CAN THO UNIVERSITY
SCHOOL OF FOREIGN LANGUAGES
DEPARTMENT OF ENGLISH LANGUAGE AND CULTURE
FINAL GROUP PROJECT
Class: Writing Scientific Reports (SG277) – 01
AI dependence's effect on EFL students' writing
skills
Instructor: Đỗ Xuân Hải, Ph.D.
Students: Nguyễn Văn Nguyên – B1912467
Trần Quỳnh Như – B1912469
Nguyễn Phước Thịnh - B2111438
Lê Đỗ Tường Vi - B2111449
Can Tho, September 2024
INTRODUCTION
1
The integration of artificial intelligence (AI) tools in academic writing has
sparked significant debate due to the potential negative implications associated with
their use. While AI-driven applications, such as grammar checkers and automated
feedback tools, offer support to students in enhancing writing efficiency and accuracy,
they also pose certain risks. Pitukwong and Saraiwang (2024) have demonstrated that
AI-based digital writing tools can significantly enhance writing performance. However,
this improvement often comes at the expense of independent critical thinking and
creativity. Similarly, studies conducted by Zou and Huang (2023) have raised concerns
regarding the potential of AI tools to diminish student engagement with the writing
process, leading to a reliance on algorithms rather than fostering deeper learning and
skill acquisition. Furthermore, research by Li, Li, and Cho (2023) has highlighted the
inadvertent weakening of students' ability to develop coherent arguments and
synthesize complex ideas, essential skills in academic writing, as a result of the use of
AI tools. This growing reliance on AI may lead to a superficial engagement with the
writing process, potentially resulting in learners becoming overly dependent on
technology, thereby risking the erosion of their problem-solving skills and creativity.
The impact of AI dependence on the development of writing skills among EFL
(English as a Foreign Language) students is a crucial area of study in the context of
modern education. As technology continues to evolve, the integration of AI tools in
language learning has become increasingly prevalent. Our topic “AI dependence's
effect on EFL students' writing skills” seeks to explore how reliance on these tools
influences students' writing abilities, particularly focusing on their capacity to express
ideas clearly and accurately in English. As students increasingly turn to AI tools for
learning, it’s essential to grasp their experiences, attitudes, and expectations. Being the
main users of these technologies, students' interactions with AI can profoundly impact
their writing abilities. Their reliance on AI influences not only their approach to writing
tasks but also their critical thinking skills and overall language development. By
exploring students’ perspectives, educators can gain insights into the advantages and
potential challenges of incorporating AI into language instruction. This knowledge is
vital for creating effective curricula and teaching strategies that harmonize
technological support with the cultivation of independent writing skills.
2
This study was chosen due to the growing use of AI tools by EFL students, such
as grammar checkers, translation services, and content generators, which, while
beneficial, may hinder the development of essential writing skills. Over-reliance on AI
can lead to a decline in grammar, vocabulary, and sentence structure proficiency,
affecting students' academic and professional success. This research aims to understand
the balance between using AI as a supportive tool and fostering independent writing
skills. By examining this relationship, educators can develop strategies that integrate
AI effectively, ensuring students still engage in critical writing practice. The study is
timely as educational institutions increasingly incorporate AI, making it crucial to
evaluate its impact on language learning outcomes.
This project explores the impact of AI dependence on EFL (English as a Foreign
Language) students' writing skills, particularly how the use of AI tools affects critical
thinking and independent language proficiency. The research will begin by reviewing
the advantages of AI in enhancing writing efficiency, such as improved grammar and
structure, as shown in various studies. However, it will also address concerns raised by
researchers about the potential negative effects of over-reliance on AI, including
weakened critical thinking, creativity, and problem-solving skills. The project will
analyze both student and teacher perspectives, examining the balance between using AI
as a supportive tool and ensuring that students actively engage with the learning process
to develop essential writing abilities. The research will culminate in recommendations
for integrating AI in language education in a way that enhances learning while
minimizing the risks associated with technology overdependence.
3
SUMMARIES
4
Summary 1
Barrett & Pack (2023) investigated the perspectives of students and teachers on
the use of generative artificial intelligence (GenAI) in the writing process. The study
involved 226 participants, with 158 students from a public research university in the
United States and 68 teachers from various institutions worldwide including the
majority of undergraduate students (97%), native English speakers (96%) and primarily
second language teachers. Data collection was conducted through a cross-sectional
questionnaire, and the responses were analyzed using statistical methods such as the
Mann–Whitney U test and principal component analysis. These methods allowed the
researchers to compare perceptions and identify patterns in the data. The findings
indicated that both students and teachers generally agreed on the acceptable use of
GenAI for brainstorming and outlining tasks. However, there was a significant
difference in opinions about its use for writing and revising, with teachers expressing
more concern about the potential misuse of GenAI compared to students. This study
underscored the views on integrating AI in educational settings, highlighting the
importance of establishing clear guidelines and ethical considerations to ensure that
GenAI tools are used effectively and responsibly in the writing process. The results
suggested a need for ongoing dialogue and collaboration between educators and
students to navigate the evolving landscape of AI in education.
Words in text: 211
AWL words: 17.06%
5
Summary 2
Kim, Yu, Detrick and Li (2024) conducted an insightful study on students'
attitudes toward utilizing generative AI for academic writing support. The study
focused on 20 Chinese students attending a Sino-British research university in Suzhou,
China. These students were enrolled in English-language academic programs, with
varying degrees of writing proficiency and different levels of familiarity with AI
technologies. The research aimed to explore how these students interacted with a
ChatGPT4-embedded writing system and how they perceived its utility in assisting with
academic writing tasks. Data for the study was collected through semi-structured
interviews conducted via Zoom. During the interviews, students were asked to reflect
on their experiences using the AI system after completing academic writing
assignments designed for the study. The researchers also analyzed essays and screen
recordings of the students' writing processes, which served as reference points during
the interviews to better understand their engagement with the AI system. The analysis
of the data followed both inductive and deductive thematic approaches, helping the
researchers identify recurring patterns and key themes in the students’ perceptions. The
findings revealed that students generally regarded the AI as a multifaceted tool,
functioning as a writing assistant, virtual tutor, and even a digital peer. While they
appreciated the AI’s ability to streamline the writing process and boost performance,
students also raised concerns about the system's limitations, particularly its struggles
with higher-order thinking and nuanced contextual understanding.
Words in text: 236
AWL Words: 17.37%
6
Summary 3
Tarchi, Casado and Brante (2024) explored the use of ChatGPT as a tool for
assisting undergraduate students with source-based writing (SBW) tasks. SBW
involved producing written work based on previously read texts and requires mastering
writing conventions and accurately synthesizing information from source material. The
research aimed to categorize the strategies employed by students when using ChatGPT
and examine how these strategies influenced the quality and content of their written
outputs. The participants were twenty-seven non-expert psychology undergraduate
students (average age of 20.37 years) completing an SBW task with access to ChatGPT
from the University of Florence, Italy. Data was collected through a combination of
sociodemographic questionnaire, an academic writing motivation scale and a measure
of perceived prior knowledge. After the writing task, students participated in a
retrospective think-aloud interview with the researcher, reflecting on how they utilized
ChatGPT during the task. The authors explored that ChatGPT are capable of multiple
tasks that can scaffold students’ reading and writing competences. However,
participants seemed hesitant to use ChatGPT due to ethical concerns surrounding it
(using ChatGPT feels like cheating). Then, strategic use of ChatGPT did not
significantly improve the quality of the written products because most students
struggled with integrating arguments and counterarguments in their essays. Therefore,
the authors suggested giving priority to studies investigating how students can be taught
to use ChatGPT as a partner when writing, by collaborating with it rather than non-
critically using its content in their written reflections.
Words in text: 248
AWL words: 15.32%
7
Summary 4
Seo (2024) conducted an initial study to explore how ChatGPT can be integrated
into writing education. The study focused on analyzing students' usage patterns and the
influence on their writing skills. The study involved 44 university students in South
Korea, the research aimed to fill a gap in understanding how ChatGPT-assisted
narrative writing affects college-level L2 (second language) writing. The study
categorized student-initiated prompts into three main types: requests for language use,
requests for revision, and requests for information. Employing paired sample t-tests and
Wilcoxon signed rank tests, the researchers assessed the effects of this intervention on
various aspects of writing performance. Notably, the findings reveal potential negative
effects on students' writing skills. The reliance on ChatGPT for basic editing and
proofreading may hinder the development of independent editing abilities, fostering an
over-dependence on AI tools for surface-level corrections. Additionally, a significant
decrease in the mean length of T-units (MLT) indicated a regression in syntactic
complexity, raising concerns about the overall sophistication of students' writing. This
decline could be misinterpreted as a decrease in writing proficiency, potentially leading
to inaccurate assessments. Moreover, the focus on surface-level feedback may detract
from higher-order writing skills and critical thinking. Although there was progress in
adapting to the narrative genre, these findings underscore the necessity for a balanced
instructional approach that promotes both technical and creative competencies in
writing education.
Words in text: 234
AWL words: 15.81%
8
Summary 5
Punar Özçelik and Yangın Ekşi (2024) investigated the effect of ChatGPT, an
AI-driven chatbot, on students' learning of register knowledge in various English
writing tasks. Using a one-shot case study approach, the research involved eleven
undergraduate students aged 19 to 21 from the faculties of Engineering, Economics and
Administrative Sciences, and Education at a state university in Turkey. Data were
collected qualitatively through observations and unstructured interviews, then analyzed
using Creswell’s qualitative analysis method. The study aimed to explore the benefits
and limitations of ChatGPT, focusing on students' feedback and suggestions for its
improvement.Findings revealed that ChatGPT helped enhance students' formal writing
skills, with participants showing enthusiasm for its use in academic settings. Despite
these positive outcomes, the study identified several challenges, such as technical issues
and the chatbot's limited ability to handle informal and neutral registers effectively.
Moreover, the small sample size, influenced by the February 2023 earthquake, limits
the generalizability of the results. Differences in students’ proficiency across various
writing contexts also impacted the outcomes. Lastly, the study's focus on self-editing
did not fully address all aspects of the writing process, suggesting that further research
with larger, more diverse groups and better prompt designs is needed to gain a deeper
understanding of ChatGPT’s potential in improving writing skills.
Words in text: 214
AWL words: 18,69%
9
Summary 6
Garcia et al.(2024) investigated how a form of AI technology, ChatGPT
influenced the completion of school tasks for primary education. The study evaluated
the pros and cons of AI use, emphasizing learning outcomes among students. The study
included five schools in the city of Tacna, Peru, using non-probability convenience
sampling with 110 basic education teachers participated. Data collection was conducted
through Semi-structured interviews and analyzed using grounded theory with the
assistance of Atlas. ti 23 software. The data revealed that ChatGPT provided many
benefits in terms of a dynamic inforesource leading to improvements in teaching and
learning strategies. However, the study also warned a note of caution about an over-
reliance on AI tools, warning that students might grow less skilled at critical thinking
and independent research by outsourcing the work to machines. Moreover, risks were
listed such as the promotion of academic plagiarism and students go pass lessen
cognitive processes when solve school tasks. This study recognized the positive
implications of AI tool, particularly for the purposes of AI in writing and other related
skills, ultimately it stresses that there is still a great deal to be done when it comes to
the associated challenges around.
Words in text: 200
AWL words: 15,00%
10
Summary 7
Lee, Davis, and Lee (2024) conducted a study to examine Korean university
students' attitudes towards the use of artificial intelligence (AI)-based tools in English
writing courses. The research involved 80 participants, primarily sophomores, enrolled
in general English classes. The study utilized both quantitative and qualitative methods
for data collection. An online survey assessed students' perceptions regarding the
effectiveness of three specific tools: Google Translate, Naver Papago (machine
learning-based), and Grammarly (generative AI). Additionally, focus group interviews
were conducted to obtain more insights into the tools' usage and their perceived impact.
The analysis employed analysis of variance (ANOVA) to identify correlations between
students' English proficiency levels and their evaluations of the AI tools. Thematic
analysis was used to interpret the interview data, categorizing the strengths and
weaknesses of each tool. The findings revealed that participants considered Google
Translate and Naver Papago particularly effective for enhancing vocabulary,
identifying errors, and improving overall writing proficiency. Conversely, Grammarly
was appreciated for its more sophisticated feedback on writing structure and coherence.
Despite the benefits, concerns emerged regarding potential over-reliance on these AI
tools, which could hinder students' ability to develop independent writing skills. The
study underscored the importance of guiding students on how to use AI-based tools as
a supplement to their learning, rather than as a substitute for fundamental writing skill
acquisition. This highlighted the need for a balanced approach to integrating AI in
language education, ensuring that it aids, rather than diminishes, students' linguistic
growth.
Words in text: 246
AWL words: 22,6%
11
Summary 8
Li, Li and Cho (2023) investigated the potential impact of ChatGPT on
improving Chinese writing skills among Chinese language learners (CLLs) from low-
income families, focusing on whether using ChatGPT after school could enhance their
abilities. Four 9th-grade students, aged 14 to 16, learning Mandarin for the first time
and fluent in English and Spanish, participated in the study. Data were gathered from
writing scores, samples, and weekly reflections. During the intervention phase,
participants used ChatGPT twice a week to generate writing ideas and revise their work,
though direct copying was prohibited. The results indicated that all participants
improved their Mandarin writing scores during the intervention. Students with lower
baseline scores saw significant, immediate progress, while those with higher scores
showed gradual improvement. Despite a slight decline during the reversal phase, scores
remained above baseline, suggesting a lasting positive impact. Participants were also
able to self-correct errors with ChatGPT’s feedback and appreciated the immediate
assistance it provided. However, some negative findings emerged, including continued
difficulties with paper-based writing, which retained numerous errors. This highlighted
a gap between students' ability to use ChatGPT effectively online and their offline
performance. Additionally, concerns about academic integrity arose, as students
required guidance to use ChatGPT appropriately without over-relying on it.
Words in text: 210
AWL words: 18.57%
12
Summary 9
Zou and Huang (2023) explored how doctoral students accept ChatGPT for
writing, using the Technology Acceptance Model (TAM) as a guiding framework.
Their goal was to identify the factors that affect students' intentions to use ChatGPT in
their writing. The study involved 242 doctoral students, aged 24 to 43, who were
enrolled in a required academic writing course at a technological university in China.
Data were gathered through an online survey conducted in March 2023, which included
demographic questions and various TAM-related aspects. The responses were
measured using a 6-point Likert scale. The researchers analyzed the data using SPSS
and Mplus software, applying descriptive statistics, correlation analyses, and structural
equation modeling (SEM) to understand the relationships between variables, while
controlling for gender and utilizing bias-corrected bootstrapping for mediation analysis.
While the overall findings were positive regarding students' acceptance of ChatGPT,
several challenges were identified. The study revealed a notable lack of direct influence
from perceived usefulness and ease of use on students' intentions, suggesting that these
factors mainly work through their attitudes. Furthermore, the focus on prior experience
indicated that less experienced users might struggle to recognize the tool's benefits,
limiting its wider adoption. As a result, although doctoral students showed strong
intentions to use ChatGPT, issues related to usability and experience may hinder its
successful integration into their academic writing.
Words in text: 224
AWL words: 19.64%
13
Summary 10
Pitukwong and Saraiwang (2024) conducted a comprehensive study to evaluate
the effectiveness of digital tools in helping Thai university students overcome common
writing challenges. These challenges included grammatical errors, developing ideas,
and organizing paragraphs. The study concentrated on two widely used tools: Paragraph
Punch and ProWritingAid. Data collection involved administering questionnaires,
writing assessments, and interviewing 53 undergraduate students. The findings
revealed that these digital tools had a positive impact on enhancing students' writing
skills, with a particular focus on the benefits of real-time feedback and structured
guidance. However, the study also highlighted several concerns. These included the
risk of students becoming overly reliant on digital tools, the ongoing need for teacher
feedback, and technical issues that might impede the learning process. The article
stresses the potential of digital tools to improve English as a Foreign Language (EFL)
writing skills but also calls attention to the importance of balancing technology with
traditional teaching approaches. This study offers valuable insights for educators and
researchers interested in incorporating technology into language learning. It provides
empirical evidence that outlines both the strengths and limitations of digital tools in
fostering writing proficiency in EFL contexts, emphasizing that technology alone is not
a complete solution for improving students' writing abilities.
Words in text: 205
AWL words: 19,02%
14
REFERENCES
Barrett, A., & Pack, A. (2023). Not quite eye to A.I.: Student and teacher perspectives
on the use of generative artificial intelligence in the writing process. International
Journal of Educational Technology in Higher Education, 20(1), 59.
https://doi.org/10.1186/s41239-023-00427-0
Garcia, R. A., Mayta, N. A., Bartesaghi, W. M., & Llapa, M. P. (2024). Exploration of
ChatGPT in basic education: Advantages, disadvantages, and its impact on
school tasks. Contemporary Educational Technology, 16(3), ep511.
https://doi.org/10.30935/cedtech/14615
Kim, J., Yu, S., Detrick, R., & Li, N. (2024). Exploring students’ perspectives on
Generative AI-assisted academic writing. Education and Information
Technologies. https://doi.org/10.1007/s10639-024-12878-7
Lee, Y.-J., Davis, R. O., & Lee, S. O. (2024). University students’ perceptions of
artificial intelligence-based tools for English writing courses. Online Journal of
Communication and Media Technologies, 14(1), e202412.
https://doi.org/10.30935/ojcmt/14195
Li, X., Li, B., & Cho, S.-J. (2023). Empowering Chinese Language Learners from Low-
Income Families to Improve Their Chinese Writing with ChatGPT’s Assistance
Afterschool. Languages, 8(4), 238. https://doi.org/10.3390/languages8040238
Pitukwong, K., & Saraiwang, S. (2024). Exploring the effectiveness of digital writing
tools on Thai EFL students’ writing. Contemporary Educational Technology,
16(3), ep519. https://doi.org/10.30935/cedtech/14808
Punar, N., & Yangın, G. (2024). Cultivating writing skills: The role of ChatGPT as a
learning assistant—a case study. Smart Learning Environments, 11(1), 10.
https://doi.org/10.1186/s40561-024-00296-8
Seo, J.-Y. (2024). Exploring the Educational Potential of ChatGPT: AI-Assisted
Narrative Writing for EFL College Students. Language Teaching Research
Quarterly, 43, 1–21. https://doi.org/10.32038/ltrq.2024.43.01
15
Tarchi, C., Zappoli, A., Casado Ledesma, L., & Brante, E. W. (2024). The Use of
ChatGPT in Source-Based Writing Tasks. International Journal of Artificial
Intelligence in Education. https://doi.org/10.1007/s40593-024-00413-1
Zou, M., & Huang, L. (2023). To use or not to use? Understanding doctoral students’
acceptance of ChatGPT in writing through technology acceptance model.
Frontiers in Psychology, 14, 1259531.
https://doi.org/10.3389/fpsyg.2023.1259531
16
