An investigation into efl learners’ perceptions of utilizing ai powered tools in learning writing skills a case study at the banking academy of vietnam

DECLARATIONS I hereby declare that the graduation thesis entitled “An investigation into EFL learners’ perceptions of utilizing AI-powered tools in learning writing skills - a case study

INTRODUCTION

Background of the study

The recent surge of AI technology has brought about significant changes in education, including teaching and learning English, the most popular language in the world This groundbreaking technology isn't just accelerating language acquisition; it's also enriching the learning experience for English as a Foreign Language (EFL) learners Integrating AI into English language learning is an inevitable trend, not only enhancing the quality of education but also creating favorable conditions for learners to conquer this international language (Wang, 2021)

In the trend of developing AI technologies in language learning, many AI- powered tools have been created to support EFL learners in the process of learning writing skills These tools offer personalized feedback and assistance, allowing learners to practice writing at their own pace and receive instant guidance Furthermore, EFL learners also highlight the convenience and accessibility of AI-powered tools, which allow them to practice writing at any time and from anywhere (Su et al., 2019) Thus, it cannot be denied that the use of AI-powered tools in learning writing is becoming increasingly widespread

As one of the leading universities in Vietnam, Banking Academy also offers English modules to equip students with essential foreign language skills for future careers In line with the general trend, the use of AI-powered tools to support English language learning, particularly for writing skills, has also become popular in the Banking Academy Therefore, understanding how students perceive and use these AI tools is very important to find appropriate teaching and learning methods, enhancing the quality of training as well as improving students' abilities

All of the preceding points have been the driving force behind this study, which is titled "An investigation into EFL learners’ perceptions of utilizing AI-powered tools in learning writing skills - a case study at the Banking Academy of Vietnam."

Significance of the study

By understanding the perceived utilization of AI-powered tools, the study can highlight how AI tools can potentially assist EFL learners in learning writing and inform strategies for maximizing their effectiveness in developing writing abilities This may involve identifying specific writing tasks where AI tools are most beneficial and developing guidelines for their responsible use Besides, instructors can offer some advice and assistance to help learners overcome barriers that hinder their optimal use

In addition, the findings can guide educators in developing appropriate teaching methods that leverage AI tools alongside traditional approaches This may involve creating blended learning environments that combine AI-powered feedback with teacher guidance and peer interaction Furthermore, since this study explores the use of AI writing tools without educators’ guidance, it also encourages the practice of using AI tools as a useful self-learning method for EFL learners.

Objectives of the study

The study aims to figure out how EFL learners perceive the utilization of AI- powered tools in learning writing skills as well as the challenges that hinder them from taking advantage of these tools Some suggestions would be proposed to increase the effectiveness of making use of AI tools in improving writing skills Furthermore, the study is conducted in the hope of assisting educators and learners in identifying best practices for utilizing AI tools to maximize their benefits for learners

Research questions

In order to shed light on the goal of the study, two questions are addressed as follows:

1 How do EFL learners at the Banking Academy of Vietnam perceive the utilization of AI-powered tools in learning writing skills?

2 What challenges do EFL learners at the Banking Academy encounter in utilizing AI-powered tools for learning writing skills?

Scope and limitations of the study

The research primarily focuses on exploring how EFL learners at the Banking Academy of Vietnam perceive the use of AI tools in learning writing as well as the challenges they have encountered when utilizing these tools The students of different years taking English writing classes at the Banking Academy were invited to take part in the research and provide their personal views The study was carried out from April 8 to April 24, 2024

The study has some limitations First, due to time constraints, the study was only conducted within the Banking Academy, so it may not be generalizable to the large population of EFL learners Second, the study only explores EFL learners' perceptions at a specific point in time Further studies would be needed to understand how perceptions and usage patterns change over time Third, because the research was carried out at the end of the semester, many students were busy studying for exams and doing projects, so few students could spend time participating in interviews Consequently, qualitative methods may not be optimally utilized.

Structure of the study

The thesis includes five chapters, as follows:

Chapter 1 Introduction: This chapter sets the stage for the thesis writing, providing the background of the study, the purpose of the study, the research questions, its significance, as well as the scope and limitations of the research

Chapter 2 Literature Review: This chapter provides an overview of theories that are relevant to EFL learners and Writing skills, AI-powered tools, the Technology Acceptance Model as well as the previously published works on the topic

Chapter 3 Methodology: This chapter typically presents the roadmap for conducting the research, including the research setting, research methods, research instruments, data collection procedure and data analysis

Chapter 4 Findings and Discussion: This chapter aims to present the results of the investigation and analyze their meaning It also includes a discussion of the results obtained

Chapter 5 Conclusion and Recommendations: This chapter provides closure to the thesis writing It summarizes the key points, highlights the significance of the findings, provides recommendations, and paves the way for future research

LITERATURE REVIEW

EFL learners and Writing skills

2.1.1 Importance of writing skills for EFL learners

There are four fundamental linguistic skills when it comes to learning a language: speaking, listening, reading, and writing Among them, speaking and writing are typically seen as productive skills since they both require the production of language in the form of words, phrases, sentences, or paragraphs Nevertheless, writing is a more planned and structured process that frequently takes more time for preparation and polish It is also considered a challenging skill to master, demanding dedicated attention and consistent practice Therefore, learning writing skills is very important for language learners to be proficient in that language

In today’s globalized world, English is the most common language and is regarded as a powerful tool for international communication, making it essential for individuals to develop English writing skills in order to facilitate the exchange of ideas with others Particularly for EFL learners who learn English as a foreign language, learning writing skills is of the utmost importance because it plays an indispensable part in their English classes as well as brings great benefits to their future careers

Writing skills are integral to the EFL courses since learners are required to complete written assignments and assessments, which can vary in length from several paragraphs to multiple pages Advanced writing skills are one of the essential needs for higher academic achievement and other writing-related activities (National Assessment of Educational Progress, 2002) Likewise, Ismail (2011) asserted that writing has always been considered a critical skill in EFL education, in which students are required to spend

6 a considerable amount of time to improve their writing abilities Writing is undoubtedly an important part of learning English as a foreign language The importance stems from the fact that it helps learners reinforce the grammatical structures and develop the student’s vocabulary Therefore, writing skills should be given more priority in EFL lessons in order to better equip students with the abilities and necessary skills for handling intercultural communication

Furthermore, writing skills are essential to social interaction and career advancement According to Messum et al (2016), there is an increase in job advertisements that emphasize skill requirements, such as "good command of written English" or "excellent written communication." Schmidgall and Powers (2020) polled

200 professionals (100 in each study) in various organizations where English is the official working language and found that approximately 80% of respondents rated English writing skills as "important" or "very important" in their workplaces These findings are consistent with the study results of Moore et al (2015), in which 83% of 71 surveyed participants assumed that English writing skills are "important" (38%) or "very important" (45%) for freshly employed graduates in their professional occupation In a similar manner, Kim (2019) conducted a survey and found that English writing skills were evaluated as the most necessary skill in the 16 tasks listed in the questionnaire In actuality, writing certain things, such as a message, an email, a report, a formal letter, or a statement, is necessary for communication and business activities Writing has inserted itself into every facet of life, from just a short email to literary masterpieces It serves as a powerful means of communication and helps connect people from different backgrounds Thus, having strong writing skills is crucial for EFL learners, both in their daily lives and in the future

2.1.2 Difficulties encountered by EFL learners in learning writing skills

Nunan (1999) believes that writing is the most difficult of the four fundamental linguistic skills because it requires a diversity of tasks and specific abilities for

7 expressing thoughts and feelings in written form Learning how to write effectively has always been considered a challenge for language users, whether they are first, second, or foreign language learners

Research has indicated that vocabulary deficiency is a challenge encountered by most EFL learners in learning writing skills (Santos, 1988; Astika, 1993) and that having a rich vocabulary is one of the most significant features that ensure the quality of writing (Maamuujav, 2021) In line with this, Nhi and Dao (2023) surveyed 258 English-major students at An Giang University in order to find out common difficulties in writing The results revealed that most students face vocabulary-related problems in word choice, word meaning, word repetition, simple words, and vocabulary deficiency They also find it challenging to adapt their vocabulary based on the audience and purpose of their writing Thus, it is necessary for EFL learners to be mindful of using appropriate vocabulary that resonates with the context and their readers

Some EFL learners are also having trouble with grammar Hans et al (2017) defined grammar as “the set of structural rules governing the composition of clauses, phrases, and words in any given natural language” (p 39) Grammar is one of the most key components of written language because of its role in making the writing clear, concise, and easy to read However, it is also a burdensome task for EFL learners due to the variety of rules and structures According to Ellis (2006), grammar difficulty can be distinguished into two types: “the difficulty learners have in understanding a grammatical feature” and “the difficulty they have in internalizing a grammatical feature so that they are able to use it accurately in communication” (p 88) In a similar manner, Abushihab (2014) conducted a study in order to identify and categorize grammatical mistakes in writing generated by students at the Faculty of English Language in Gazi University of Turkey In this study, a randomly selected group of 20 learners were asked to compose an essay of 200 to 250 words regarding the problems they experience while learning English The author and two other raters analyzed the written data and found a

8 total of 179 grammatical errors, with the five most prevalent errors as follows: tenses, prepositions, articles, active and passive voice, and morphology

Organization is another cause of anxiety for EFL learners when learning writing skills According to Brown (2004), organization covers the effectiveness of introduction, conclusion, logical sequences of ideas, and appropriate length of writing It is about how to arrange the ideas so that they are connected and make your writing clear A well- organized writing supports the writers to communicate their ideas effectively However, EFL learners often face challenges in organizing their writing, as highlighted in various research papers Studies have shown that learners commonly make errors in organization, such as incoherence, disunity, and misuse of cohesive devices (Saud, 2015; Giang et al., 2023) EFL learners may struggle with writing organization due to complexities like lack of time, research background, and critical thinking skills (Asnas et al., 2023)

In addition, a lack of regular guidance, feedback, and editing from teachers is another difficulty faced by EFL learners in learning writing skills With a large number of students in today's classrooms, it is a significant challenge for educators to allocate sufficient time to offering detailed feedback on each student's written work Moreover, the different levels of learners further complicate the task of providing appropriate feedback to address learners’ needs (Asep, 2014) Without receiving specific feedback, students are uncertain about their mistakes and areas for improvement, leading to repeated errors and a lack of learning motivation Writing is a skill that requires time and consistent practice When students lack direction and motivation, they may struggle to make progress, which can negatively impact their overall learning outcomes

AI-powered tools

2.2.1 Overview of Artificial Intelligence (AI)

The concept of "Artificial Intelligence" (AI) first appeared in the summer of 1956 when an American computer scientist, John McCarthy, introduced it at the Dartmouth Conference to describe the science and technology of creating intelligence machines that can imitate human behavior Since then, there has been no official definition of artificial intelligence, but it can be understood depending on each person's perspective According to Bellman (1978), artificial intelligence is the automation of abilities that are consistent with human thinking, such as learning, analyzing, problem-solving, decision-making, and even creativity Rich and Knight (1991) assumed that artificial intelligence is the field of study that aims to make computers capable of doing things better than humans, especially in areas that require human expertise Briefly, AI is a field of computer science and technology that specializes in solving problems and performing tasks we typically think of as needing human intelligence

After decades of development, AI has crept into every corner of human life, from chatbots, virtual assistants such as Siri (Apple) and Alexa (Amazon) to driverless cars The explosion of AI is inevitable in the context of international integration and the rapid development of the 4.0 industrial revolution In recent years, constant technical innovation, strong computer processing, and internet network infrastructure have created a solid launch pad for the development of AI As a matter of fact, AI breakthroughs and applications have become pioneering technology trends, strongly promoting a wave of innovation in all spheres of life

In education, AI is playing an increasingly important role, not only improving the teaching and learning process but also creating new opportunities for customizing the experiences of learners and educators AI applications in education include a wide range of tools and programs, from personalized learning systems, test creation support to educational chatbots For example, a personalized learning system can track a student's

10 learning progress and recommend content and exercises appropriate to their ability and level of progress Besides, AI also helps reduce the work burden for teachers, helping them focus more on interacting with students and improving the quality of teaching Thanks to AI, teachers can receive detailed data analysis of their students' learning progress, helping them recognize and respond promptly to any difficulties students may be experiencing With the constant development of technology, there is no limit to the ways in which AI can contribute to the education industry in the future

2.2.2 AI-powered tools for writing

With the rapid development of AI in education, many AI-powered tools have been created to bring benefits to learners, especially for those who are on the journey to conquer English, a global language The key to learning English effectively is to regularly practice the skills of listening, speaking, reading, and writing In particular, writing skills are one of the common "weaknesses" of many English learners whose mother tongue is not English Therefore, the emergence of many AI-powered tools has greatly assisted EFL learners in the process of learning writing skills

For that reason, many studies have been conducted and have discovered popular

AI tools and their diverse functions that help EFL learners overcome difficulties they encounter in learning writing skills For example, Zulfa et al (2023) found that Grammarly is the most frequently used tool in EFL learners’ writing process This tool helps learners detect vocabulary and grammar errors in their writing It is also loved by users for its feature of explaining errors and suggesting better alternatives This not only helps learners improve the quality of their writing but also helps them remember vocabulary and grammar better (Maulidina and Wibowo, 2022) Another popular AI tool is Quillbot, an online paraphrasing tool that is perceived favorably by EFL learners It is evident from the articles that learners' paraphrasing abilities have considerably improved in terms of parallel construction, word order, and sentence structure (Hieu et al., 2022)

In addition, ChatGPT and Gemini, the two leading generative AI applications today,

11 show potential for enhancing writing skills in educational settings Research indicates that these tools can significantly improve fluency, flexibility, and narrative originality in writing tasks among students, highlighting their assistance in verbal creativity EFL learners can utilize these AI tools by aiding in summarizing complex concepts, improving grammar, and effectively communicating ideas, making them a useful tool for developing writing skills (Levy et al., 2023) Besides, there are a number of other AI- powered tools, such as ProWritingAid, Paraphraser, EssayGrader, etc., that also have various useful functions and can be utilized by EFL learners in learning writing.

Technology acceptance model (TAM)

The Technology Acceptance Model (TAM) is one of the most commonly used theories in the study of technology user behavior This model helps explain the factors that motivate users to adopt and use new technology, including artificial intelligence (AI) technology TAM was first developed by Fred Davis in 1986, based on the Theory of Reasoned Action (TRA), and has been widely applied in research in many fields (Phuong, 2023)

According to Davis (1986), TAM consists of the following key variables:

Perceived Usefulness (PU): The degree to which users believe that using a particular technology will help them achieve their goals and improve work efficiency

Perceived Ease of Use (PEU): The degree to which users believe that using a particular technology will be easy and convenient

Attitude Toward Using (ATU): The user's level of liking or disliking the technology

Behavioral Intention (BI): The degree to which the user intends to use the technology in the future

Actual use (AU): The actual level of use of the technology by users

In the TAM model, Davis et al (1989) assumed that attitude is the factor that modestly affects user intention and behavior, while perceptions of ease of use and usefulness of technology have a direct influence on their behavioral intention In the same way, Mathieson (1991) proposed that removing the attitude factor from TAM would not have a significant impact on the reliability of the model However, several recent studies have figured out that attitude is an important predictor of behavioral intention, especially in situations where technology use is not mandatory but entirely voluntary (Athiyaman, 2002) Therefore, the attitude factor will still be used in the model to measure EFL learners’ perceptions since the use of AI-powered tools in learning writing skills is voluntary

Figure 2.1: Technology Acceptance Model (TAM) (Davis et al., 1989)

Students' perceptions play a crucial role in shaping their behaviors and actions (Smith et al., 2020) For instance, research indicates that students' perceptions of social media impact can influence their social behavior (Hashim et al., 2016) Additionally, students' perceptions of on-campus foodservice operations, such as food quality, ambience, and service quality, ultimately affect their dining frequency and behavior (Smith et al., 2020) Furthermore, students' positive perception of technology indeed correlates with increased motivation, enthusiasm, and academic achievement Research also indicates that students' attitudes and motivation significantly influence the integration and development of technology in learning (Nozaki et al., 2023)

According to Teo (2018), when students utilize learning tools, they often examine the technologies' usability and usefulness in addition to their relevance to the learning objectives, the users' technical proficiency, and the accessibility of facilities This relates to the concept of the technology acceptance model (TAM), which could be applied to study how technology is adopted and used Thus, the TAM would be an appropriate model to be applied in this research as it could help find out how students perceive AI- powered tools and how they are utilizing these tools in learning writing skills

In the framework of this research, perceived usefulness illustrates EFL learners' belief and confidence in using AI-powered tools to improve their performance in writing Perceived ease of use represents the degree to which EFL learners believe that using AI- powered tools is easy and does not require much effort Attitudes and behavioral intentions toward using AI-powered tools in developing writing skills are created by beliefs about the usefulness and ease of use of these tools EFL learners will be more likely to use the AI tools if they believe these technologies are simple to use and have been proven to increase the quality of their writing The TAM not only measures EFL learners' perceptions toward utilizing AI-powered tools for learning writing skills but also identifies factors preventing or facilitating their adoption of this technology.

Related studies

With the development and increasing popularity of AI-powered tools for enhancing writing skills, a lot of research has been conducted to explore EFL learners’ perceptions of using these tools

Le (2023) conducted an investigation into students' perceptions of AI technology application in English writing classes, aiming to figure out the benefits of this technology as well as the challenges students face when applying AI to support writing skills The study used both quantitative and qualitative data to get answers from participants For the quantitative phase, the researcher used a questionnaire presented on a Likert scale to explore how students perceive the advantages and disadvantages of this technology For

14 the qualitative phase, the interview approach was utilized to clarify the data obtained from the previous phase The research took place in four General English classes at Vietnam National University with 100 participants The findings showed that the majority of students believed that AI-based learning tools play an important role in improving their writing skills and increasing the quality of their writing In addition to the beneficial features of these tools, their accessibility, versatility, and simplicity were all highly appreciated by learners Regarding interest and motivation, most students also agreed that AI-powered tools could increase their engagement and enthusiasm for studying On the contrary, the study also indicated some objective and subjective factors that might hinder their application of these tools The two biggest challenges were inaccurate feedback and dependence on technology Overall, the research has provided valuable insights into the benefits and drawbacks of using AI technology in English writing classes Although these tools still have limitations, the great benefits they bring cannot be denied Therefore, developing AI in education is a general trend Along with that, it is very important to improve teachers' capacities for applying AI in the classroom and strengthen policies to ensure fairness

Another study is “The use of AI and technology tools in developing students’ English academic writing skills” by Zulfa et al (2023) This research examined how students leverage AI and technology tools to enhance their English writing abilities To gain a deep understanding of the research topic, the authors employed a mixed-methods approach, utilizing both quantitative and qualitative methodologies The questionnaires were completed by 73 students at one of the State Islamic Universities in Indonesia, and

11 of them engaged in interviews through the Zoom application The results showed the various AI tools used by students in the English writing process, and the most popular tools are Grammarly and QuillBot The study highlighted the beneficial functions of these tools based on students’ experiences It also emphasized students’ improvements in grammar, language structures, vocabulary usage, and idea organization by utilizing

AI technology in English academic writing This study implies that AI-based tools have

15 the potential to revolutionize the development and evaluation of writing skills Students could improve self-awareness, self-learning, and self-confidence when utilizing this technology However, it is essential that students make appropriate use of these tools to ensure the development of writing skills

In line with this, Utami et al (2023) conducted a study to investigate Indonesian students’ perspectives on utilizing AI technology in academic writing classes This study applies the Technology acceptance model of Davis et al (1989) to explore students' perceptions of AI tools in three aspects: the ease of use, the usefulness and the attitude toward usage 58 eleventh-grade students from three senior high schools in Semarang, Indonesia, were invited to participate in the study These students were required to work in groups of two to four members and apply AI-based learning tools in writing inside and outside the classroom Both a questionnaire and an interview were used to obtain the data The findings indicated that most participants had a positive perception of using AI tools in writing They confirmed that AI-powered tools could assist them in all four basic stages of writing, including planning, drafting, revising, and editing Furthermore, the majority of students found it quite easy to access AI tools because of the availability of the internet and their proficiency in technical skills Especially, the average mean score of attitudes toward using was pretty high, at 4.35, indicating that a significant number of students enjoyed using AI-based learning tools Based on the results, the author recommended that students could use these AI technologies in writing classes because of their numerous benefits However, they should only be used as supplementing tools, and teachers still play an important role in giving systematic instructions to students

With the same concern, Al-Raimi et al (2024) carried out research to explore EFL learners’ perspectives on utilizing AI tools for enhancing writing skills This investigation was launched at a university in the Sultanate of Oman at the end of the first semester of the school year 2023–2024 The researcher collected and analyzed data using quantitative techniques, administering a 5-point Likert questionnaire to 61 university students The findings revealed that the students generally approved of using AI for

16 writing development Interestingly, there wasn't a significant difference in perception between genders, as these tools have many beneficial functions, such as paraphrasing texts, checking for spelling and grammatical errors The study also suggested using AI- powered tools to help students overcome difficulties in academic writing These findings imply that EFL educators and learners can employ AI writing tools to assist EFL learners in improving their overall writing abilities Besides, additional training and understanding of these AI technologies are required in order to utilize them correctly and effectively

In summary, despite being conducted in different countries with varying participants ranging from high school to university students, all the aforementioned studies suggest that EFL learners generally exhibit positive perceptions towards utilizing AI-powered tools in learning writing skills This technology is viewed favorably by students for enhancing their English language proficiency, improving content quality, reduce mistakes, and develop language usage in their writing Additionally, incorporating AI-powered tools in English writing classes could improve student learning motivation and performance Although there are still difficulties and limitations when using these tools, the majority of EFL learners still perceive AI-powered tools as beneficial aids in enhancing their writing skills and language learning experiences However, further research with larger sample sizes would be necessary to acquire a more profound understanding of AI-powered tools’ effectiveness for English language learners, as well as to identify any potential drawbacks

METHODOLOGY

Research methods

This study utilizes a mixed-methods approach, employing questionnaires for quantitative data collection and semi-structured interviews for qualitative data collection

Quantitative research is a way of getting information about the research object in the form of arithmetic and statistical data This data, often gathered through large-scale surveys using questionnaires, allows researchers to analyze trends and draw generalizable conclusions about a population In other words, it involves quantifying the data collected to serve statistical and analytical purposes Quantitative research is considered a more structured and objective approach to data collection The use of statistical analysis helps researchers rely on measurable phenomena, reducing the possibility of personal bias or opinion interfering with the findings Besides, using large sample sizes may increase the accuracy of statistical analysis, helping to make important decisions Quantitative data is generated quickly, and the analysis process is often simple, making it a great choice for researchers who need immediate results

However, quantitative research only collects numerical data and does not explore the underlying reasons for the answers This limits the ability to gain rich insights into research participants' thoughts and attitudes Thus, the researcher combined quantitative and qualitative methodologies to dive deeper into this study Qualitative research is a method of gathering 'non-numerical' information regarding the object of study for the purpose of in-depth analysis or assessment This information is often collected through interviews, direct observations, or focus group discussions using open-ended questions and is often applied in cases of small, focused research samples The qualitative research

18 method allows researchers to gain rich insights into what underlies respondents' decision-making It can also help explain the relationships between these variables discovered in quantitative studies.

Research locale and population

In line with the general trend of AI technologies, many EFL students have chosen to utilize AI-powered tools for studying writing skills, with the expectation of learning more effectively and achieving higher results Therefore, the research was conducted at the Banking Academy of Vietnam to find out how the students perceive and use these tools The research was carried out at the end of the second semester of the academic year 2023-2024, targeting students currently enrolled in writing skills classes offered at the Banking Academy, who were eager to try any learning strategy that would help them enhance their writing skills An estimated 600 students were taking writing skills classes at the time of the study.

Resesearch sample

The phrase "sample size" refers to the number of people included in a research study to reflect a certain demographic Determining the appropriate sample size is one of the most critical aspects of statistical analysis A sample size that is too small will not produce accurate results or sufficiently represent the realities of the community under study On the other hand, while larger sample sizes are more representative and have narrower error margins, an overly large sample size may significantly increase the cost and duration of the research project

In this study, Andrew Fisher's formula was used to determine the minimum required sample size, with a 95% confidence level, a standard deviation of 0.5, and a confidence interval (margin of error) of ± 8% The confidence level (95%) was converted into a Z-Score (1.96)

Sample size = ( Z−score) 2 × StdDev × (1−StdDev)

In this study, a sample size of at least 151 people would be necessary.

Research instruments

The questionnaire was built based on the Technology Acceptance Model and adapted from Le (2023), Utami et al (2023), and Al-Raimi (2024) to collect data on a large sample The goal of the questionnaire is to explore EFL learners' perceptions of utilizing AI-powered tools in learning writing skills The questionnaire is divided into three main sections, with a total of 41 items

Section 1 consists of five items to collect demographic and background information about the participants, including gender and year of study Additionally, this section also explores which AI-powered tools participants have used as well as the level of their use

Section 2 is divided into five small parts Part 1 includes six items to find out how EFL learners perceive the simplicity and convenience of using AI tools Part 2 consists of 12 items to explore students' perceptions of the usefulness of AI tools in learning writing skills Part 3, including five items, is to explore respondents' attitudes, and Part

4, including five items, is to learn about their behavioral intentions in utilizing AI in developing writing skills Finally, Part 5 contains five items to explore the difficulties preventing learners' optimal use of AI tools in learning writing

Section 3 includes three open-ended questions to learn more about the benefits and difficulties for EFL learners in leveraging AI tools to develop writing skills At the same time, this section also explores participants' thoughts or suggestions to help improve the effectiveness of using AI tools in learning writing

To further explore participants' thoughts and reasons for their choices in the questionnaire, a semi-structured interview was conducted The interview was based on a list of primary questions mirroring the topic that needed to be covered Additional questions starting with “how” and “why” will be asked to delve deeper into the participant's thinking However, the order and way of asking questions depended on the context and characteristics of the interviewee Below are the main questions asked in the interviews

The first question is, “Do you find AI-powered tools easy to access and use?”

The second question is, “In what ways do you think AI-powered tools are helpful when you learn writing skills?”

The third question is, “Among the AI tools you have used, which one do you find most effective?”

The fourth question is, “What challenges have you encountered when utilizing AI-powered tools to learn writing skills?”

The fifth question is, “Do you think AI tools should be applied more inside and outside the classroom?

And the final question is, “What would you suggest or recommend to optimize the effectiveness of using AI-powered tools in learning writing skills?”

These questions are all aimed at better understanding participants' thoughts and experiences in leveraging AI-powered tools to learn writing skills

Data collecting procedure

The research is divided into two phases to collect quantitative data and qualitative data for analysis

Figure 3.1: Quantitative data collection procedure

Based on the Technology Acceptance Model and adapted from Le (2023), Utami et al (2023), and Al-Raimi (2024), the questionnaire was designed to address the five aformentioned research questions It employed a variety of question formats, including single-answer, multiple-choice, Likert scale, and open-ended questions In addition, the questionnaire is divided into small parts to improve readability and prevent respondents from feeling overwhelmed by its length The researcher also utilized Google Form, a user-friendly and widely used platform for creating online surveys

The draft questionnaire was sent to the instructor for feedback and then sent to three participants for pilot testing Those participants pointed out some grammatical errors and inappropriate word usage They also identify a few unclear or ambiguous statements Finally, taking those comments into account, the researcher modifies the questionnaire

The finalized questionnaire was distributed online to the target participants via messaging applications such as Zalo, Messenger, and email Furthermore, the researcher proactively printed copies of the questionnaire and went to the school to conduct a paper- based survey This comprehensive approach ensured the survey reached a wider audience, fostering a more representative sample

After 17 days, 152 responses were collected through the Google Form The participants' answers were fully recorded in a Google Sheet linked to the online survey platform, making it convenient for the researcher to synthesize and extract data into an Excel file for data processing and analysis

Figure 3.2: Qualitative data collection procedure

Based on the quantitative data and adapted from Le (2023), open-ended questions for the semi-structured interview were developed to encourage participants to elaborate on their experiences, attitudes, and beliefs toward the utilization of AI tools when learning writing skills The questions closely align with the research topic and aim to address the research questions.

Convenience sampling is used by the researcher to recruit participants for interviews The selected students were those who participated in the survey in the

23 quantitative phase Students who are easily accessible and most likely to agree to an interview are invited by the researcher Finally, there were 5 students volunteering to participate in the interview

Step 3: Setting up the interview

Individual interviews are conducted online via Messenger audio call, in accordance with the participants' preferences.This format also allowed for flexibility in scheduling and ensured a private and comfortable environment for participants to share their thoughts openly and honestly

During the interview, the researcher will ask the main questions previously designed and add additional questions to gain a deep understanding of the participant's thoughts The researcher gathered the data by taking notes on the interviewers' main ideas and recording them with informed consent from the participants.

Data processing and analysis

The data in the Excel file was imported into SPSS version 20, a specialized software for data statistics often used in scientific research fields Then, Cronbach's alpha was used to measure the reliability of the scale This useful tool will help check whether the observed variables in a factor are reliable or not Additionally, it shows which of the observed variables of a factor contribute to the measurement of the factor concept and which variables do not Numerous scholars concur that when Cronbach alpha is from 0.8 to nearly 1, the measurement scale is good; from 0.7 to nearly 0.8 is acceptable The correlation coefficient for all variables must be at least 0.3 According to some other researchers, when the topic being tested is novel or unfamiliar to respondents in the study setting, a Cronbach alpha of 0.6 or higher is appropriate (Nunnally, 1978; Peterson, 1994; Slater, 1995).In addition, “Cronbach's Alpha if Item Deleted” should be taken into

24 account When it is greater than Cronbach's Alpha, it indicates that the measured variable is not positively contributing to the reliability of the research questionnaire In this case, eliminating that variable can increase the scale's reliability

Then, descriptive statistics were utilized in the next step It is a statistical method that includes concise descriptive coefficients to summarize a set of data, which may be representative of the entire population or a sample of the population Specifically, it is used to shorten, organize, simplify, describe, and present data in numbers or visual charts It shows the average value score (mean), maximum value (max), minimum value (min), and standard deviation (std deviation) of each observed variable Descriptive statistics provide an overview of the data, allowing researchers to present the data in a concise and understandable way, and easily grasp the main characteristics of the data The use of descriptive tables for Likert scales is to compare the average value of observed variables in the same group This helps identify which variables have the highest and lowest levels of agreement among the surveyed population From the data processed through SPSS version 20, the researcher analyzed the results, thereby addressing the research questions and achieving the goal of the study

Table 3.1: Interpretation of the mean score

According to Pinto and Stein (2015), transcribing recordings after an interview is crucial to accurately document and analyze the information gathered, ensuring details are preserved and accessible for further review and investigation Thus, the recordings from the interviews were transcribed, enabling the researcher to take the interview's contents and analyze them efficiently The researcher listened back to the audio recordings while typing out the transcription This helps the researcher better grasp the nuances of the conversation and the emotions of the participants

In the next step, the researcher coded the qualitative data by highlighting words and phrases into themes that pertain to the research topic: perceived ease of use, perceived usefulness, attitude toward using, behavioral intention, and challenge in using This division can help researchers generalize how students perceive the use of AI tools in learning writing skills, whether they feel positive, negative, or neutral

Based on this analysis, valuable recommendations and suggestions can be formulated for the research paper's conclusion Furthermore, the qualitative data from interviews was integrated with the statistical analysis to provide a more comprehensive understanding

FINDINGS AND DISCUSSION

Quantitative results

Frequency of using AI-powered tools

Table 4.1 summarizes the demographic information of the 152 people participating in the survey, providing an overview of the sample

It can be seen that the number of females accounted for the highest proportion, at 78.3% of the total number of respondents in this study The figures for males and others who refuse to reveal their genders are slightly lower, at 17.1% and 4.6%, respectively Regarding year of study, the majority of participants in this survey are fourth-year students (37.5%), followed by first-year students (30.9%) and third-year students (17.8%) Second-year students make up the remaining 13.8% of surveyed participants Interestingly, 39.5% of the surveyed takers confirm that they often use AI-powered tools to learn writing skills Meanwhile, only 9.9% of participants said that they seldom use these AI tools to develop their writing skills The percentages of students who occasionally use AI tools and students who use these tools very often are 37.5% and 13.2%, respectively

Table 4.2: Reliability analysis of perceived ease of use (PEU)

Scale Variance if Item Deleted

Cronbach's Alpha if Item Deleted

Table 4.3: Reliability analysis of perceived usefulness (PU)

Scale Variance if Item Deleted

Cronbach's Alpha if Item Deleted

Table 4.4: Reliability analysis of attitude toward using (ATU)

Item-Total Statistics Scale Mean if

Scale Variance if Item Deleted

Cronbach's Alpha if Item Deleted

Table 4.5: Reliability analysis of behavioral intention (BI)

Scale Mean if Item Deleted

Scale Variance if Item Deleted

Cronbach's Alpha if Item Deleted

Table 4.6: Reliability analysis of challenge in using (C)

Scale Mean if Item Deleted

Scale Variance if Item Deleted

Cronbach's Alpha if Item Deleted

As can be seen from the five above tables, the Cronbach’s alpha coefficients of all groups of studied factors are greater than 0.8 This shows that the reliability of the scales is high In general, Cronbach's alpha coefficient is in the range of 0 to 1 If it is greater than 0.7, the measurement scale is acceptable The higher the Cronbach’s alpha coefficient, the higher the reliability of the scale (Nunnally & Bernstein, 1994)

In addition, the test results show that all observed variables have an appropriate total correlation coefficient (Corrected Item - Total Correlation > 0.3), and no variable has a Cronbach's alpha coefficient if Item Deleted greater than the total Cronbach's alpha coefficient, proving that the variables are of good quality and there is no need to remove any items (Nunnally & Bernstein, 1994)

Table 4.7: Summary of descriptive statistics Factors Items N Minimum Maximum Mean Std Deviation

As is presented in Table 4.7, the average values of the observed variables all range from 3 to 5 on the 5-point Likert scale This indicates that the majority of students who responded to the survey tended to agree with the statements in the questionnaire The item PEU 1: “AI-powered tools are easy to use.” and PEU 3: “AI-powered tools can easily be used on various devices.” get the highest mean of 4.03 while item C1: “I encounter technical issues such as slow loading, server downtime, etc while using AI- powered tools for learning writing skills.” get the lowest mean of 3.28

In addition, it is noticeable that the values in column N are equal to the sample size (N = 152), proving that these items do not have missing values The maximum and minimum values in the table are both reasonable and do not exceed the scale limit (from

4.1.4.1 AI-powered tools used by EFL learners in learning writing skills

Figure 4.1: AI-powered tools used by students in learning writing skills

The chart illustrated that there were a variety of AI-powered tools used by students when learning writing skills Interestingly, each student often utilized two or more tools to assist them in the process of learning writing This indicated that the use of AI tools was increasingly widely adopted among EFL students

It is noticeable that the most popular tool was Grammarly, which was chosen by

117 students, corresponding to 77% of the surveyed respondents ChatGPT was next on the list at 75.7%, making it slightly less popular than Grammarly QuillBot came next, with 93 students using it Ranked in fourth place was Paraphraser, with 48 out of 152 students favoring it, just ahead of Google Bard In addition, some students also

35 mentioned a few other tools they used to learn writing, such as Bing, Copilot, Claude, etc

Figure 4.2: AI-powered tools ranked by effectiveness

Students' preferences for AI-powered tools to improve their writing were clearly depicted in the pie chart ChatGPT was rated as the most effective tool for learning writing skills, at 38.2%, followed by Grammarly at 28.3% QuillBot trails behind at 18.4%, and Google Bard was ranked fourth with 8.6% This suggested that these AI- powered tools could offer valuable functionalities and contribute to students’ writing development The distribution also revealed a clear interest among students in leveraging the functionalities offered by these AI tools

4.1.4.2 EFL learners’ perceived ease of using AI-powered tools in learning writing skills

Table 4.8: EFL learners’ perceived ease of using AI-powered tools in learning writing skills

No Question items Mean Standard deviation Interpretation

1 AI-powered tools are easy to use 4.03 829 Agree

2 AI-powered tools are easily accessible 4.01 873 Agree

3 AI-powered tools can easily be used on various devices 4.03 845 Agree

4 AI-powered tools have user-friendly interfaces 3.84 849 Agree

5 AI-powered tools have easy-to- follow instructions 3.89 907 Agree

6 AI-powered tools provide direct and understandable responses 3.73 920 Agree

As shown in the above table, the simplicity and convenience of using AI-powered tools to learn writing skills were highly appreciated by students (M = 3.92) The items

“AI-powered tools are easy to use” and “AI-powered tools can easily be used in various devices” received the highest mean score of 4.03, followed by the item “AI-powered tools are easily accessible” with a mean of 4.01 This is correlated to the availability of the internet and smart devices in the 4.0 era

Additionally, most students agreed that AI-powered tools have user-friendly interfaces (M = 3.84, SD = 0.849) and easy-to-follow instructions (M = 3.89, SD 37

0.907), so students could easily interact with these tools Regarding the last item, a large number of students also consented that they could understand the direct feedback provided by AI tools, which have a mean of 3.73 and a standard deviation of 0.920

4.1.4.3 EFL learners’ perceived usefulness of AI-powered tools in learning writing skills

Table 4.9: EFL learners’ perceived usefulness of AI-powered tools in learning writing skills

No Question items Mean Standard deviation Interpretation

1 AI-powered tools help me enhance my writing skills 3.64 902 Agree

2 AI-powered tools improve the quality of my writing 3.88 879 Agree

3 AI-powered tools help me in achieving my learning objectives 3.72 870 Agree

4 AI-powered tools increase my ability to organize the ideas 3.69 871 Agree

5 AI-powered tools can enrich my vocabulary 3.82 917 Agree

6 AI-powered tools offer me the ability to diversify my word choices 3.92 888 Agree

7 AI-powered tools help me understand the use of transition words 3.82 849 Agree

8 AI-powered tools improve my ability to use various sentence patterns 3.73 1.003 Agree

9 AI-powered tools help to reinforce my grammar 3.84 872 Agree

10 AI-powered tools increase my fluency in writing 3.87 926 Agree

11 AI-powered tools keep me active and motivated in learning writing 3.57 988 Agree

AI-powered tools give me the flexibility to learn writing skills anytime and anywhere

The mean score of 3.77 revealed that a large number of participants hold a high opinion of AI-powered tools for enhancing their writing skills They agreed that AI- powered tools helped increase their performance in writing (M = 3.64, SD = 0.902) as well as improve the quality of their written works (M = 3.88, SD = 0.879) Furthermore, many participants felt that they could reach their learning goals when utilizing these AI tools (M = 3.72, SD = 0.870)

The effectiveness of AI-powered tools in developing vocabulary was appreciated by many students They claimed that AI-powered tools could enrich their vocabulary (M

= 3.82, SD = 0.917) and help them understand the use of transition words (M = 3.82, SD

= 0.849) Especially, the item “AI-powered tools offer me the ability to diversify my word choices” got the highest mean score of 3.92, with a standard deviation of 0.888

In addition, many participants believed that AI-powered tools helped reinforce their grammar (M = 3.84, SD = 0.872) and improve their ability to use various sentence patterns (M = 3.73, SD = 1.003), demonstrating the AI tools' capacity to help them minimize errors in their writing and refine sentences to make them better

Furthermore, students claimed that they could utilize some AI-powered tools to develop and organize ideas (M = 3.69, SD = 0.871) They also agreed that AI tools helped enhance their fluency in writing (M = 3.87, SD = 0.926) Moreover, these AI-powered

39 tools give students the flexibility to learn writing skills anytime and anywhere (M = 3.75,

SD = 0.915) and keep them active in learning (M = 3.57, SD = 0.988) This indicated that AI tools might contribute to making students' self-study of writing more effective and convenient

4.1.4.4 EFL learners’ attitudes toward using AI-powered tools in learning writing skills

Table 4.10: EFL learners’ attitudes toward using AI-powered tools in learning writing skills

No Question items Mean Standard deviation Interpretation

1 I enjoy using AI-powered tools for developing writing skills 3.83 804 Agree

2 I feel comfortable using AI-powered tools for learning writing skills 3.81 820 Agree

I believe that AI-powered tools bring great benefits to me in learning writing

I believe that utilizing AI-powered tools for enhancing writing skills is a good idea

I believe that AI-powered tools have the potential for greater application in writing classes

With an average score of 3.85, the surveyed participants had a positive attitude toward the use of AI-powered tools in the process of learning writing skills Many of

40 them had a liking for using AI tools (M = 3.83, SD = 0.804) and had pleasant experiences in developing writing skills with these technologies (M = 3.81, SD = 0.820) They also felt that it was a good idea to use these tools to develop their writing (M = 3.85, SD 0.882)

As can be clearly seen, the highest mean score was attributed to the benefits of AI-powered tools, indicating that these technologies could fulfill students’ needs by addressing the difficulties they often encounter when learning writing (M = 3.89, SD 0.634) Also, many students believe that AI-powered tools have the potential for greater application in writing classes in the future (M = 3.89, SD = 0.896)

4.1.4.5 EFL learners’ behavioral intentions toward using AI-powered tools in learning writing skills

Table 4.11: EFL learners’ behavioral intentions toward using AI-powered tools in learning writing skills

No Question items Mean Standard deviation Interpretation

I intend to continue utilizing AI- powered tools in learning writing skills

2 I am willing to use AI-powered tools more often in learning writing skills 3.84 931 Agree

I intend to explore more features of

AI-powered tools to develop writing skills

I would recommend using AI- powered tools for writing development to others

I intend to apply what I have learned from using AI-powered tools in academic and business writing

It is noticeable from Table 4.11 that participants showed a high intention to utilize AI-powered tools in learning writing skills, with an average mean score of 3.85

Qualitative findings

There were five students participating in the interview They all have one thing in common: they have studied writing skills classes at the Banking Academy However, their level of use of AI-powered tools when learning writing skills varied from

“occasionally” to “very often." Students A, B, and C stated that they frequently use AI tools when learning writing skills Student D uses these tools very often, and student E occasionally uses them when necessary

4.2.1 AI-powered tools used by EFL learners in learning writing skills

When asked which tools they rated as most helpful in improving their writing skills, 2 students chose Grammarly, 2 students chose ChatGPT, and the remaining students found QuillBot most beneficial This is completely similar to the quantitative results, when these 3 tools received the most votes because of their usefulness

Student D explained why she considered ChatGPT the most effective tool for her writing development: “ChatGPT can understand quite exactly what I want to learn It responds very quickly and often analyzes each aspect of the problem I ask in a careful and detailed way.” This opinion was similar to that of student A

Student C shared: “One of the most effective AI tools that I have used is Grammarly This tool provides me with a range of features that help me improve my writing I find that Grammarly is easy to use and provides me with constructive feedback that I can use to enhance my skills.” Student E also had the same view

Along with Grammarly and ChatGPT, QuillBot was also highly appreciated by student B: “To me, I think QuillBot is the most effective tool because it provides diverse functions that suit my needs.”

From what the interviewees shared, an overall trend can be seen is that there is growing popularity and influence for the use of AI tools among EFL learners, who actively seek out and evaluate different options to identify the most effective one for their needs

4.2.2 EFL learners’ perceived ease of using AI-powered tools in learning writing skills

All 5 interview participants agreed with the simplicity and convenience of AI tools

In quantitative results, most respondents stated that accessing and using AI tools is very easy This was also approved by interviewee A: “I find it easy to access and use these tools because I almost always have my phone with me, and it is also connected to 4G.” In addition, interviewee E stated that “I can easily access AI tools as they are mostly available online.” Furthermore, “I think the interface of these tools is quite simple and easy to use, because as soon as I open a feature, it will display instructions on what to do next and where to click,” student A gave her view

It can be clearly seen that students are all familiar with and capable of using learning support technologies proficiently, including AI-powered tools This correlates with the current technological advancements and the widespread use of internet- connected smart devices

4.2.3 EFL learners’ perceived usefulness of AI-powered tools in learning writing skills

When asked about the usefulness of AI tools in helping them improve their writing skills, students all mentioned one to several benefits

Student A emphasized: “These tools often suggest many appropriate vocabulary and phrases for me, so I can expand my vocabulary and know how to use them flexibly in many different contexts.They also help me minimize mistakes and provide many types of sentence structures for me.”

This idea was supported by student D, who said:”They help me a lot in acquiring synonyms and antonyms I can take advantage of them when I'm stuck for words.” She also stated, “They help me check grammar, spelling errors and even provide better ideas for me I can make my sentences longer and better thanks to these AI tools.”

Student C shared that “AI tools provide effective and constructive suggestions to improve my writing, such as synonyms, paraphrases, sentence rewrites, and so on.”

Student B gave her opinion: “They help me a lot I can check grammar by Grammarly tool, consult ideas from Gemini or ChatGPT, or paraphrase by QuillBot.”

She also expressed that “I find using these tools makes my writing sound more professional and clear.”

However, student E gave a neutral response: “I only use them occasionally when

I need to check grammar and vocabulary errors, so I don't find it very helpful in improving my writing skills.”

Overall, the majority of students hold a favorable opinion of AI tools’ benefits However, one participant, due to infrequent use, held a neutral opinion Similarly, the quantitative results showed a large portion of students finding these tools helpful, while others provided neutral responses or did not find them useful in improving their writing skills

4.2.4 EFL learners’ attitudes toward using AI-powered tools in learning writing skills

Participant D shared her special love and positive attitude toward the use of AI tools as follows:

“Writing is a quite difficult skill that requires logic and the refinement of words and sentences This is a huge barrier for me That's why I’m really into using AI-powered tools because it helps me a lot in the process of learning writing while also saving me time and money.”

Participant C had a positive view toward the AI application in the future, as mentioned below:

“In the future, I believe that the application of AI technology in learning English writing will become even more popular, both inside and outside the classroom With the convenience and effectiveness that these tools bring, students may no longer need to wait for their teacher to check their writing Instead, they can use AI tools to receive instant feedback and improve their skills.”

Generally, most interviewees expressed a positive attitude toward using these AI tools Only student E had a neutral attitude This might be due to the differences in individual experiences and learning styles

4.2.5 EFL learners’ challenges in using AI-powered tools for learning writing skills

When asked about the difficulties of using AI tools when learning writing skills, student C shared that “I feel more dependent on AI tools to create ideas when using these tools in learning writing.” Likewise, student E stated, “Using AI tools too much may hinder my creativity.”

“I don't want to spend money on premium, but the free version doesn't have many functions to use,” student D expressed

Student C said: “Sometimes, the suggestions made by the AI tools may not be appropriate or may not align with my intentions.”

Discussion

Based on the analyzed quantitative and qualitative results, it becomes clear that EFL students are quite familiar with AI-powered tools since most students revealed that they sometimes or often used these tools to learn writing skills This is correlated with the significant development of AI technology and the availability of the internet and smart devices in the current 4.0 era The findings also revealed the most popular AI tools

48 are ChatGPT, Grammarly, and QuillBot, as these tools provide a variety of functions that could meet students’ needs and benefit their writing skills It is in line with Zulfa

(2023), who found that students often utilized Grammarly, QuillBot, and ChatGPT for assisting students in increasing their writing skills with automated corrections, feedback, and suggestions

Moreover, the research indicated that most students have positive perceptions of utilizing AI-powered tools for enhancing writing skills This can be clearly seen from the following indicators: perceived ease of use, perceived usefulness, attitude toward usage, and behavioral intention to use Regarding the perceived ease of use, the majority of students highly appreciated the simplicity, availability, and convenience of AI- powered tools They found it easy to access and use these technologies on different devices These findings are consistent with those of Utami et al (2023), who discovered that students could easily access these resources because they had acquired enough technical skills and were accustomed to various learning tools In addition, many students agreed that these AI-powered tools have friendly interfaces for users and easy- to-follow instructions They also believed that AI-powered tools provided direct and understandable responses This is quite similar to the research of Le (2023), in which students felt satisfied with the simple and friendly features of AI tools They could easily understand on-screen instructions as soon as they opened a feature (Utami et al., 2023)

In terms of their perceived usefulness, a large number of students considered AI- powered tools helpful for their writing development Despite not getting a strong agreement, it can still be seen that AI-powered tools could help enhance students’ writing skills and improve the quality of their writing Students highlighted these technologies’ advantages in expanding their vocabulary and minimizing grammatical errors These findings align with those of Le (2023), who found that students hold a favorable opinion of AI applications, which could benefit their linguistic competency and grammar knowledge Additionally, students believe that AI tools could enhance their abilities to use various sentence structures and diversify word choices They also agreed that these

49 tools could organize ideas and make their writing more fluent In the same way, the study by Al-Raimi et al (2024) indicated that the majority of participants confirmed that supportive and constructive suggestions provided by AI had significantly strengthened students’ competence in using appropriate words and sentences, contributing to their fluency in writing Moreover, students also reported that AI tools could become assistants for their self-studying The integration of AI technology has created favorable conditions for students to learn writing skills independently (Zulfa et al., 2023; Mohammad, 2024)

In addition, students embraced a positive attitude towards utilizing AI-powered tools for learning writing skills, as proven by an average high mean score They expressed their interest in using AI tools, as they believed in the convenience and usefulness of these technologies Utami et al (2023) also discovered similar results, in which participants were enthusiastic about taking advantage of these tools to overcome their difficulties in writing However, a small proportion of students were neutral on this statement because they felt that AI tools still could not fulfill their needs They wanted these functions to be enhanced and more diverse Despite not receiving full approval, we can see that the majority of participants believed in the potential for greater application of AI inside and outside the classroom This is a positive sign since students love using this technology and look forward to its future development

Furthermore, the results revealed that a significant number of students intended to continue using and exploring AI-powered tools to support their writing development This indicates that students recognized the benefits of AI-powered tools and were willing to make use of them Some even expressed eagerness to explore more features and AI tools to assist them in the process of learning writing Moreover, students also confirmed that they would apply the knowledge gained from AI tools to benefit their academic and business writing It can be noted that AI tools could offer students an easy-to-use approach to improve their writing skills and help narrow the gap between theoretical knowledge and real-world applications These findings are also supported by Le (2023),

50 who asserted that EFL learners would increase their interaction time with AI-powered tools as they recognized the benefits of these technologies and had positive feelings toward using them

However, the results also demonstrated some challenges encountered by students when utilizing AI-powered tools for learning writing skills While the majority of students had no difficulty accessing AI tools, a sizable proportion stated that they sometimes faced lag and connectivity issues In line with this, the research paper by Nabhan (2017) emphasized that these sophisticated technologies sometimes had technical malfunctions that might affect the process of taking advantage of them Additionally, students complained that the free versions of these tools were limited in functionality, but the premium ones were too expensive for them to afford They also agreed that AI tools sometimes provided inaccurate feedback or misleading information Besides, ethical concerns such as plagiarism and misinformation were another obstacle for students These findings go in line with the discoveries of Le (2023), indicating that

AI technologies still have drawbacks and highlighting the importance of students exercising their critical judgment when evaluating AI suggestions for improvement The study further revealed a troubling trend: a large number of students felt reliant on AI technologies when they employed them This issue was also addressed in the research conducted by Le (2023) Accordingly, the overuse of AI-based tools might hamper students' writing abilities Thus, this study implies that students should appropriately control the level of usage so that they can take advantage of these technologies to improve their writing ability It is good that students recognized the challenges, and some of them proposed several suggestions to overcome these challenges

CONCLUSION AND RECOMMENDATIONS

Conclusion

The study reveals that the majority of students exhibit a positive perception of utilizing AI-powered tools to learn writing skills, which can be clearly seen from the following indicators: perceived ease of use, perceived usefulness, attitude toward usage, and behavioral intention to use Regarding the perceived ease of use, the simplicity, availability, and convenience of AI-powered tools are highly appreciated by most students AI tools have user-friendly interfaces and provide understandable feedback, making it easy for students to access and use these tools on various devices In terms of perceived usefulness, most students have a high opinion of AI-powered tools as they can assist students in the process of learning and increase their writing competence Students highlight the benefits of AI tools in many aspects of writing, which may create favorable conditions for students to learn writing independently Furthermore, it is noticeable that most students embrace positive attitudes and behavioral intentions toward AI applications Despite some challenges in using AI-powered tools, such as technical malfunctions, ethical concerns, and technology dependence, they still show their interest in taking advantage of AI tools for writing and believe in the potential development of these technologies These things have motivated students to continue using AI tools and apply what they learn from AI to their academic and business writing.

Recommendations for learners and educators

Based on the analyzed results, some recommendations for learners and educators are proposed to increase the effectiveness of utilizing AI-powered tools in learning writing skills

First, it is necessary for learners to learn how to appropriately utilize AI tools to improve their writing ability and avoid dependence on them These tools should only be considered supplemental tools Students can actively use them to give suggestions during the writing process, then ask them to correct errors and provide feedback after completion

Second, AI-powered tools still have limitations, so they may provide inaccurate or inappropriate responses Thus, learners should evaluate suggestions before accepting them Furthermore, learners can also combine the use of multiple AI tools to check whether the answers provided by the tools are reasonable

Third, it is advisable that learners take notes of new words, phrases, and sentences that the AI tools suggest This can help learners easily remember those words, phrases, and sentences in specific contexts and apply them in future writing

Finally, learners should set their learning goals and identify the specific writing skills that they want to improve (for example, grammar, spelling, and style) Then, they can learn about and choose AI tools that align with their goals

First, educators should provide guidance or advice to help students address the difficulties and optimize the use of AI-powered tools to improve their writing skills In addition, educators could also encourage learners to share experiences and support each other in using these tools as assistants for independent learning

Second, as the development and use of AI tools are inevitable, educators may consider developing appropriate teaching methods that leverage AI tools alongside traditional approaches This may involve creating blended learning environments that combine AI-powered feedback with teacher guidance and peer interaction

Third, it is important to evaluate learners' writing skills on a regular basis to track their progress and identify areas in need of improvement This will allow educators to adapt their teaching strategies and provide more support for learners as needed

Overall, there should be more cooperation between educators and learners to fully harness the potential of AI-powered tools for enhancing writing skills.

Recommendations for further studies

Generally, the study reveals that the majority of EFL students have positive perceptions of utilizing AI-powered tools in learning writing skills However, it primarily focuses on exploring students' thoughts and perceptions and does not explore the results students achieved after applying AI tools to learning writing skills Therefore, future research can address this gap by investigating students' writing proficiency as well as their achievements by leveraging AI tools In addition, the study only explores EFL learners' perceptions at a specific point in time Longitudinal studies would be needed to understand how perceptions and usage patterns change over time This might expand the field of research and give a more comprehensive understanding of using AI to learn writing skills

Abdulwahed Ahmed Ismail , S (2011) Exploring Students’ Perceptions of ESL Writing

English Language Teaching, 4(2), 73 https://doi.org/10.5539/elt.v4n2p73

Abushihab, I (2014) An Analysis of Grammatical Errors in Writing Made by Turkish

Learners of English as a Foreign Language International Journal of Linguistics,

Al-Raimi, M., Badri Abdulhakim Mudhsh, Yasir Al-Yafaei, & Salim Al-Maashani

(2024) Utilizing artificial intelligence tools for improving writing skills: Exploring Omani EFL learners’ perspectives Forum for Linguistic Studies, 6(2), 1177–1177 https://doi.org/10.59400/fls.v6i2.1177

Asep, S (2014) The Challenges in Teaching Writing Skill at Junior High School:

Asnas, S A M., Mustofa, M., & Ubaidillah, M F (2022) Investigating Academic

Writing in EFL Contexts: Students’ Voices on Complexities and Coping Strategies Vision: Journal for Language and Foreign Language Learning, 11(2), 139–162 https://doi.org/10.21580/vjv12i116326

Astika, G G (1993) Analytical Assessments of Foreign Students’ Writing RELC

Athiyaman, A (2002) Internet users’ intention to purchase air travel online: an empirical investigation Marketing Intelligence & Planning, 20(4), 234–242 https://doi.org/10.1108/02634500210431630

Bellman, R (1978) An Introduction to Artificial Intelligence Boyd & Fraser Publishing

Brown, H D (2004) Language Assessment: Principles and Classroom Practices New

Bùi Văn Hiếu, Hoàng Mai Huy, & Hằng, T (2022) ÁP DỤNG ỨNG DỤNG

QUILLBOT NHẰM CẢI THIỆN KỸ NĂNG TÁI DIỄN ĐẠT TRONG VIẾT BÀI LUẬN HỌC THUẬT ĐỐI VỚI SINH VIÊN CHUYÊN NGÀNH NGÔN NGỮ ANH, TRƯỜNG NGOẠI NGỮ, ĐẠI HỌC THÁI NGUYÊN TNU Journal of Science and Technology, 227(13), 116–124 https://doi.org/10.34238/tnu- jst.6717

David, F D., Bagozzi, R P., & Warshaw, P R (1989) User acceptance of computer technology: A comparison of two theoretical models Management Science, 35(8), 982–1003

Ellis, R (2006) Current Issues in the Teaching of Grammar: An SLA Perspective

University of Auckland Auckland, New Zealand https://ojcsteve.tripod.com/sitebuildercontent/sitebuilderfiles/ellis_grammar.pdf

Hans, A., & Hans, E (2017) Role of Grammar in Communication -Writing Skills

International Journal of English Language, Literature and Humanities, 5(1) https://doi.org/10.24113/ijellh.v5.issue1.42

Hashim, K., Al-Sharqi, L., & Kutbi, I (2016) Perceptions of Social Media Impact on

Social Behavior of Students International Journal of Virtual Communities and

Social Networking, 8(2), 1–11 https://doi.org/10.4018/ijvcsn.2016040101

Kim, B.-J (2019) A Basic Study on the Selection of Educational Contents for

Cultivating Global Business Communication Competence Journal of Digital Convergence, 17(2), 91–100 https://doi.org/10.14400/JDC.2019.17.2.091

Le, P T N (2023) Students’ Perceptions of the AI Technology Application in English

Writing Classes Proceedings of the AsiaCALL International Conference, 4, 45–

Lien, T., Thai Cong Dan, & Han, B (2023) AN INVESTIGATION INTO

ORGANIZATION ERRORS IN EFL LEARNERS’ PARAGRAPH WRITING:

A CASE IN A FOREIGN LANGUAGE CENTER IN CAN THO CITY, VIETNAM European Journal of Applied Linguistics Studies, 6(1) https://doi.org/10.46827/ejals.v6i1.420

Maamuujav, U (2021) Examining lexical features and academic vocabulary use in adolescent L2 students’ text-based analytical essays Assessing Writing, 49 https://doi.org/10.1016/j.asw.2021.100540

Mathieson, K (1991) Predicting user intentions: Comparing the technology acceptance model with the theory of planned behavior Information Systems Research, 2(3), 173–191 https://doi.org/10.1287/isre.2.3.173

Maulidina, P., & Wibowo, H (2022) THE USE OF GRAMMARLY TOOLS TO

ENRICH STUDENT’S WRITING ABILITY Lingua, 18(2), 179–189 https://doi.org/10.34005/lingua.v18i2.2246

Messum, D., Wilkes, L., Peters, K., & Jackson, D (2017) Content analysis of vacancy advertisements for employability skills: Challenges and opportunities for informing curriculum development Journal of Teaching and Learning for

Graduate Employability, 7(1), 72 https://doi.org/10.21153/jtlge2016vol7no1art582

Michặl Lévy, Yeh, A., Hawkes, C H., Lechner‐Scott, J., & Giovannoni, G (2023) Is using ChatGPT to help write papers and grants is useful, and ethical? Multiple Sclerosis and Related Disorders, 76, 104873–104873 https://doi.org/10.1016/j.msard.2023.104873

Mohammad, T (2024) Challenging Traditional EFL Writing Classroom Using Al

Mediated Tool: A Paradigm Shift World Journal of English Language, 14(2),

Moore, T., Morton, J., Hall, D., & Wallis, C (2015) Literacy practices in the professional workplace: Implications for the IELTS reading and writing tests In

57 www.academia.edu https://www.academia.edu/27217846/Literacy_practices_in_the_professional_w orkplace_Implications_for_the_IELTS_reading_and_writing_tests

Nhi, T T Q., & Dao, N T T (2023) AN INVESTIGATION INTO ENGLISH-

MAJORED UNDERGRADUATES’ DIFFICULTIES IN ENGLISH ESSAY WRITING TNU Journal of Science and Technology, 228(12), 458–465 https://doi.org/10.34238/tnu-jst.8454

Nozaki, K., Matsuzaki Wuttipong, Heman Hiroyuki, Kawachi, A., & Kawachi, Y

(2023) A Study of Student Learning Attitudes and Motivation in Technological Processes in Learning World Psychology, 1(3), 71–88 https://doi.org/10.55849/wp.v1i3.380

Nunan, D (1999) Second language teaching & learning Boston, Mass : Heinle &

Heinle Publishers https://archive.org/details/secondlanguagete0000nuna

Nunnally, J C (1978) Psychometric Theory (2nd ed.) New York: McGraw-Hill

Nunnally, J C., & Bernstein, I H (1994) Psychometric theory McGraw-Hill, Inc., New

Persky, H R., Daane, M C., & Jin, Y (2002) The Nation’s Report Card: Writing 2002

U.S Department of Education Institute of Education Sciences National Center for Education Statistics https://nces.ed.gov/nationsreportcard/pdf/main2002/2003529.pdf

Peterson, R A (1994) A Meta-Analysis of Cronbach’s Coefficient Alpha Journal of

Consumer Research, 21, 381-391 http://dx.doi.org/10.1086/209405

Phuong, D T M (2023, July 26) Mô hình MTAM: Lý thuyết tiềm năng về hành vi của người dùng công nghệ Economy & Forecast Review

Pinto, L, L., & Stein, L, L (2015) The Theoretical Basis of the Technique of Context

Reinstatement in the Cognitive Interview Avances En Psicología Latinoamericana, 33(2), 283–299 https://doi.org/10.12804/apl33.02.2015.08

Rich, E., & Knight, K (1991) Artificial Intelligence McGraw-Hill

Santos, T (1988) Professors’ Reactions to the Academic Writing of Nonnative-

Speaking Students TESOL Quarterly, 22(1), 69 https://doi.org/10.2307/3587062

Saud, W (2015) Cohesion in the Descriptive Writing of EFL Undergraduates King

Khalid University, Abha, Saudi Arabia https://www.researchgate.net/publication/332572551_Cohesion_in_the_Descrip tive_Writing_of_EFL_Undergraduates

Schmidgall, J., & Powers, D E (2020a) Predicting communicative effectiveness in the international workplace: Support for TOEIC® Speaking test scores from linguistic laypersons Language Testing, 38(2) https://doi.org/10.1177/0265532220941803

Schmidgall, J., & Powers, D E (2020b) TOEIC® Writing test scores as indicators of the functional adequacy of writing in the international workplace: Evaluation by linguistic laypersons Assessing Writing, 46 https://doi.org/10.1016/j.asw.2020.100492

Slater, S F (1995) Issues in conducting marketing strategy research Journal of

Strategic Marketing, 3(4), 257–270 https://doi.org/10.1080/09652549500000016

Smith, R A., White-McNeil, A., & Ali, F (2020) Students’ perceptions and behavior toward on-campus foodservice operations International Hospitality Review, ahead-of-print(ahead-of-print) https://doi.org/10.1108/ihr-06-2019-0010

Su, Z., Miao, L., & Man, J (2019) Corrigendum: Artificial Intelligence Promotes the

Evolution of English Writing Evaluation Model IOP Conference Series: Materials Science and Engineering, 646 https://doi.org/10.1088/1757-

Teo, T (2018) Students and Teachers’ Intention to Use Technology: Assessing Their

Measurement Equivalence and Structural Invariance Journal of Educational

Computing Research, 57(1), 201–225 https://doi.org/10.1177/0735633117749430

Utami, S P T., Andayani, A., Winarni, R., & Sumarwati, S (2023) Utilization of artificial intelligence technology in an academic writing class: How do Indonesian students perceive? Contemporary Educational Technology, 15(4), ep450 https://doi.org/10.30935/cedtech/13419

Wang, D (2021) Changes and Challenges: a Study on the Application of Artificial

Intelligence Technology in College English Teaching Association for Computing

Zulfa, S., Dewi, R S., Hidayat, D N., Hamid, F., & Defianty, M (2023) The Use of AI and Technology Tools in Developing Students’ English Academic Writing Skills

International Conference on Education, 1(1), 47–63 https://jurnalfaktarbiyah.iainkediri.ac.id/index.php/proceedings/article/view/181

THE QUESTIONNAIRE

Thank you for your time in participating in this questionnaire!

I am currently conducting an investigation into EFL learners' perceptions of utilizing AI- powered tools in learning writing skills Your participation will be greatly appreciated and will contribute significantly to my research I pledge that all information you provide will be kept confidential and will only be used for research purposes

Once again, I would like to thank you for your help and wish you a good day!

(Cảm ơn bạn đã dành thời gian cho cuộc khảo sát này!

Hiện tại mình đang nghiên cứu về nhận thức của người học tiếng Anh về việc sử dụng các công cụ AI trong học kỹ năng viết Sự tham gia của bạn sẽ giúp mình rất nhiều trong quá trình làm bài Mình xin cam kết mọi thông tin mà bạn cung cấp sẽ được bảo mật và chỉ phục vụ cho mục đích nghiên cứu

Một lần nữa, mình xin cảm ơn sự giúp đỡ từ bạn và chúc bạn có một ngày tốt lành!) Section 1: Demographic and background questions (Thông tin cơ bản)

1 What is your gender? (Giới tính của bạn?) o Male (Nam) o Female (Nữ) o Prefer not to say (Không muốn tiết lộ)

2 Which year are you in? (Bạn là sinh viên năm mấy?) o Year 1 (Năm 1) o Year 2 (Năm 2) o Year 3 (Năm 3) o Year 4 (Năm 4)

3 Please tick the AI-powered tools that you have used in learning writing skills

(Hãy đánh dấu các công cụ AI bạn đã từng sử dụng khi học kỹ năng viết) – Chọn tất cả các đáp án phù hợp

4 Among the AI-powered tools you have used, which one do you find most effective? (Trong số các công cụ AI bạn đã sử dụng, bạn thấy công cụ nào là hiệu quả nhất?) o Quillbot o Grammarly o Paraphraser o ChatGPT o Gemini (Google bard) o ProWritingAid o Essaygrader o Upscore o Others (Khác)

5 How often do you use AI-powered tools when you learn writing?

(Bạn có thường xuyên dùng các công cụ AI khi học kỹ năng viết không?)

62 o Rarely (Hiếm khi) o Sometimes (Thi thoảng) o Often (Thường xuyên) o Usually (Rất thường xuyên)

Section 2: EFL learners’ perceptions on utilizing AI-powered tools in learning writing skills (Cảm nhận của người học tiếng Anh về việc sử dụng các công cụ AI để học kỹ năng viết)

Please tick the number that best describes your choice (1= Strongly disagree, 2 Disagree, 3= Neutral, 4= Agree, 5= Strongly agree)

Hãy đánh dấu vào con số mô tả đúng nhất sự lựa chọn của bạn (1= Rất không đồng ý, 2= Không đồng ý, 3= Trung lập, 4= Đồng ý, 5= Rất đồng ý)

Part 1 Perceived ease of using AI-powered tools in learning writing skills

(Mức độ dễ dàng khi sử dụng của các công cụ AI để học kỹ năng viết.)

1 AI-powered tools are easy to use (Các công cụ AI dễ sử dụng.)

2 AI-powered tools are easily accessible (Các công cụ AI dễ truy cập.)

3 AI-powered tools can easily be used on various devices (Các công cụ AI có thể dễ dàng sử dụng trên nhiều loại thiết bị.)

4 AI-powered tools have user-friendly interfaces (Các công cụ AI có giao diện phù hợp và thân thiện với người dùng.)

5 AI-powered tools have easy-to-follow instructions (Các công cụ

AI có hướng dẫn dễ hiểu và dễ dàng thực hiện theo.)

6 AI-powered tools provide direct and understandable responses

(Các công cụ AI cung cấp phản hồi trực tiếp và dễ hiểu.)

Part 2 Perceived usefulness of using AI-powered tools in learning writing skills (Mức độ hữu ích của việc sử dụng các công cụ AI trong việc học kỹ năng viết.)

1 AI-powered tools help me enhance my writing skills (Các công cụ AI giúp tôi nâng cao kỹ năng viết.)

2 AI-powered tools improve the quality of my writing (Các công cụ AI giúp cải thiện chất lượng bài viết của tôi.)

3 AI-powered tools help me in achieving my learning objectives

(Các công cụ AI giúp tôi đạt được mục tiêu học tập.)

4 AI-powered tools increase my ability to organize the ideas (Các công cụ AI giúp tôi nâng cao khả năng sắp xếp và trình bày ý tưởng.)

5 AI-powered tools can enrich my vocabulary (Các công cụ AI giúp tôi tăng vốn từ vựng.)

6 AI-powered tools offer me the ability to diversify my word choices (Các công cụ AI giúp tôi có khả năng đa dạng hóa trong việc lựa chọn từ.)

7 AI-powered tools help me understand the use of transition words

(Các công cụ AI giúp tôi hiểu cách sử dụng các từ nối, từ ngữ liên kết.)

9 AI-powered tools improve my ability to use various sentence patterns (Các công cụ AI giúp nâng cao khả năng sử dụng nhiều kiểu câu khác nhau trong bài viết.)

8 AI-powered tools help to reinforce my grammar (Các công cụ AI giúp tôi củng cố ngữ pháp.)

10 AI-powered tools increase my fluency in writing (Các công cụ

AI giúp tôi viết trôi chảy hơn.)

11 AI-powered tools keep me active and motivated in learning writing (Các công cụ AI khiến tôi chủ động và có động lực hơn trong học viết.)

12 AI-powered tools give me the flexibility to learn writing skills anytime and anywhere (Các công cụ AI giúp tôi có thể linh hoạt học kỹ năng viết mọi lúc mọi nơi.)

Part 3 Attitudes towards the use of AI-powered tools in learning writing skills (Thái độ đối với việc sử dụng các công cụ AI trong việc học kỹ năng viết.)

1 I enjoy using AI-powered tools for developing writing skills (Tôi thích sử dụng các công cụ AI để nâng cao kỹ năng viết.)

2 I feel comfortable using AI-powered tools for learning writing skills (Tôi cảm thấy thoải mái khi sử dụng các công cụ AI trong việc học kỹ năng viết.)

3 I believe that AI-powered tools bring great benefits to me in learning writing (Tôi tin rằng các công cụ AI giúp ích cho tôi rất nhiều khi học viết.)

4 I believe that utilizing AI-powered tools for enhancing writing skills is a good idea (Tôi tin rằng việc sử dụng các công cụ AI để phát triển kỹ năng viết là một ý tưởng tốt.)

5 I believe that AI-powered tools have the potential for greater application in writing classes (Tôi tin rằng các công cụ AI có tiềm năng lớn để được ứng dụng rộng rãi hơn trong các lớp học viết.)

Part 4 Behavioral intentions in using AI-powered tools in learning writing skills (Ý định hành vi trong việc sử dụng các công cụ AI để học kỹ năng viết.)

1 I intend to continue utilizing AI-powered tools in learning writing skills (Tôi dự định tiếp tục sử dụng các công cụ AI để học kỹ năng viết.)

2 I am willing to use AI-powered tools more often in learning writing skills (Tôi sẵn sàng sử dụng các công cụ AI thường xuyên hơn trong việc học kỹ năng viết.)

3 I intend to explore more features of AI-powered tools to develop writing skills (Tôi dự định khám phá thêm nhiều tính năng khác của các công cụ AI để học kỹ năng viết.)

4 I would recommend using AI-powered tools for writing development to others (Tôi sẽ khuyên người khác nên sử dụng các công cụ AI để học kỹ năng viết.)

5 I intend to apply what I have learned from using AI-powered tools in academic and business writing (Tôi dự định sẽ áp dụng những gì tôi đã học được từ các công cụ AI vào học tập và công việc.)

Part 5 Challenges in using AI-powered tools for developing writing skills

(Thách thức khi sử dụng công cụ AI để phát triển kỹ năng viết.)

1 I encounter technical issues such as slow loading, server downtime, etc while using AI-powered tools for learning writing skills (Tôi gặp phải các vấn đề kỹ thuật như tốc độ tải chậm, máy chủ ngừng hoạt động, v.v khi sử dụng các công cụ AI để học kỹ năng viết.)

2 It costs a great deal to use advanced functions of AI powered tools (Việc sử dụng các chức năng nâng cao của các công cụ AI tốn nhiều chi phí.)

3 AI-powered tools can provide misleading or inaccurate feedback

(Các công cụ AI có thể cung cấp phản hồi gây hiểu lầm hoặc không chính xác.)

4 The use of AI-powered tools can raise ethical concern, such as plagiarism, misinformation, etc (Việc sử dụng các công cụ AI có thể gây ra các vấn đề về đạo đức, chẳng hạn như đạo văn, thông tin sai lệch, v.v.)

INTERVIEW QUESTIONS

1 Do you find AI-powered tools easy to access and use?

2 In what ways do you think these tools help you when learning writing skills?

3 Among the AI tools you have used, which one do you find most effective? Can you say why it worked best for you?

4 What challenges have you encountered when utilizing AI-powered tools to learn writing skills?

5 Do you think AI tools should be applied more inside and outside the classroom? Why?

6 What would you suggest or recommend to optimize the effectiveness of using AI- powered tools in learning writing skills?