Generative Artificial Intelligence and ChatGPT in Language Learning:

EFL Students' Perceptions of Technology Acceptance

Anh Vo and Huong Nguyen

University of Danang, Vietnam

Abstract
The study conducted in the centre of Vietnam in 2023 aims to explore Editors

English-majored students’ perceptions towards the application of ChatGPT Section: Special Issues
in developing students’ English language. The research was grounded on Editor in Chief: Dr Joseph Crawford
Guest Editor: Dr Marios Kremantzis
the Technology Acceptance Model developed by Davis (1998) and further
revised by Shoufan (2023). The quantitative approach was selected for the Publication
study with the questionnaire as the instrument. 369 English-majored Received: 10 December 2023
students who have experienced using ChatGPT in language learning were Revision: 2 March 2024
selected for the study. The findings reveal that although students found it Accepted: 13 March 2024
Published: 30 April 2024
easy to use ChatGPT as well as perceived it as a useful tool for their
language learning, they had neutral opinions on its usefulness. Furthermore, Copyright

students had a high level of attitudes towards the use of ChatGPT in © by the authors, in its year of first
language learning while the negative attitudes towards it were found to be at publication. This publication is an
open access publication under the
a low level. While genders were recognised not to have impacts on students’
Creative Commons Attribution CC
perceptions, class levels had significant effects on those of students It is BY-ND 4.0 license.
highly recommended that teaching strategies, careful monitoring, and a
balance between the use of ChatGPT and human interaction should be
under consideration for an effective adoption of ChatGPT in language
learning and teaching.

Citation
Vo, A., & Nguyen, H. (2024). Generative Artificial Intelligence and ChatGPT in Language Learning: EFL Students' Perceptions of
Technology Acceptance. Journal of University Teaching and Learning Practice, 21(6).
 Introduction
Artificial intelligence-generated content (AIGC) is the latest created technology that allows users
to request personalized requirements to generate certain content (Wu et al., 2023). Based on
the development of Artificial Intelligence Content Generation (AIGC), various products were
developed by the world's leading technology companies. Among them is OpenAI with ChatGPT,
a tool that can respond to human requests in natural languages. Since the appearance of
ChatGPT in late 2022, the application has boosted great changes in many fields including
language education (Sami et al., 2023). Teachers and students have been trying to make use of
ChatGPT to teach and learn foreign languages.
Many studies have been conducted on the use of ChatGPT to develop students’ language
proficiency (Baskara, 2023; Kohnke, 2023; Kasneci et al., 2023; Kohnke et al., 2023; Ho, 2024;
Hong, 2023). The studies found that ChatGPT offers an interactive, engaging and productive
language learning environment for learners to develop language skills and knowledge despite
some potential challenges to academic integrity. While these studies introduce preliminary
meaningful discussions on ChatGPT’s impact on language learning, their findings about the
benefits and challenges of ChatGPT are theoretically based, or they still lack empirical evidence
or methodological rigor. Moreover, students’ perceptions on ChatGPT, which can influence the
effective use of ChatGPT and its improvement (Shoufan, 2023), have not been learned enough.
Therefore, there is a need to have more knowledge on using ChatGPT to learn a foreign
language from students’ perspectives.
The study was carried out to explore students’ perceptions towards the use of ChatGPT to
develop English language skills in the Vietnamese context where technology advancements
have been updated a little slowly as compared to developed nations. It aims to address the
research question: “What are students’ perceptions towards the use of ChatGPT in language
learning?” Two hypotheses were tested in the study.
(1) Students’ perceptions towards the use of ChatGPT are affected by their genders.
(2) Students’ perceptions towards the use of ChatGPT are affected by class level.
The paper presents a review of ChatGPT, and its pros and cons before introducing the
Technology Acceptance Model (TAM) which forms the theoretical framework to investigate
students’ perceptions of ChatGPT in language learning. Quantitative and qualitative data are
used to illustrate the research findings, which are expected to be useful for researchers and
practitioners to have a more efficient use of ChatGPT to teach and learn foreign languages. The
study provides an insight into the students’ perceptions of ChatGPT use in language learning,
which will give recommendations for teachers, learners and policymakers on how ChatGPT can
be used to achieve its effectiveness and adaptability in English language teaching settings in
general and the Vietnamese context in particular. The research also contributes to
advancements in the use of technology in language learning as well as the broader field of
artificial intelligence and natural language processing which presents a vast ground for future
research.

Literature Review
ChatGPT and its Merits in Education and Language Learning
The last few years have seen the rise of ChatGPT, a noteworthy milestone in the field of natural
language processing. ChatGPT is an advanced, complex and powerful, generative language
model that can comprehend and produce natural language (Kasneci et al., 2023; Lund & Wang,
2023; Mhlanga, 2023; Pavlik, 2023). It can produce thoughtful and contextually relevant
answers to a question by synthesizing a vast amount of text data. ChatGPT is a versatile
application that can perform human-like specific tasks including providing explanations,
answering factual questions, making creative suggestions, and engaging in both formal and
informal conversations (Lund & Wang, 2023).
ChatGPT has been widely applied in many fields including education. In education, it can serve
as a virtual tutor, offering explanations and guidance on a range of subjects. Rahman and
Watanobe (2023) maintain that the ChatGPT model can be a professional tutor with whom
learners can argue about the given explanations, solutions, and other suggestions. Therefore,
learners receive interactive support from ChatGPT. Jeon and Lee (2023) who investigate the
complementary roles of both ChatGPT and teachers in education mentioned that ChatGPT took
four roles as material supplier, assessor, interlocutor and assistant. Sallam (2023) indicates
some advantages of ChatGPT in healthcare education including developing academic research
papers, utilizing healthcare research, personalizing learning, and boosting critical thinking.
ChatGPT can facilitate the teaching and learning process and develop students’ varied skills
including reading, writing, information analysis, critical thinking, problem-solving, and research
(Rahman & Watanobe, 2023; Shoufan, 2023). In digital writing, Johinke at al., (2023) claim that
technologies including ChatGPT “can be harnessed to reflect the values of education –
openness, individual autonomy, and the power of knowledge” (p. 2).
In language learning and teaching, ChatGPT can be a valuable tool in language learning,
offering several advantages to both teachers and students. Many authors including Kohnke et
al. (2023) and Kohnke (2023) demonstrated how the teacher could use ChatGPT to support
their teaching by creating fun and engaging activities and developing teaching materials,
translating them into the students’ first language to motivate students’ interest in learning the
topic and promote communication in the classroom. Ali et al. (2023) proposed a model in which
teacher could perform a flexible role in creating an enjoyable and interactive learning
environment.
For learners, ChatGPT has been increasingly become popular (Hong, 2023). ChatGPT offers an
interactive, engaging and productive language learning experience (Baskara, 2023; Kohnke,
2023; Kohnke et al., 2023; Kasneci et al., 2023; Shaikh et al., 2023; Shoufan, 2023). Students
can initiate conversations in any topic and receive instant human-like responses from ChatGPT
regarding a variety of aspects of language learning including language practice, language skills
as well as communication skills. Students can develop their mastery of vocabulary and grammar
by requesting explanations of forms, meanings, and uses of a word, a phrase or a grammatical
point. Furthermore, ChatGPT can provide authentic written materials for students to improve
their reading and writing skills (Baidoo-Anu & Owusu Ansah, 2023; George & George, 2023). Ali
et al. (2023) showed that ChatGPT generally motivated learners to develop their reading and
writing skills, while it had a neutral impact on their listening and speaking skills. In the meantime,
Kohnke et al., (2023) stated that students who engage in regular voice interactions with
ChatGPT have improved their pronunciation and listening comprehension. ChatGPT can also
translate materials or texts into students’ mother tongues, which can provide more language
practice opportunities (Ali et al., 2023). As ChatGPT can simulate real-life conversations,
students can engage in conversations with ChatGPT on diverse topics, which requires students’
fluency and use of appropriate language. The interaction can enhance students’ communication
skills (Baskara, 2023) in diverse conversations with ChatGPT. Particularly, ChatGPT can offer
answers related to cultural aspects, including idioms, customs, or social norms. This promotes
students’ understanding of cultural issues which are necessary to their communicative language
ability. Generally, students can benefit from ChatGPT’s feedback on their written language
practice exercises to improve their language competence.
ChatGPT provides ample opportunities for student’s self-study outside class, focusing on
personalized learning experiences with instant feedback. ChatGPT can meet the learning needs
of students with varied interests, learning pace, and academic performance (Ali et al., 2023;
Hoang et al., 2023; Mohammed et al., 2023; Kohnke et al., 2023). ChatGPT can identify
students’ weaknesses in language competence and help them achieve their learning goals
(Baskara, 2023) by providing immediate, real-time feedback on students’ language use. It can
point out students’ mistakes and suggest alternatives or improvements (Ali et al., 2023;
Mohammed, 2023; Kohnke et al., 2023; Tlili et al., 2023). In this way, ChatGPT helps scaffold
students’ learning and boosts their performance (Ali et al., 2023). According to Hong (2023),
while human feedback may not be efficient due to time-consuming work, ChatGPT is
immediate, interactive, and effective. Therefore, students may become more motivated in their
learning (Ali et al., 2023) and become more independent in their own learning (Ali et al, 2023;
Fauzi et al., 2023; Kohnke et al., 2023).
Challenges of ChatGPT
While acknowledging the benefits of ChatGPT in optimizing the teaching and learning process,
some authors refer to ChatGPT as a disruption of academic integrity (Hong, 2023). Students’
overreliance on ChatGPT’s responses causes great concerns about the originality of their own
work (Cassidy, 2023; Pavlik, 2023, Shoufan, 2023), which raises ethical questions about
plagiarism generated due to using ChatGPT in education. To reduce this problem, many
institutions bring back in-class oral and written examinations and apply relevant rules (Rudolph
et al., 2023). Furthermore, although ChatGPT’s responses normally demonstrate critical
thinking, clarity, precision, coherence, relevance, and depth, they can be inaccurate, biased,
and inappropriate (Shaikh et al., 2023). Some authors attributed this to ChatGPT being an
unsupervised learning model or the presence of stereotypes in the vast corpora of texts
(Baskara, 2023). ChatGPT may not have the cultural awareness to understand human language
as well as abstract and complex ideas (Baskara, 2023). ChatGPT’s answers are also not
properly cited with references (Kohnke et al, 2023). As ChatGPT presents both advantages and
challenges, teachers and students need to learn how to effectively monitor and evaluate the use
of ChatGPT (Baskara, 2023) or use it ethically and competently. Kohnke et al (2023) insist on
the need for digital competencies, instructions or guidelines for using ChatGPT in teaching and
assessment, enhanced awareness of the risks and drawbacks of ChatGPT.
Review of Previous Studies
The use of ChatGPT has been mainly explored from educators’ viewpoints (Hong, 2023;
Shoufan, 2023) and few studies examine students’ reactions or perceptions to see whether
there is any resistance by students so that further improvement to ChatGPT, and its uses can
be made. Shoufan (2023) highlighted the significance of investigating students’ experience in
using ChatGPT and their perceptions of it. The author explained that students’ perceptions can
greatly impact their motivation, engagement, and performance. While positive attitudes can
promote students’ eagerness to learn and their academic achievement, negative viewpoints of
ChatGPT can lead to disengagement, demotivation, and little chance of academic success.
However, students’ perceptions of ChatGPT have not yet been fully addressed by previous
relevant studies (Shoufan, 2023).
It is worthwhile to acknowledge some initial attempts to elicit students’ ideas and comments on
their use of ChatGPT in education. Rahman et al. (2023) carried out a survey with students and
teachers to explore the impact of ChatGPT in a programming course. The study found students
had favorable attitudes to ChatGPT which is learner-friendly and beneficial to personalized
learning, assessment, and evaluation. Even though the study presented varied information
about the benefits as well as threats of ChatGPT, its findings about students’ perceptions can
be challenged by the lack of coherent data collection and analysis. Furthermore, programming
students’ opinions cannot be representative for language students.
Shoufan (2023) explored how students in a computer engineering program perceived ChatGPT
and assessed its potential benefits and challenges. The study developed a 27-item
questionnaire based on the themes and codes identified from students’ responses in a previous
study. Students demonstrated divided opinions on the demerits of ChatGPT on academic
integrity and threat to their studies, jobs, and life. The author concluded that ChatGPT can and
should be used in learning while its limitations should be well recognized and addressed.
Shoufan’s study contributes to research in exploring students’ perceptions of education in
general by proposing varied themes and topics in the use of ChatGPT.
Ali et al. (2023) who acknowledged a big gap in the literature review about dimensions in the
use of ChatGPT in English language learning explored the impact of ChatGPT on learning
motivation among teachers and students of English as a foreign language (EFL) in Saudi Arabia
through quantitative research design. The study described useful dimensions including
enhancing language skills and scaffolding learning. However, it only presented the perceptions
of 20 university professors who have diverse backgrounds, motivations, and expectations from
the ChatGPT in its early phase. The study did not depend on data collected from students. In
fact, Ali et al. (2023) explain that research on the ChatGPT is still in its infancy.
Shaikh et al (2023) which is among very few exploratory studies on the actual use of ChatGPT
for language learning uses a questionnaire to ask participants of diverse backgrounds about
their feedback on usefulness and effectiveness of ChatGPT after doing a series of tasks related
to English language learning with ChatGPT in formal contexts such as conversation, writing,
grammar, and vocabulary. The results demonstrated ChatGPT's capability for producing
accurate and cogent responses on its own. ChatGPT's conversation interface made it possible
to have interactive conversations and write naturally. Because just a small subset of university
students majoring in computer science performed the study, their perspectives on the subject
may differ to some extent from those of students majoring in English as a second language, for
instance.
Within the context of Vietnam, AI technologies including Chatbot and ChatGPT have become an
interest to language educators and researchers. Tran and Tran (2023) investigated both
teachers’ and students’ opinions about the role of ChatGPT in enhancing learners’ critical
thinking, digital literacy, and digital environment navigation. From the perspectives of the
participants, the authors revealed that ChatGPT has a role in the enhancement of critical digital
literacies. However, participant bias and the small scale of the study, as admitted by the
authors, can be a flaw. Ho (2024) presented a comprehensive study on non-major English
learners’ perceptions and behaviours in using ChatGPT in language learning. The study found
favourable attitudes among students who could find ChatGPT to promote their specialised
vocabulary learning. The students also become more confident in their learning, using ChatGPT
as a translation machine, grammar checker, and paraphrasing tool. Even though these studies
offer some meaningful findings about students’ perceptions in the use of ChatGPT, future
research in Vietnam and other contexts can explore the perceptions of varied types of students,
particularly English majors in many other related aspects of using ChatGPT in order to give a
more profound statement of the efficacy of ChatGPT in language learning.
There is a vast ground for investigating the use of ChatGPT in language learning at the tertiary
level in general and learners’ perceptions of it. Hong (2023) suggested that foreign language
educators need to discuss the merits, demerits as well as risks of Chat GPT with students rather
than avoid the topics. The author calls for further research into learners’ acceptance of the new
technology as language learners may perceive ChatGPT in different ways. In fact, O’Dea and
O’Dea (2023) who reviewed recent literature about the use of AI technologies in higher
education including tertiary language education found that there is a low acceptance rate of AI
in higher education due to obstacles at different levels. They particularly recommend the
investigation of students’ perceptions on AI technology to support language learning.
Technology Acceptance Model
The technology acceptance model (TAM) was first proposed by Davis (1989) who investigates
the individual technology acceptance behavior in various information systems constructs. Based
on the theory of social psychology, the model is intended to study the relationship among
cognitive and affective factors and user’s application of technology. Technology acceptance
involves three stages, starting with external factors triggering cognitive responses (i.e.,
perceived usefulness and perceived ease of use). Then, an effective response (or the user’s
intention or attitude toward using technology) is formed, which impacts the user’s behaviour.
Perceived usefulness and perceived ease of use are two main factors in the model. Perceived
usefulness describes a potential user’s subjective assessment of an application system’s ability
to improve performance in their daily lives or at work. Perceived ease of use is the degree to
which the potential user anticipates that the technology will be simple to use. External elements,
such as social, cultural, and political issues, have an impact on these two aspects. Attitude to
use reflects the user’s assessment of how desirable it is to utilize a specific information system
application while behavioral intention measures the probability that an individual will use the
application.
The model implies that the more an application is regarded to be user-friendly, the more
probable it is that the user will find it beneficial, which will increase the likelihood that the
technology will be accepted. (Davis, 1993; 1998). Generally, both the usefulness and the ease
to use of technology affect the behavioral intention, thereby affecting the final user experience.
Figure 1

Technology Acceptance Model (TAM) developed by Davis (1998)

Perceived
Usefulness
(U)

Attitude Behavioral Actual

External
Toward Intention to System
Variables
Using (A) Use (BI) Use

Perceived
Ease of Use
(E)

Since its first proposal, TAM has been further developed in many subsequent studies which
have proposed new variables. Some examples include “compatibility” (Agarwal & Prasad,
1998), “experience, self-efficacy, perceived risk and social influence” and “long-term and near-
term usefulness” (Chau, 1996). To evaluate the use of ChatGPT among senior students in a
computer engineering program, Shoufan (2023) who did not explicitly acknowledge the TAM
model might adapt TAM constructs flexibly in education. Shoufan’s model consists of
requirements, interactions, impact on learning, long-term impact, and affection. The model is
presented with detailed elaborations of both positive and negative aspects of the five constructs.
Multiple rounds of coding are used to form themes that go through multiple refinements.
Figure 2

Study Summary as a Framework for Using ChatGPT in Education by Shoufan (2023)

Interestingness

Good Admiration
Explanation
Positivity

Ease of Use Helpfulness motivation

Human-Like Efficient Helpfulness at

Optimism
Interaction Learning Work

Impact on Long-Term
Requirement Interactions Affection
Learning Impact

Human Tricky Laziness in Threat to job

Negativity

Concerns
intelligence Prompting Studying Opportunities

Background Answer Academic Malicious Use Uncertainty

Knowledge Accuracy Dishonesty

This study is largely based on Davis’ and Shoufan’s models to develop the measurement scale
of students’ perceptions of ChatGPT in English language learning. Requirement and long-term
impact are excluded. University students recruited are assumed to have adequate cognitive
abilities and technical skills to perform ChatGPT actions. The study is also limited to students’
language learning rather than their future work as part of long-term impact.

Method
The setting of the Research
The research was conducted in August 2023 when the new school year had just started in
Vietnam. The study limited itself to universities which
educate foreign language students in Vietnam. Academic Editors
Research Approach Section:
Senior Editor: [First Last]
A quantitative approach was adopted in the study
Associate Editor: [First Last]
with the aim of exploring EFL students’ perceptions
over the use of ChatGPT in English language Publication
learning. The quantitative method allowed the
Received: [DD MONTH YYYY]
researchers to conduct the study on a large scale.
Revision:
Accepted:
Published:
Copyright: © by the authors, in its year of first
publication. This publication is an open access
publication under the Creative Commons
Attribution CC BY-ND 4.0 license.
Research Instrument Development and Validation
As stated in the previous section, the study was grounded in the theory of the TAM model
developed by Davis (1998) and further revised by Shoufan (2023). Generally, the study
developed a questionnaire on three main constructs including usefulness, ease of use,
attitudes, and behaviour intention in relation to constructs which are specific to language
learning. In terms of usefulness, students can assess ChatGPT’s effectiveness in developing
English skills including listening, speaking, reading and writing. In terms of ease of use, students
can evaluate the difficulty and convenience of using ChatGPT as well as its interaction, quality
of answers and its ability to generate authentic language materials. In terms of attitudes,
students can reveal both positive and negative attitudes towards the use of ChatGPT in their
English learning. Students can express their admiration for ChatGPT’s helpfulness and
effectiveness, acknowledging it as a useful complementary learning resource and a vast source
of responses. In the meantime, students can reflect on their monitoring of ChatGPT, their
confusion, the appropriateness of ChatGPT’s answers and academic cheating. Finally, in terms
of behavior intention, students can evaluate the comfort of the learning environment with
ChatGPT, their motivation and their intention of ChatGPT use inside and/or outside class. The
questionnaire concludes with open-ended questions about their other concerns and suggestions
for a better use of ChatGPT in language learning.
The questionnaire was first designed with 3 parts. The first part seeks demographic information
like genders and class level. The second part consists of 32 items, which were supposed to be
divided into 3 dimensions. The first dimension named “Ease of use” has 9 items expressing
students’ opinions on ChatGPT interaction. The second dimension “Impact on learning” has 16
items. The final dimension “behavior intention” consists of 7 items. The questionnaire has an
acceptable reliability with Cronbach alpha .928 for the whole questionnaire, .690 for the first
dimension, .900 for the second dimension, and .834 for the final dimension. The third part has 3
open-ended questions.
To identify the statistical eligibility of items, Corrected Item – Total Correlation was calculated.
Five items had the value of corrected item-total correlation less than .3; therefore, such items
were excluded from the questionnaire. In the next step, Exploratory Factor Analysis (EFA) was
processed. In order to have variables with good significant statistics, factor loading was kept at
0.5 instead of the acceptable level 0.3. KMO value is .926, over .5 and the significant level for
Bartlett’s Test of Sphericity is .000, less than .05, which denotes that EFA is meaningful, and
there is a substantial correlation in the data. There are 5 components extracted with
Eigenvalues greater than 1. These 5 components explain 59.558 % of the variance. Therefore,
it can be concluded that the items of the questionnaire can be divided into 5 components, which
are called 5 dimensions in the later part of this article. 5 items make up the dimension
“Perceived Usefulness” with a range of values from .53 to .83. The second dimension “Ease of
Use” has 7 items with values between .50 and .71. The third dimension “Positive Attitudes”
includes 5 items with the values from .51 to .71. The dimension “Negative Attitudes” has 5 items
with the range of values from .51 to .79 while the final dimension “Behavior Intention” comprises
of 5 items with values between .58 and .735.
Three items were deleted from the questionnaire as they had a value of under .5. The
questionnaire now has three parts: Demographic information, Students’ perceptions towards the
use of ChatGPT, and open-ended questions. The second part has 27 items divided into 5
dimensions: Perceived Usefullness, Perceived Ease of Use, Positive Attitudes, Negative
Attitudes, and Behavior Intention.
The level of reliability of the final version of the questionnaire is excellent, with .923 for the
whole questionnaire and .833 for the dimension “Perceived Usefulness”, .825 for the dimension
“Ease of Use”, .807 for the dimension “Positive Attitudes”, .731 for “Negative Attitudes” and .811
for the final dimension “Behavior Intention”.
Research Participants
The participants of the study were selected on some criteria. Firstly, they must actually use
ChatGPT in their process of learning English at the tertiary level. Secondly, their major is
English. The online questionnaire on Google Forms was sent to them through email,
messenger, or Zalo. The questionnaire had responses from 369 students, among whom are 340
female students (92.2%), and 29 male students (7.8%). The great imbalance in the number of
male and female students reflects the normal rate in the language education sector.
221 of them (59.9%) are fourth-year students; 107 (29%) are third-year students, and the rest
are 41 second-year students (11.1%). At the time of the research, the first-year students had not
started their school year; therefore, there were no first year students taking part in the research.
The numbers of students in different class levels reflect that the longer students are at
university, the more likely they are to use ChatGPT in their English language learning.
Results
Students’ Perceptions towards the Use of ChatGPT in English Language Learning
Students’ perceptions towards the application of ChatGPT in English language learning are
reflected in 5 dimensions: Perceived Usefulness, Ease of Use, Positive Attitudes, Negative
Attitudes, and Behavior Intention.
Table 1

Descriptive Statistics of Dimensions

Mean Std. Deviation

Perceived Usefulness 3.30 .75
Ease of Use 3.54 .63
Positive Attitudes 3.76 .68
Negative Attitudes 3.34 .66
Behavior Intention 3.43 .74

The mean values for all the dimensions range from 3.30 (Perceived usefulness) to 3.76 (positive
attitudes). On a scale 1 to 5, the interval is 0.8 (Interval=N-1/N, N=5). Students’ perception
towards the perceived usefulness of ChatGPT has the lowest mean value, denoting that most
students selected “neutral” for their answers. In other words, students had neutral opinions on
the usefulness of ChatGPT in developing their English language skills.
Table 2

Descriptive Statistics of Perceived Usefulness

Mean Std. Deviation

My English skills have improved due to using
3.38 .89
ChatGPT.
The ChatGPT is useful to develop my speaking
3.17 1.02
skills
The ChatGPT is useful to develop my writing
3.63 .93
skills
The ChatGPT is useful to develop my reading
3.43 .91
skills.
The ChatGPT is useful to develop my listening
2.92 1.10
skills
Specifcally, students have neutral opinions on whether their English skills can be improved
using ChatGPT with a mean value of 3.38. Students disagree that ChatGPT can help them
develop listening and speaking skills (mean values 2.92 and 3.17, respectively). Yet, they
believe that ChatGPT is useful for them to enhance reading and writing skills with mean values
of 3.43, and 3.63.
Regarding the second dimension, students perceive that it is easy to use ChatGPT with a mean
value of 3.54. As shown in Table 3, students find it easy, and convenient to use ChatGPT
(mean values = 3.31, and 3.76 respectively). Furthermore, the mean values for the rest of the
items have a range from 3.40 to 3.79, expressing that ChatGPT answers are recognised to be
human-like friendly, accurate, authentic, and well-structured.
Table 3
Descriptive Statistics of Ease of Use

Mean Std. Deviation

I have no difficulty in using Chat GPT in my
3.31 .98
language learning.
Using the ChatGPT to learn English is
3.76 .90
convenient.
ChatGPT makes human-like friendly
3.47 .93
impressions.
ChatGPT provides good explanations. 3.79 .90
ChatGPT answers are well- structured. 3.65 .90
ChatGPT answers are accurate. 3.40 .83
ChatGPT can generate authentic language
3.40 .89
materials.
The second lowest mean value (3.34) belongs to students’ negative attitudes towards the use of
ChatGPT. Among the 5 items, the first item on the requirement for careful monitoring has the
highest mean value with 3.63, followed by the item on the probability of academic cheating
facilitated by ChatGPT (mean value=3.45). Yet, students have a neutral viewpoint over the
issue of the quality of answers generated by ChatGPT with the mean values between 3.17 and
3.29.
Table 4
Descriptive Statistics of Negative Attitudes

Mean Std. Deviation

The use of Chat GPT requires careful
3.63 .900
monitoring.
ChatGPT affects learning negatively because I
3.29 1.029
can find answers and solutions without efforts.
I am confused about the answers of ChatGPT. 3.17 .864
Chat GPT can produce biased and
3.19 .918
inappropriate content.
ChatGPT will make academic cheating easier. 3.45 1.057

Yet, most students have high levels of positive attitudes towards the use of ChatGPT in English
language learning with the highest mean value (3.7683). With the mean values from 3.58 to
3.94, most students agree that ChatGPT is a good technological tool functioning as a
complementary learning resource and helpful, effective support for language learning.
Table 5
Descriptive Statistics of Positive Attitudes

Mean Std. Deviation

ChatGPT has amazing capabilities. 3.75 .86
ChatGPT is a helpful and effective technology
3.94 .83
for language learning.
ChatGPT is good as a complementary learning
3.80 .87
resource.
Asking follow-up questions helps ChatGPT find
3.58 .84
the answer.

In addition, the mean value for the dimension “behavior intention” is 3.43, expressing that most
students chose “Agree” for this dimension, which denotes students’ intention to use ChatGPT to
develop the English language. Students are found to be interested, and motivated to use
ChatGPT as expressed by the mean values in Table 6.
Table 6
Descriptive Statistics of Behavior Intention

Mean Std. Deviation

ChatGPT is a comfortable learning environment. 3.64 .89
I feel motivated to use ChatGPT more. 3.43 .92
ChatGPT makes learning English enjoyable. 3.48 .94
I like using ChatGPT to practice English inside the
3.17 .98
classroom
I like using ChatGPT to practice English outside the
3.45 .97
classroom.

Among 369 students, 233 students (63%) have the wish to keep using ChatGPT for their
language learning. According to these students, ChatGPT is useful in generating ideas for their
writing tasks, providing them with helpful hints for their reading exercises, as well as authentic
materials or guidelines for their language study.
By contrast, 103 students do not want to use ChatGPT (27.9%). These students believe that
ChatGPT can be helpful for other subjects, but ineffective with language learning because they
are man-made, and there is a lack of interaction, which is very important for developing a
language. Reasons for such a decision may be explained by what is found in the open-ended
questions. Regarding the question on students’ challenges to using ChatGPT, the typical
answer is the concern on users’ dependence on ChatGPT, leading to reduced creativity in
developing productive skills like writing and reading, and laziness. “It may affect my way of
learning since I may rely on it too much if I got lazy and wasn’t careful enough.”, one answer
cited from a fourth-year student’s response. “I am scared if I use ChatGPT too much because it
can make me reduce my imagination and creativity without effort, so it is a big challenge when I
take a test without it”, a third-year student is worried about her use of ChatGPT. The second
common challenge for students in using ChatGPT is the accuracy of the information provided.
“Sometimes it can be really tricky to find the best way to express my questions or requests to
the chatbot, and if I use inaccurate words, it also results in misleading, unrelated information.
Also, I believe that using ChatGPT too frequently would make me rely on the app more and
maybe I cannot enhance my skills if I get used to using the chatbot. I think looking for a method
to use this app without depending too much on it is essential” (A challenge of a second-year
student).
33 students (8.9%) do not know whether they will continue using ChatGPT in the future. The
reasons are their insufficient experience with the application of ChatGPT to learn languages.
Gender Difference
Independent sample T Test was employed to test whether genders have any effects on
students’ perceptions towards the use of ChatGPT to learn the English language. As shown in
Table 7, there are some differences between the perception of female and male students when
mean differences are identified. Yet, such a difference is not significant when the significant
value is .626, much higher than 0.05. In addition, there is a great imbalance between the
number of male and female students (7.8% and 92.2%, respectively) which may impact the
result.
Table 7

Gender difference

Gender N Mean Std. Deviation Std. Error Mean

Perceived Male 29 3.24 .57 .10
usefulness Female 340 3.31 .77 .04

Ease of use Male 29 3.52 .58 .10

Female 340 3.54 .63 .03
Positive Male 29 3.65 .71 .13271
attitudes Female
340 3.77 .67833 .03679

Negative Male 29 3.22 .51 .09

attitudes Female 340 3.35 .67 .03
Behavior Male 29 3.35 .66 .12
intention Female 340 3.44 .75 .044

Class Level Difference

MANOVA was applied to test whether the class level (i.e. second year, third-year, fourth-year
students) has effects on students’ perception towards the use of ChatGPT to develop the
English language. As indicated in Table 8, the sig. level is .002, less than .005, meaning that the
class level has significant effects on students’ perception towards the application of ChatGPT to
study the English language.
Table 8

MANOVA Test: Level Difference

Value F Hypothesis df. Error df. Sig.

.906 2.424 15.000 996.963 .002

To sum up, EFL students surveyed in Vietnam have neutral opinions on the usefulness of
ChatGPT to develop the English language, and have a low level of negative attitudes towards it
and a high level of positive attitudes towards the use of ChatGPT is recognised. In addition,
students find it easy to use ChatGPT as well as perceive it as a useful tool for their language
study, resulting in a high rate of them who keep using it in the future (63%). Genders are
identified not to have effects on students’ perceptions whereas the class level has certain
impacts on students’ perceptions.
Discussion
The fast development of technology has brought enormous changes to all fields of life,
especially education. The appearance of ChatGPT marks a new era for education in general
and language learning in particular (Shaikh et al., 2023; Shoufan, 2023). The findings of the
research show that language learners accept the new technological tool though they have not
had much experience in using it in their process of learning foreign languages. This is in line
with Shaikh et all (2023) who acknowledge ChatGPT as a friendly, accessible, and useful tool
from the viewpoints of English language learners. The participants in their study appreciated the
usefulness of ChatGPT’s prompts while they completed English language tasks. Our study also
confirms the promising and practical role of ChatGPT in language learning. However, how to
use the tool effectively is of great concern among language teachers and learners. Shoufan
(2023) indicated students’ lack of background knowledge to provide appropriate prompts and
evaluate responses to ChatGPT. Therefore, the author emphasizes that ChatGPT should not be
the only source for learning. In fact, Kohnke et al (2023) insist on the need for digital
competencies, instructions, or guidelines for using ChatGPT. In our study, neutral results can be
because students in the study might have not received adequate guidance on how to use
ChatGPT in their language learning before they evaluated ChatGPT’s functions. Therefore, it
would be beneficial to provide students chances to be exposed to formal instructions in
language learning so that the usefulness of ChatGPT can be fully exploited.
With limited experience in using ChatGPT to learn a foreign language, students in Vietnam tend
to have neutral ideas on its perceived usefulness, especially when foreign language learning
requires much interaction and sources provided may not be guaranteed to be reliable. Students
find ChatGPT more useful for learning reading and writing rather than speaking and reading,
which is similar to the findings of many other studies including Baidoo-Anu & Owusu Ansah
(2023), George and George (2023), Ali et al. (2023), Bin-Hady et al. (2023). While some studies
(Baskara, 2023; Kohnke et al., 2023) mention the role of ChatGPT in improving oral
competence and interaction/communication skills, this study does not reinstate this idea. This
aligns with Ho (2024) who recognises ChatGPT’s drawbacks in enhancing English listening and
speaking skills and oral communication. This can be explained by the fact that ChatGPT does
not support functions to develop these skills or that most students only access free versions of
ChatGPT with limited capabilities. It is worthwhile that institutions should invest in technological
environments to fully exploit the benefits of AI technologies (O'Dea & O'Dea, 2023).
Nevertheless, students are found to be interested, and motivated to use ChatGPT in their
language learning, which may promote their independent learning. This aligns with the insights
of many studies (Ali et al, 2023; Bin-Hady et al., 2023; Fauzi et al.., 2023; Kohnke et al., 2023).
ChatGPT helps develop language skills, “improving time efficiency and effectiveness, and
providing support and motivation” (Fauzi et al., 2023, p.14889). In the same vein, Shaikh et al
(2023) who emphasize the promising potential of ChatGPT in language learning and teaching,
maintain that ChatGPT can be used
to enhance learners’ language skills, scaffold their learning process and provide a support
tool for language learning. Bin-Hady et al., (2023) say it can provide an “enjoyable learning
environment” (p. 8) and promote “learner autonomy” (p. 9).
In this study, students doubt the quality of ChatGPT answers which require an adequate level of
monitoring. The issue of accuracy in ChatGPT’s content raises the need to educate students to
effectively create prompts and evaluate ChatGPT’s responses (Shoufan, 2023). Some students
are concerned about their overreliance on ChatGPT’s responses which may not help improve
their language skills or they referred to plagiarism related to ChatGPT’s use which is also found
in other studies (Cassidy, 2023; Pavlik, 2023, Shoufan, 2023). In fact, ChatGPT’s “responses
are not entirely original but paraphrases of sources that have not been appropriately cited (i.e.,
plagiarism)” (Kohnke et al., 2023, p. 8). Therefore, Ho (2024) who acknowledged students’
concerns about ChatGPT-related plagiarism proposed that it is worthwhile for teachers to raise
students’ awareness of the issue and guide them to go through different steps in evaluating
suggestions by ChatGPT.
Yet, concerns about the application of ChatGPT for foreign language learning do not prevent
English majored students in Vietnam from appreciating the adoption of ChatGPT in studying
foreign languages. Vietnamese students are identified as having a high level of positive
attitudes towards the use of ChatGPT in learning a foreign language. Currently, the rate of
Vietnamese students who have the intention to keep using ChatGPT for their language learning
is high, which is intricately linked to the moderate rates for usefulness and ease of use, the two
deciding factors for students’ behavioral intention. ChatGPT has witnessed a high degree of
acceptance and favorable attitudes in many fields as well as in language learning. Shaikh et al.
(2023) who explored Norwegian students’ perceptions of ChatGPT demonstrate that most
“participants’ ratings were on the higher end of the Likert scale, which shows participant
satisfaction with the system usability in terms of ease of learning, ease of use and usefulness”
(p. 1592), emphasizing learner intention to use the tool for language learning. Therefore, this
well fits TAM that indicates both the usefulness and the ease to use of technology may affect
the behavioral intention, thereby affecting the learner experience. To effectively promote
ChatGPT in language learning, a higher degree of usefulness is expected, though.
Conclusion
In conclusion, ChatGPT is perceived as a positive tool for learning a foreign language though
concerns on its use are still found, especially in developing countries like Vietnam where people
have limited access to new technology and ChatGPT was just introduced for a short time (Wu et
al., 2023). ChatGPT is found to be a promising learning tool when students taking part in the
research are keen on keeping using it for their language learning.
Yet, as ChatGPT is a new tool, making it a more useful and meaningful learning and teaching
tool requires teachers’ and learners’ much effort. Practitioners should be aware of how they will
incorporate ChatGPT in their lessons efficiently and learners need to make a wise use of it. To
make feasible and effective for the use of ChatGPT in language teaching and learning, teachers
and learners need to pay attention to these recommendations Firstly, new teaching strategies
to integrate ChatGPT in the lessons should be explored sufficiently before teachers and
learners apply it in their teaching and learning (Shaikh et al., 2023). Applying technology in
language teaching has a long history, moving from the initial use of basic technological tools like
videos, and non-internet software to the very advanced online ones. Teachers have upgraded
themselves with methodologies to integrate technology into language teaching (Vo et. al.,
2020). As ChatGPT has just been developed, its use in language teaching and learning requires
both practitioners and policymakers to work on approaches for the incorporation of AI in general
and ChatGPT in particular for effective teaching strategies created (Hoang et al., 2023; Ho,
2024; Lodge et.al., 2023). Secondly, careful monitoring from teachers would be also one matter
to be considered for the use of ChatGPT in language learning to avoid being distracted by
unreliable information that may be generated by ChatGPT as well as prevent cheating from
students with the help of ChatGPT (Lodge et.al., 2023). Thirdly, users should not abuse
ChatGPT to avoid being overreliance on it (Fauzi et al., 2023). Over-dependence on technology
can be reduced by students’ self-control and teachers’ monitoring (Vo, 2023). In addition,
human interaction should be kept the priority in language teaching; therefore, a balance
between the use of technological tools like ChatGPT and teacher-student interaction can be
maintained.
Policy-makers in the educational system are advised to incorporate AI tools like ChatGPT in the
existing educational programs, especially teacher education ones. Being able to use technology
like AI tools in learning, and then applying them in a teaching career later is a “must” skill a pre-
service student needs to possess (Vo et al., 2020). Courses on how to integrate technology in
general and ChatGPT, in particular, should be added to teacher education programs.
Together with the updated methodologies to suit the new teaching context, the application of
ChatGPT to learn languages will be largely facilitated and enhanced if new functions are
continuously added to the chatbox. As shown from this study, ChatGPT is not very useful in
developing students’ speaking and listening skills due to its lack of functions. In case students
can have chances to practise these skills with voice functions, ChatGPT will play a more crucial
role as a teaching tool. Even though ChatGPT was created to serve people’s needs in a wide
range of aspects, educational purposes are advised to play a more crucial role in its
development.
This study which has provided information on students’ perceptions towards the use of
ChatGPT in language learning has set the initial step on the application of ChatGPT in language
teaching and learning. Yet, although quantitative research can help gather information on a
large scale, the study could have provided more useful information if qualitative data had been
collected for deeper understandings and further explaining what has been discovered in the
quantitative study. In addition, as ChatGPT has just been applied in language teaching and
learning for a short time, students’ responses to the questionnaire may not fully reflect reality.
ChatGPT can be employed to facilitate the language teaching and learning process more
effectively if teaching methodologies can incorporate the use of ChatGPT as a technology tool
in teaching practice. Studies on the application of ChatGPT in enhancing specific language
skills like writing, listening, speaking, and reading skills as well as language-related subjects
such as translation and interpretation, grammar, or linguistics are in need of further knowledge
of ways to apply ChatGPT in language teaching.
Conflict of Interest
The author(s) disclose that they have no actual or perceived conflicts of interest. The research
is funded by the Partnership for Higher Education Reform (PHER). No AI has been used to
produce this manuscript.
References
Agarwal, R., & Prasad, J. (1998). A conceptual and operational definition of personal
innovativeness in the domain of information technology. Information Systems Research,
9(2), 204–215. https://doi.org/10.1287/isre.9.2.204.
Ali, J. K. M., Shamsan, M. A. A., & Hezam, T. A., & Mohammed, A. A. (2023). Impact of
ChatGPT on learning motivation: teachers and students’ voices. Journal of English
Studies in Arabia Felix, 2(1), 41–49. http://dx.doi.org/10.56540/jesaf.v2i1.51.
Baidoo-Anu, D., & Owusu Ansah, L. (2023). Education in the Era of Generative Artificial
Intelligence (AI): Understanding the Potential Benefits of ChatGPT in Promoting Teaching
and Learning. https://doi.org/10.61969/jai.1337500.
Baskara, R. (2023). Exploring the implications of ChatGPT for language learning in higher
education. Indonesian Journal of English Language Teaching and Applied Linguistics, 7(2),
343–358.
Bin-Hady, W. R. A., Al-Kadi, A., Hazaea, A., & Ali, J. K. M. (2023). Exploring the
dimensions of ChatGPT in English language learning: a global perspective. Library
Hi Tech. . https://doi.org/10.1108/LHT-05-2023-0200.
Casidy, C. (2023). Australian universities to return to ‘pen and paper’ exams after students
caught using AI to write essays. The Guardian Online.
Chau, P. Y. K. (1996). An empirical assessment of a modified technology acceptance
model. Journal of Management Information Systems, 13, 185‐204.
https://doi.org/10.1080/07421222.1996.11518128.

Davis, F. D. (1993). User acceptance of information technology: system characteristics, user

perceptions and behavioral impacts. International Journal of Man-Machine Studies, 38(3),
475–487. https://doi.org/10.1006/imms.1993.1022.
Davis, F. D. (1998). Perceived usefulness, perceived ease of use, and user acceptance of
information technology. MIS Quarterly, 13(3), 319–340. http://dx.doi.org/10.2307/249008
Fauzi, F., Tuhuteru, L., Sampe, F., Ausat, A. M. A., & Hatta, H. R. (2023). Analyzing the role of
ChatGPT in improving student productivity in higher education. Journal on Education, 5(4),
14886–14891. http://dx.doi.org/10.31004/joe.v5i4.2563.
George, A. S., & George, A. H. (2023). A Review of ChatGPT AI’s Impact on Several Business
Sectors. Partners Universal International Innovation Journal, 1(1), 9–23.
https://doi.org/10.5281/zenodo.7644359.
Ho, P. X. P. (2024). Using ChatGPT in English Language Learning: A Study on I.T. Students’
Attitudes, Habits, and Perceptions. International Journal of TESOL & Education, 4(1), 55–
68. https://doi.org/https://i-jte.org/index.php/journal/article/view/432.
Hoang, N. T., Duong, N. H., & Le, D.-H. (2023). Exploring Chatbot AI in improving vocational
students’ English pronunciation. AsiaCALL Online Journal, 14(2), 140–155.
https://doi.org/https://doi.org/10.54855/acoj.231429.
Hong, W. C. H. (2023). The impact of ChatGPT on foreign language teaching and learning:
opportunities in education and research. Journal of Educational Technology and
Innovation, 5(1), 37–45.
Jeon, J., & Lee, S. (2023). Large language models in education: A focus on the complementary
relationship between human teachers and ChatGPT. Education and Information
Technologies, 28, 15873–15892. http://dx.doi.org/10.1007/s10639-023-11834-1.
Johinke, R., Cummings, R., & Di Lauro, F. (2023). Reclaiming the technology of higher
education for teaching digital writing in a post—pandemic world. Journal of University
Teaching & Learning Practice, 20(2). https://doi.org/https://doi.org/10.53761/1.20.02.01
Kasneci, E., Seßler, K., Küchemann, S., Bannert, M., Dementieva, D., Fischer, F., &
Kasneci, G. (2023). ChatGPT for good? On opportunities and challenges of large
language models for education. Learning and Individual Differences, 103.
. http://dx.doi.org/10.1016/j.lindif.2023.102274.

Kohnke, L. (2023). L2 learners’ perceptions of a chatbot as a potential independent language

learning tool. L2 Learners’ Perceptions of a Chatbot as a Potential Independent Language
 Learning Tool, 17(2), 214–226. https://doi.org/10.1504/ijmlo.2023.128339.
Kohnke, L., Moorhouse, B. L., & Zou, D. (2023). ChatGPT for language teaching and learning.
RELC Journal, 54(2). http://dx.doi.org/10.1177/00336882231162868.
Lodge, J. M., Barba, P., & Broadbent, J. (2023). Learning with Generative Artificial Intelligence
Within a Network of Co-Regulation. Journal of University Teaching and Learning Practice,
20(7). https://doi.org/10.53761/1.20.7.02
Lund, B. D., & Wang, T. (2023). Chatting about ChatGPT: how may AI and GPT impact
academia and libraries? Library Hi Tech News. Library Hi Tech News, 40(3), 26–29.
Mhlanga, D. (2023). Open AI in Education, the Responsible and Ethical Use of ChatGPT
Towards Lifelong Learning. SSRN Electronic Journal.
Mohammed, A., Al-ghazali, A., & Alqohfa, K. (2023). Exploring ChatGPT uses in higher studies:
A case study of Arab postgraduates in India. Journal of English Studies in Arabia Felix,
2(2), 8–16.
O’Dea, X., & O’Dea, M. (2023). Is Artificial Intelligence Really the Next Big Thing in Learning
and Teaching in Higher Education? A Conceptual Paper. Journal of University Teaching &
Learning Practice, 20(5). https://doi.org/https://doi.org/10.53761/1.20.5.05
Pavlik, J. V. (2023). Collaborating With ChatGPT: Considering the Implications of Generative
Artificial Intelligence for Journalism and Media Education. Journalism & Mass
Communication Educator, 78(1). https://doi.org/10.1177/10776958221149577.
Rahman, M. M., & Watanobe, Y. (2023). ChatGPT for education and research: Opportunities,
threats, and strategies. Applied Sciences, 13(9), 57–83.
Rudolph, J., Tan, S., & Tan, S. (2023). War of the chatbots: Bard, Bing Chat, ChatGPT, Ernie
and beyond. The new AI gold rush and its impact on higher education. Journal of Applied
Learning and Teaching, 6(1). https://doi.org/10.37074/jalt.2023.6.1.23.
Sallam, M. (2023). ChatGPT utility in healthcare education, research, and practice: systematic
review on the promising perspectives and valid concerns. Healthcar, 11(6).
https://doi.org/10.3390/healthcare11060887.
Sami, A., Uddin, I., Fayyaz, N., Bilal, M., Shahid, M., & Ali, I. (2023). Getting to Know ChatGPT :
An Introduction to Implementation and Working Getting to Know ChatGPT : An Introduction
to Implementation and Working. Institute of Computer Sciences and Information
Technology Journal. International Conference on Computing Technologies, Tools and
Applications.
Shaikh, S., Yayilgan, Y., Klimova, B., & Pikhart, M. (2023). settingsOrder Article Reprints Open
AccessArticle Assessing the Usability of ChatGPT for Formal English Language Learning.
EKHPE, 13(9). https://doi.org/10.3390/ejihpe13090140.
Shoufan, A. (2023). Exploring Students¬¬¬’ Perceptions of CHATGPT: Thematic Analysis and
Follow-Up Survey. IEEE, 38805–38818.
Tlili, A., Shehata, B., Adarkwah, M. A., Bozkurt, A., Hickey, D. T., Huang, R., & Agyemang, B.
(2023). What if the devil is my guardian angel: ChatGPT as a case study of using chatbots
in education. Smart Learning Environments, 15(23), 1–24.
Tran, T. N., & Tran, H. P. (2023). Exploring the role of ChatGPT in developing critical digital
literacies in language learning: a qualitative study. Proceedings of the AsiaCALL
International Conference, 1–17.
Vo, T. K. A. (2023). English-majored students’ self control over their online learning. Journal of
Science and Technology, 21(2), 14–18.
Vo, T. K. A., Pang, V., & Lee, K. W. (2020). Evaluating Vietnam’s pre-service English Teacher
Education program for technology integration in education. Call-Ej, 21(3), 8–22.
Wu, T., He, S., Liu, J., Sun, S., Liu, K., Han, Q. L., & Tang, Y. (2023). A Brief Overview of
ChatGPT: The History, Status Quo and Potential Future Development. IEEE/CAA Journal
of Automatica Sinica, 10(5), 1122–1136. https://doi.org/10.1109/JAS.2023.123618.