with a p-value of 000, which is less than the threshold of 05 This strong agreement between the two raters underscores the reliability of the posttest scoring process.
49 both groups can be established and confirmed In addition, the scores by rater 1 were picked for the following analysis
The post-test scores of the Control Group (CG) and Experimental Group (EG) were analyzed using a Normal Q-Q Plot to verify the normality of the underlying population distribution The results, illustrated in Figure 4.2, show that the scores for both groups align closely along a straight line, confirming that they are normally distributed.
Figure 4.2 Normal Q-Q plots for TOEIC speaking posttest results
With the reliability of the speaking posttest scores validated, the mean scores of
The posttest results indicated a significant difference between the control group (CG) and the experimental group (EG) The CG achieved a mean score of approximately 11.91, while the EG scored an average of 14.44, with the EG's mean score (M=14.44, SD=3.70, N=2) being notably higher than that of the CG.
An independent samples T-test was conducted to assess the statistical significance of the difference between two values, CG (M=91, SD=3.09, N=2) Levene’s test for equality of variances yielded a significance value of 117, which is greater than the 05 threshold Therefore, it can be concluded that the conditions for equal variance are met.
A T-test for equality of means revealed a statistically significant difference between the posttest means of the control group (CG) and the experimental group (EG), with a p-value of 004, which is less than the 05 threshold This indicates that the speaking scores of the two groups were significantly different.
CG and EG after the treatment notably changed In other words, the speaking results in posttests of EG was substantially better than that of CG
Table 4.7 Descriptive statistics of posttest scores
Table 4.8 Independent Sample T-Test of posttest results
That kind of improvement can be visually seen from the following chart
Figure 4.3 Comparison of means of pretest and posttest scores
To assess the reliability and internal consistency of the questionnaire responses, the researcher utilized Cronbach’s Alpha for the six constructs The results, calculated using SPSS version 26, are presented in the table below.
Variable Cronbach's Alpha N of Items
4.1.2.1 Learners’ perceptions on using Cake to enhance TOEIC Speaking competence a Learners’ perceptions on using Cake to enhance TOEIC Speaking scores
To investigate learners' perceptions of the CAKE app's impact on improving TOEIC speaking skills, we analyzed two sets of questions focusing on scores (items 1-5) and performance (items 6-10) The findings are visually represented in Figure 4.4, which outlines the initial score construct.
Figure 4.4 Learners’ perceptions on using Cake to enhance TOEIC Speaking scores
Chart 4.4 reveals that approximately two-thirds of learners endorsed the use of Cake to enhance their scores on the TOEIC Speaking test Notably, 78.13% of participants reported an improvement in their scores, with a mean score of 3.88 and a standard deviation of 0.871.
A significant majority of participants, 71.88% (M=3.91, SD=0.963), reported improved scores in responding to questions, particularly for questions 1 and 2, which involved reading a text aloud However, lower agreement was noted for questions 8, 9, and 10, where only 50% agreed that they could respond using provided information (M=3.78, SD=0.906), and for questions 3 and 4, with 46.88% agreement on describing a picture (M=3.78, SD=0.975) Similarly, question 11, which focused on expressing an opinion, garnered only 46.88% agreement (M=3.78, SD=0.870), with 34.375% of participants indicating they did not notice significant progress Overall, despite some hesitance, most learners perceived the use of Cake positively for enhancing their TOEIC Speaking scores.
The next construct was about learners’ perceptions on using Cake to enhance TOEIC Speaking performance, as described in Figure 4.5 as follows
Figure 4.5 Learners’ perceptions on using Cake to enhance TOEIC Speaking performance
The Cake app was perceived as beneficial by students for enhancing their fluency in the TOEIC Speaking test, with a mean score of 3.84 and a standard deviation of 0.884; 56.25% of learners agreed and 18.75% strongly agreed with this statement.
Discussion of findings
This section discusses the findings from the previous chapter in the context of existing research on mobile app-assisted language learning The analysis is structured around two key themes linked to the research questions: the enhancement of TOEIC Speaking competence and learners' perceptions of the Cake app.
4.2.1 Research question 1: To what extent does Cake app impact on enhancing TOEIC speaking competence?
The Cake app significantly enhances TOEIC Speaking skills, as evidenced by pretest-posttest data and user feedback While both the experimental and control groups showed improvements in their scores, the experimental group experienced a notable increase from M=9.34 to M=9.44, attributed to their use of the app for practicing pronunciation, intonation, and fluency This led to remarkable advancements in their performance on specific tasks like reading aloud and responding to questions However, features aimed at reducing anxiety and improving test readiness received mixed feedback, indicating that psychological progress during testing relies not only on the app but also on individual capabilities.
The significant improvement in grades observed after 33 sessions of the TOEIC A+ course highlights the collaborative efforts of both teachers and students Notably, the posttest results showed a marked increase compared to the pretest, supporting findings from previous studies, including Metruk (2021), Schmidgall and Powers (2021), Fitria et al (2021), and Syafiq et al This reinforces the effectiveness of mobile app-assisted language learning in enhancing educational outcomes.
4.2.2 Research question 2: How do learners perceive Cake app in their TOEIC speaking practice?
The Cake app significantly influences students' preparation for the TOEIC Speaking Test by enhancing their learning experience It serves as an effective tool for practicing key skills such as pronunciation, intonation, stress, grammar, vocabulary, and cohesion Additionally, it emphasizes the importance of good listening and fluent speaking abilities, particularly in relation to questions 8, 9, and 10, where students must listen to a voice prompt and respond as if engaged in a real conversation.
The Cake app significantly influences student learning behaviors, particularly through intrapersonal interaction Most participants noted that the app allows for self-correction, facilitates optimal learning via short videos, and enables them to practice real-life English anytime and anywhere However, there was notable disagreement regarding the ability to subscribe to preferred channels and select specific content, attributed to the app's limited channel options and unrelated recommended videos, as highlighted by S5 in the interview Despite these limitations, users find value in the app as a tool to boost their confidence and reflexes in everyday communication, as expressed by S2.
Students find that the Cake app enhances their learning experience on their personal phones, fostering a sense of belonging while allowing them to manage notifications and reminders The app's AI voice recognition feature provides grades, and users can adjust video speed to match their learning levels However, some students face challenges installing the app due to outdated phone versions, which poses a barrier to widespread adoption of this educational tool.
Students using the Cake app value the blend of real and virtual interaction, noting that it introduces them to new vocabulary and phrases daily through diverse channel subscriptions This engagement fosters a sense of responsibility and involvement in their classes However, some students feel that classroom communication activities can be ineffective due to the varying contexts of the sentences they bring, which may hinder meaningful exchanges.
A detailed analysis of student interviews conducted via the Cake app reveals that allowing students to select their preferred content and learning methods significantly enhances the learner-centered approach This shift mitigates the drawbacks of traditional teaching, which typically involves a one-way lecture from instructor to students By adopting this model, learners engage more actively in the educational process, both in the classroom and beyond.