HO THI MY HANG USE OF AI CHATBOT IN DEVELOPING HIGH SCHOOL STUDENTS’ ENGLISH WRITING SKILLS MASTER OF ENGLISH LANGUAGE TEACHING Binh Dinh – 2024... HO THI MY HANG USE OF AI CHATBOT IN
INTRODUCTION
Rationale
English is one of the most widely spoken languages globally, with billions of speakers, making it essential for students to learn due to its global significance Mastering English opens doors to numerous opportunities in education, careers, and personal growth For high school students, proficiency in English is crucial as it lays the groundwork for future academic and professional success.
High school students develop essential language skills—listening, speaking, reading, and writing—within the English curriculum, with writing being particularly vital for their overall proficiency Early exposure to writing becomes increasingly important as students progress in their studies, with paragraph writing playing a crucial role This skill helps students organize their thoughts, formulate clear arguments, and communicate effectively Mastering well-structured paragraphs is essential not only for academic achievement but also for real-life problem-solving, as it enables students to present their ideas logically and persuasively (Wirantaka, 2016).
Many Vietnamese high school students encounter significant challenges in paragraph writing, struggling with the organization of their ideas and maintaining logical flow This often leads to frustration and negatively affects their writing proficiency and academic performance As an English teacher observing these difficulties, I have been seeking effective tools to help students enhance their writing skills both in and out of the classroom.
I have found was AI chatbot
In the era of rapid technological advancement, AI chatbots are increasingly utilized as writing aids, offering features such as grammar and spelling checks, idea suggestions, and examples Their effectiveness and user-friendly nature save individuals significant time and effort, making them valuable tools for enhancing writing productivity (Gayed et al., 2022; Jeanjaroonsri, 2023).
AI chatbots serve as valuable tools for enhancing English writing skills among EFL students These intelligent systems provide instant feedback and support, enabling students to refine their writing efficiently Specifically, learners can inquire about various aspects of their texts, including grammar, vocabulary, syntax, content, and structure, ultimately leading to improved writing proficiency (Hosseini et al., 2023; Strobl et al., 2019; Thorp).
2023), and tips on how to improve their writing skills (Akgun & Greenhow, 2022; Nazari et al., 2021; Rudolph et al., 2023)
Recent studies have explored the impact of AI tools on the writing skills of English as a Foreign Language (EFL) students Researchers have generally found that integrating AI-powered writing tools can enhance student writing abilities (Kurniati & Fithriani, 2022; Shi et al.).
While some researchers have expressed concerns about the potential side effects of AI in writing, such as plagiarism and inaccuracies, there is a notable gap in the literature regarding its impact on content and organization Most studies have primarily focused on AI's influence on grammar and syntax, overlooking the equally crucial aspects of writing This lack of comprehensive understanding hinders both students and educators from effectively leveraging AI to enhance writing skills.
This study aimed to fill that gap by providing a deeper look in how AI chatbot’s support can be used to improve all four aspects of students’ writings
To do it, I made use of ChatGPT, a famous AI Chatbot, as an assistant for my students during English paragraph writing practice.
Research Aim and Objectives
This research aims to study the use of ChatGPT as a tool to improve the English writing skills of Grade 10 students at Phan Boi Chau High School
1 To investigate the impact of ChatGPT in improving English writing skills of Grade 10 students at Phan Boi Chau High School
2 To explore the perceptions of Grade 10 students at Phan Boi Chau High School regarding the use of ChatGPT as a tool to develop their writing skills.
Research Questions
1 To what extent does the use of an ChatGPT improve the English writing skills of Grade 10 students at Phan Boi Chau High School?
2 What are the perceptions of Grade 10 students at Phan Boi Chau High School regarding ChatGPT as a tool to develop their writing skills?
Scope of the Study
The study focused on 90 Grade 10 students from Phan Boi Chau High School in Pleiku City, Gia Lai Province, Vietnam Conducted over a 12-week period, the research utilized Global Success 10, Book 2, starting from Unit 1.
This study investigates the influence of ChatGPT, a popular AI chatbot, on various elements of students' paragraph writing, including vocabulary, structure, grammar, creativity, idea development, and coherence and cohesion.
Significance of the Study
This study contributes to our understanding of AI technology's role in language learning, particularly in writing It highlights ChatGPT's influence on key writing elements, including vocabulary, structure, grammar, creativity, idea development, and overall coherence and cohesion.
This study highlights the educational advantages of AI chatbots for various stakeholders in the learning process For teachers, these chatbots can enhance lessons by providing students with additional practice and personalized feedback, thereby improving the overall learning experience Students benefit from AI chatbots as they offer an accessible means to receive support in developing their writing skills, fostering confidence and providing essential resources for improvement Additionally, the research findings serve as a valuable reference for future studies on the integration of AI chatbots in education, particularly in language learning, potentially paving the way for further exploration into how AI technology can address challenges in teaching and learning English writing.
Organization of the Study
This thesis consists of five chapters: Introduction, Literature Review, Methodology, Findings and Discussion and Conclusion
Chapter 1 introduces the rationale, aim and objectives, research questions, scope and significance of the study and the organization of the thesis
Chapter 2 provides theoretical backgrounds for the study, including what had been studied and applied about writing, paragraph writing and AI chatbots, especially in EFL education It also reviews previous studies related to the topic to highlight the research gap that this research intends to address
Chapter 3 provides the justifications for the research methods used in the study The materials used for the intervention, the research design description, the site to conduct and the participants are presented first The research tools that were used to collect data for the study are then described in detail The chapter is finally concluded with the data analysis methodology, the research procedure for conducting the study, and the research timeline
Chapter 4 gives a detailed presentation and analysis of collected data, then interpret and discuss the results to answer the research questions
Chapter 5 concludes the study report by summarizing the key findings It then relates them to previous research The chapter proceeds by suggesting implications of the research and concludes with a discussion of the study's limitations and recommendations for future research.
Chapter Summary
English writing skills are crucial for high school students' academic and professional success, yet many Vietnamese students face challenges in crafting coherent and logically flowing paragraphs This study explores the potential of AI chatbots, particularly ChatGPT, to enhance writing abilities by offering grammar checks, constructive feedback, and assistance in idea generation.
This study investigates the effects of ChatGPT on vocabulary, grammar, structure, creativity, and coherence among 90 Grade 10 students at Phan Boi Chau High School over a 12-week period, while also examining the students' perceptions of the tool.
This study seeks to fill the gaps in current research by exploring the significant role of AI in content development and organization It enhances the understanding of AI's impact on education while providing valuable insights for teachers, students, and researchers The thesis is structured into five chapters, encompassing the rationale, literature review, methodology, findings, and conclusions.
LITERATURE REVIEW
Writing
Writing has always been a very important way for people to communicate and has played a big role in keeping and saving information throughout history
Historically, individuals documented events to preserve knowledge for future generations, allowing us to comprehend and connect with the lives and experiences of those who lived long ago.
Writers and poets have historically utilized writing as a powerful medium to express their ideas and showcase their talents, a legacy that continues to resonate with audiences today Unlike spoken language, which can be easily altered or forgotten, the written word preserves essential information, ensuring its consistency and longevity throughout time.
Writing is a vital form of communication that extends beyond merely putting words on paper; it serves as a medium for expressing emotions, ideas, and beliefs (Harmer, 2004) According to Nunan (2003), writing involves a mental process of organizing thoughts into coherent written language, requiring careful planning to ensure clarity and meaning Harris (2001) further emphasizes that writing is a complex process that encompasses multiple stages, including planning, drafting, and revising, to achieve an accurate and well-structured final product.
Writing in EFL Classrooms
In EFL classrooms, writing is a crucial skill that students must develop, as it enhances their ability to think critically and use the English language effectively Unlike speaking, which allows for spontaneous expression, writing demands careful consideration of word choice and sentence structure This thoughtful process not only fosters creativity and self-expression but also deepens students' understanding of the language (Maley, 2009).
Writing is an essential skill that extends beyond school, playing a crucial role in students' future academic and professional lives In college, students frequently engage in writing essays, reports, and research papers, making effective communication with professors and peers vital Additionally, strong writing skills are indispensable in the workplace, where clear and professional communication is necessary for job applications, emails, and reports Mastering English writing enables individuals to articulate their ideas effectively, fostering collaboration and enhancing career success.
2.2.2.1 Definition of Paragraph Before students can effectively craft a paragraph, it is crucial for them to have a comprehensive understanding of it Many researchers have offered their definitions for what is perceived as a paragraph
According to Oshima and Hogue (2007), a paragraph serves as the essential unit of written organization, comprising a sequence of related sentences that communicate a singular idea This central idea is introduced by a topic sentence, elaborated upon with supporting sentences, and typically wrapped up with a concluding sentence.
Zemach and Rumisek (2003) highlight that a paragraph serves as a cohesive unit of thought, requiring both unity and coherence Unity ensures that all sentences are connected to the main idea, while coherence allows for a logical and smooth flow between sentences These essential qualities contribute to the clarity of the paragraph, making it significant both on its own and within the broader context of the text.
In the same vein, Kane (2000) highlights the role of paragraphs in guiding the reader through the writer's argument or narrative, describing them as
"islands of meaning" within a sea of text He notes that effective paragraphs use transitions to maintain cohesion, helping readers see how each idea connects to the next
Among the mentioned definitions, I found Oshima and Hogue (2007)’s to be the most specific and comprehensible Therefore, I chose it as the main basis for this study
2.2.2.2 Components of Paragraph According to Mayers (2006), a well- constructed paragraph is typically composed of three key parts: the introduction, the body, and the conclusion
The introduction contains the crucial topic sentence, which establishes the main idea of the paragraph Typically positioned at the beginning, this sentence can also be effectively placed in the middle or at the end to achieve specific effects or to direct the reader's attention.
The body of a paragraph is crucial as it contains supporting sentences that enhance the main idea through examples, details, explanations, and evidence This section allows the writer to thoroughly explore and elaborate on the topic, ensuring the reader gains a comprehensive understanding of the subject matter.
In conclusion, the final sentence of a paragraph reinforces the main idea, offering closure and highlighting the key takeaway for the reader By mastering these essential components, students can enhance their writing skills, creating clear, organized, and impactful paragraphs that effectively convey their intended messages.
2.2.2.3 Types of Paragraph As stated by Wali and Madani (2020), paragraphs can come in many types However, when it comes to writing in English class, three types are most practiced: Narrative, Descriptive and Expository Each of these types serves different purposes in writing, which reflect their distinct structures, contents, and ways to convey information
A narrative paragraph is designed to tell a story or recount an event, emphasizing a logical and chronological sequence of events It incorporates key elements such as characters, settings, actions, and dialogue to craft a vivid and engaging narrative Typically, these paragraphs build towards a climax or resolution, adding dynamism to the storytelling process.
A descriptive paragraph seeks to vividly illustrate a person, place, object, or scene through detailed sensory descriptions By focusing on sights, sounds, smells, tastes, and textures, these paragraphs create a rich and precise image in the reader's mind, allowing them to visualize and emotionally connect with the subject being portrayed.
An expository paragraph serves to explain or inform readers about a specific topic by delivering clear and factual content It is typically organized around a central idea, supported by relevant details and examples The language used is straightforward and factual, prioritizing clarity and logical structure for effective communication.
Recognizing the distinctions among various paragraph types is essential, as each influences the integration of AI chatbots in writing instruction differently This study specifically examined the use of AI chatbots in relation to expository paragraphs, aligning with the focus of my students' curriculum in the new English program at the time of research.
2.2.2.4 Paragraph Writing Paragraph writing refers to the process of constructing a clear, focused unit of writing that develops a single idea It involves organizing sentences in a way that logically supports the main point and helps the reader follow the writer’s thought process Effective paragraph writing includes starting with a clear topic sentence, followed by supporting details or explanations, and often concludes with a sentence that ties everything together (Oshima & Hogue, 2007) Furthermore, the use of smooth transitions between sentences plays a crucial role in maintaining the flow and coherence of the paragraph, ensuring that ideas are presented logically and sequentially
According to Zemach and Rumisek (2016), effective paragraphs are characterized by unity and coherence, where all sentences are interconnected and revolve around a central idea, enhancing clarity The structure of a paragraph should adapt to its purpose and audience; for instance, narrative paragraphs utilize vivid details and chronological order to convey a story, while expository paragraphs aim to inform or explain through facts and examples Additionally, argumentative paragraphs necessitate the presentation of a claim, supported by evidence and reasoning, ensuring that each sentence is closely tied to the overarching argument.
Perceptions of Writing
Perception is the process through which individuals organize and interpret sensory information to create a meaningful understanding of their environment This involves not only receiving sensory input but also interpreting it according to prior knowledge, experiences, and expectations As a subjective process, perception can lead to vastly different interpretations of the same stimulus, influenced by each person's unique background and cognitive framework.
Students' perceptions of writing significantly influence their attitudes, motivation, and engagement in the writing process Many view writing as a challenging skill that demands considerable time, practice, and cognitive effort How students interpret writing—whether as a form of self-expression, a difficult academic task, or a way to showcase knowledge—affects their willingness to engage with the writing process and ultimately impacts their success in developing writing proficiency.
Students' perceptions of writing are significantly influenced by their past experiences, teacher feedback, and the relevance of writing to their personal and academic objectives Supportive and constructive feedback fosters positive views of writing, helping students recognize it as a valuable skill Conversely, criticism and an emphasis on grades can lead to negative associations, making writing seem daunting and undesirable.
Motivation significantly influences students' views on writing When students engage in creative or reflective writing tasks that hold personal significance, they become more motivated and invested in their work This ownership leads to a positive perception of writing as an enjoyable and valuable endeavor, thereby increasing their commitment to improvement In contrast, students who perceive writing as a monotonous academic obligation often lack motivation, seeing it merely as a requirement rather than a meaningful pursuit.
Self-efficacy, or students' beliefs in their writing abilities, plays a crucial role in their perception of writing Those with high self-efficacy are more inclined to take risks and experiment with language, while students with low self-efficacy often experience anxiety and may shy away from challenging or creative writing tasks, hindering their development Therefore, fostering a supportive classroom environment is essential to help students build confidence in their writing skills.
Students' perceptions of writing are influenced by various factors such as past experiences, teacher feedback, motivation, and self-efficacy Educators can significantly enhance these perceptions by offering constructive feedback, designing engaging assignments, and cultivating a supportive atmosphere that encourages students to feel motivated and confident in their writing abilities.
AI Chatbot
Artificial intelligence (AI) is recognized as a pivotal technology of the 21st century, yet it lacks a universally accepted definition Despite ongoing discussions, researchers have not reached a consensus Notably, McCarthy et al (1956) characterized AI as the science of creating machines capable of performing tasks that typically require human intelligence Similarly, Nilsson (1998) defined AI as systems designed to solve problems and make decisions, while Fogel (1995) described AI as machines that operate autonomously, mimicking functions of the human brain.
AI is increasingly prevalent in our daily lives, seamlessly integrated into various online activities that transform our interaction with technology For instance, Google’s search engine utilizes AI to deliver relevant search results, while platforms like YouTube and Netflix leverage AI to recommend content tailored to user preferences These advanced recommendation systems analyze user behavior to suggest videos or shows that align with individual interests, significantly enhancing the overall user experience.
Virtual assistants such as Siri, Alexa, and Google Assistant exemplify the practical applications of AI in everyday life Utilizing natural language processing and machine learning, these tools effectively interpret and respond to voice commands, assisting users with tasks like setting reminders and answering inquiries Their convenience and diverse features have solidified their significance in the daily routines of countless individuals.
Self-driving cars, such as those created by Waymo, exemplify an advanced application of AI technology By integrating sensors, machine learning, and sophisticated decision-making algorithms, these vehicles can navigate various driving conditions autonomously, without human intervention The evolution of autonomous vehicles highlights AI's transformative potential in enhancing safety and efficiency within the transportation sector.
AI chatbots, such as ChatGPT and Gemini, exemplify advanced AI technology by engaging in human-like conversations They are equipped to provide information, answer questions, and assist with a variety of tasks, making them invaluable in sectors like customer service and education Their sophisticated language capabilities enhance their utility across multiple domains.
2.4.2 The Concept of AI Chatbot
An AI chatbot is a software program designed to simulate conversation with users through text-based chats, allowing for interactions that feel like speaking with a real person (Rahman et al., 2017) This technology utilizes natural language processing (NLP) and machine learning (ML) to comprehend and respond to human input effectively (Caldarini et al., 2022).
AI chatbots are versatile tools capable of engaging users on a variety of topics, from product inquiries to health advice and weather updates Their responses are shaped by a programmed knowledge base and the conversation's context, ensuring relevance and a human-like interaction As highlighted by Deshpande et al (2017), the main objective of AI chatbots is to deliver personalized and accurate information, ultimately enhancing user experience and satisfaction.
AI chatbots excel in generating text based on user input, making them versatile tools for drafting emails, creating reports, and even crafting stories or essays Their ability to produce coherent and contextually relevant content enhances their utility across various sectors, such as customer service, technical support, education, and content creation As highlighted by Xu et al (2019), this flexibility enables chatbots to function effectively in settings with limited human resources, offering real-time assistance and minimizing wait times for customer support.
AI chatbots leverage advanced algorithms and machine learning to enhance their performance by learning from previous interactions This continuous improvement allows them to adapt to individual user preferences, resulting in more accurate and context-aware responses By integrating natural language processing (NLP) and machine learning, these chatbots can better understand user input and engage in dynamic conversations, ultimately boosting user engagement and satisfaction.
In summary, AI chatbots mark a major leap forward in human-computer interaction, effectively mimicking human conversation and delivering information Their text generation capabilities and ongoing enhancements through learning algorithms render them indispensable across various applications From enhancing customer service to streamlining content creation, AI chatbots are revolutionizing our engagement with technology.
AI chatbots have evolved significantly over the past fifty years, marking substantial advancements in artificial intelligence and machine learning The journey began with ELIZA, the first AI chatbot developed by Joseph Weizenbaum in the 1960s, which employed simple techniques to mimic therapeutic conversations Although ELIZA's capabilities were rudimentary compared to modern chatbots, it represented a pivotal breakthrough in demonstrating that machines could engage in human-like dialogue Today, AI chatbots have progressed rapidly and are widely utilized across various sectors.
2.4.3.1 In Other Fields In other fields such as healthcare and business,
AI chatbots have been deployed for various purposes
AI chatbots in healthcare assist doctors by analyzing patient-reported symptoms to provide preliminary advice and predict health risks based on user input and historical data This proactive approach enables better health management before issues escalate Additionally, these chatbots enhance patient care by suggesting treatment plans and sending reminders, ultimately increasing the efficiency of healthcare delivery (King, 2023).
AI chatbots enhance user experiences in various business sectors by offering personalized recommendations for food and drinks, particularly in restaurants and hotels They assist job seekers by navigating the application process, providing career guidance, and matching them with appropriate job openings In customer service, chatbots efficiently manage inquiries, resolve issues, and offer advice, which alleviates the burden on human staff and ensures seamless service Additionally, in the banking sector, AI chatbots facilitate account management, process requests, and deliver financial advice, streamlining operations for both customers and institutions.
2.4.3.2 In Education With the increasing use of technology in education,
AI chatbots are increasingly recognized for enhancing educational outcomes through personalized support and efficient management of administrative tasks Their capacity to create engaging and interactive learning experiences has garnered significant interest from both researchers and educators Numerous studies have explored the impact of AI chatbots on teaching and learning processes.
Recent studies highlight the benefits of AI chatbots in education, particularly in language learning Han et al (2022) found that these tools enhance student engagement and language proficiency through immediate feedback and interactive practice Similarly, Sandu and Gide (2019) demonstrated that AI chatbots assist teachers by managing grading and addressing routine inquiries, which enables educators to focus more on personalized instruction and student support.
ChatGPT
Selecting the right AI chatbot is essential for generating valuable and reliable study results Among the various options available, including Google’s Gemini, Microsoft’s Copilot, IBM watsonx Assistant, and Rasa, ChatGPT stands out as the most suitable choice.
ChatGPT is a highly advanced conversational agent known for its impressive capabilities and strong user engagement According to Wu et al (2023), it excels in providing detailed and contextually relevant responses, making it suitable for various functions such as language translation, code debugging, creative writing, and summarizing academic papers Its versatility and accuracy in handling diverse tasks position it as an invaluable tool for users.
The selection of ChatGPT for this study is driven by its widespread recognition in Vietnam, ensuring that both students and educators are likely familiar with the platform, which promotes easier integration into the research Its accessibility, being free and available to anyone with an internet connection, removes financial barriers, allowing all participants to engage with the chatbot Furthermore, ChatGPT's user-friendly design necessitates minimal guidance for students, thereby enhancing the efficiency of the learning process.
In conclusion, ChatGPT is an ideal choice for this research Its broad recognition, accessibility, and ease of use align well with the study’s objectives
By incorporating ChatGPT, the research is likely to yield relevant and meaningful results, offering valuable insights into the ability of AI chatbots in enhancing EFL writing skills.
Related Studies
In the past ten years, research has focused on the potential of ChatGPT to enhance the writing abilities of English as a Foreign Language (EFL) students Findings indicate various methods through which ChatGPT contributes to writing improvement, highlighting both students' perceptions and measurable advancements in their writing skills.
A study by Sumakul et al (2022) explored students' perceptions of using ChatGPT in writing classes, focusing on eight EFL students from an Indonesian university The findings revealed that students had a positive experience with ChatGPT, describing it as fun and helpful They particularly appreciated the interactive nature of the chatbot, which enhanced their learning experience and made it more engaging and effective.
A study by Silitonga et al (2023) explored the impact of ChatGPT on enhancing Malaysian college students' motivation for English writing Involving 73 participants, the quasi-experimental research revealed that ChatGPT significantly increased students' motivation The chatbot provided detailed feedback, enabling students to identify their strengths and areas for improvement, thereby facilitating their efforts to overcome writing challenges.
Kwon et al (2023) conducted a study on the effectiveness of ChatGPT in enhancing second language (L2) writing skills among 75 Korean elementary school students learning English Utilizing pre-test and post-test assessments, the findings indicated that ChatGPT significantly outperformed traditional teaching methods in improving students' writing abilities Additionally, student interviews revealed that they found the chatbot to be both useful and comfortable, highlighting their engagement and satisfaction with AI-assisted learning.
In Vietnam, the integration of AI chatbots, particularly ChatGPT, into English as a Foreign Language (EFL) writing instruction has garnered significant interest Phan (2023) conducted interviews with students at Vietnam National University, revealing that they appreciated the accessibility, adaptability, and simplicity of ChatGPT, indicating a willingness to incorporate AI tools into their learning Similarly, Nguyen (2023) interviewed EFL teachers at Van Lang University, with most expressing a positive outlook on using ChatGPT, believing it could enhance the teaching process This highlights a growing acceptance of AI in educational settings in Vietnam.
Research indicates that teachers and students in Vietnam are generally prepared to utilize AI chatbots in writing classes; however, existing studies primarily emphasize their perceptions rather than exploring the actual effects on students' writing skill development This highlights a significant gap that my study aims to address.
Chapter Summary
This chapter explores the significance of writing in EFL classrooms, highlighting its historical importance and impact on students' academic and personal growth It emphasizes the teaching of paragraph writing, focusing on essential elements such as grammar, structure, creativity, and coherence The chapter also addresses the importance of effective assessment practices and the integration of digital tools, including AI, to enhance writing skills and offer personalized support Furthermore, it underscores the role of students' perceptions of writing, which significantly influence their motivation and success, and suggests that constructive teacher feedback and supportive learning environments can positively shape these perceptions.
In conclusion, the chapter underscores the importance of writing in EFL education, addressing challenges students face and highlighting the role of technology and positive perceptions in improving writing skills.
METHODOLOGY
Research Design
This study employed a quasi-experimental design with pre- and post-tests to evaluate the impact of ChatGPT on students' writing skills, enabling the collection of realistic data without hindering the learning process (Cook & Campbell, 2007) Participants were split into two groups: the experimental group received writing lessons enhanced by ChatGPT, while the control group underwent traditional writing instruction This structure facilitated a direct comparison between the two groups, highlighting the effectiveness of ChatGPT in enhancing writing skills.
This study employed a mixed-methods approach, utilizing both quantitative and qualitative data collection to address the research questions Quantitative data were gathered through pre- and post-tests that assessed the writing skills of Grade 10 students, focusing on key areas such as vocabulary, spelling, grammar, coherence, cohesion, idea development, and creativity These tests provided a clear measure of the students' progress Additionally, questionnaires were administered to the experimental group to capture their perceptions of ChatGPT's impact on their writing For qualitative insights, semi-structured interviews were conducted with selected students from the experimental group, offering a deeper understanding of their interactions with the chatbot, its effect on their motivation, and the overall enhancement of their writing skills.
Utilizing both quantitative and qualitative methods enhances the understanding of the research problem by offering a comprehensive perspective As noted by Abowitz and Toole (2010), integrating multiple data sources yields richer and more reliable data, thereby improving the validity of the findings This study combines data from tests, questionnaires, and interviews to provide a well-rounded analysis of the impact of ChatGPT on writing teaching and learning.
Research Setting and Participants
The study took place at Phan Boi Chau High School in Pleiku, Gia Lai Province It was chosen for several key reasons
As a teacher with in-depth knowledge of the school's academic environment, challenges, and strengths, I leveraged my insider perspective to design and conduct a relevant study This familiarity was crucial in ensuring that the research findings would be practical and beneficial.
Phan Boi Chau High School in Gia Lai Province is highly regarded for its academic excellence and serves as a benchmark for other local schools, making it an ideal environment for experimenting with innovative teaching methods My extensive experience at the school facilitated participant recruitment and collaboration with school administrators, ensuring the smooth execution of the study.
The school boasts modern facilities that enhance educational activities, featuring spacious classrooms equipped with advanced technology to facilitate learning The highly qualified staff and experienced teachers foster a supportive environment, which has been crucial to my academic success.
With over twenty-one years of teaching experience, I effectively instructed both the experimental and control groups during the study My background in adapting teaching methods to accommodate diverse learning capabilities fostered a sense of familiarity and trust among the students This established rapport enhanced their comfort and cooperation throughout the research process, ultimately improving the accuracy of the findings.
The study focused on 90 Grade 10 students from Phan Boi Chau High School, specifically selected from two of my classes due to my familiarity with their learning styles and academic strengths Most participants demonstrated above-average English skills, particularly in writing, enabling an effective evaluation of ChatGPT's impact across varying levels of proficiency.
In a study comparing traditional English instruction to instruction enhanced by ChatGPT, participants were randomly divided into two equal groups: a control group and an experimental group, each consisting of 45 students The control group, comprising 26 females and 19 males, received conventional English teaching without ChatGPT, while the experimental group, made up of 25 females and 20 males, utilized ChatGPT in their writing practice A summary of the participants' demographics is presented in Table 1.
Group Total of students Gender
Years of studying English Control 45 26 females 19 males 9 years
To streamline data processing, I assigned codes to students, designating them as CS for the control group and ES for the experimental group, followed by a number indicating their order within each group (e.g., CS1 for the first student in the control group and ES45 for the last student in the experimental group).
Textbook
To provide formal face-to-face instruction in English for the participants,
I used the coursebook Global Success 10, designed for the General Education Programme 2018 In the second term, grade 10 students had to learn 5 units and
2 reviews Each unit includes 8 lessons: Getting Started, Language (Vocabulary, Pronunciation, Grammar), Reading, Speaking, Listening, Writing, Communication and Culture/CLIL, and Looking Back and Project
6 Gender equality Writing about jobs for men and women
Writing about Vietnam’s participation in international organisations
8 New ways to learn Writing about the benefits of blended learning
Table 2 illustrates that the study comprised a total of four units, each dedicated to different writing topics In Unit 6, centered on gender equality, students explored the theme of "jobs for men and women." Meanwhile, Unit 7 focused on Vietnam's relationship with international organizations, prompting students to engage in writing about this subject.
“Vietnam’s participation in international organization) For Unit 8 (New ways to learn), the topic was “the benefits of blended learning” And for Unit 9 (Protecting the environment), students explored “wildlife organizations”.
Research Instruments
To gather comprehensive data on the impact of the AI chatbot on student writing improvement, I employed three key tools: pre- and post-tests, a questionnaire, and semi-structured interviews These methods provided a holistic perspective on the effectiveness of the chatbot in enhancing students' writing skills.
3.4.1 Pre-test and Post-test
Pre- and post-tests are essential tools in educational research for evaluating the effectiveness of interventions As noted by Creswell and Guetterman (2019), these assessments enable a straightforward comparison of learning outcomes by measuring performance before and after the intervention Implementing pre- and post-tests allows researchers to gain valuable insights into the impact of educational strategies.
I can determine if ChatGPT helped improve the students’ writing skills
To assess the impact of ChatGPT on students' paragraph writing skills, pre-test and post-test evaluations were conducted The pre-test, administered at the study's onset, focused on the topic of "teacher's job" from Unit 6, establishing a baseline for students' writing abilities In contrast, the post-test, given at the conclusion of the study, required students to articulate their understanding of "blended learning," a concept introduced in Unit 8 Both assessments required students to compose paragraphs of approximately 120 words, enabling a direct comparison of their writing skills before and after the intervention.
In designing the tests, I adhered to the recommendations from Hughes (2020) and Brown and Abeywickrama (2019), which offer comprehensive guidelines on aligning test topics with specific educational objectives and student proficiency levels.
Table 3 Rubric for assessing paragraph writing
1.15-1.43 Excellent Vocabulary Poor vocabulary, incorrect word use
Limited vocabulary, some incorrect word use
Adequate vocabulary, occasional incorrect word use
Good vocabulary, rare incorrect word use
Rich vocabulary, accurate and varied word use
Grammar Frequent grammar mistakes that obstruct understanding
Some grammar mistakes, sentences may be unclear
Adequate grammar, mistakes occasionally affect clarity
Few grammar mistakes, mostly clear and effective sentences
Consistently accurate grammar, high clarity
Structure Unclear organization, no clear introduction, body or conclusion
Weak structure, some attempts at organizing ideas but often unclear or confusing
Adequate structure, ideas are generally organized with occasional lack of clarity
Good structure, ideas are logically organized with clear introduction, body and conclusion
Excellent structure, clear and logical flow of ideas with well- defined sections
Spelling Frequent spelling errors that interfere with understanding
Some spelling errors that affect clarity
Occasional spelling errors but meaning is generally clear
Few spelling errors, writing is mostly clear and easy to follow
Almost or completely no spelling errors, writing is polished and professional
Idea Ideas are vague, unsupported or lack development
Ideas are basic and lack depth, minimal supporting details
Ideas are sufficiently developed with adequate supporting details
Ideas are well developed with clear and relevant supporting details
Ideas are thoroughly developed, rich in details, depth and insight
Creativity Lack originality or engagement, overly simplistic
Limited creativity, predictable or overused ideas
Some creative elements, shows some originality but lacks impact
Creative and original, engages the reader with interesting or new ideas
Highly creative and original, writing is engaging, thought provoking and imaginative
Writing is scattered, ideas don’t connect logically
Limited coherence, ideas are connected but transitions are unclear or awkward
Ideas are generally cohesive, some transitions could be smoother
Coherent and cohesive writing, ideas flow logically with effective transitions
Highly coherent and cohesive, ideas flow seamlessly with skillful use of transitions
The rubric detailed in Table 3, inspired by the works of Rohim (2019), Rukmini and Saputri (2017), and Brown (2007), evaluates seven key aspects of student writing: vocabulary, grammar, structure, spelling, idea development, creativity, and cohesion and coherence Each criterion is assessed on a five-level scale ranging from terrible (0.0 to 0.29 points) to excellent (1.15 to 1.43 points), with a maximum possible score of 10 and a minimum of 0 Following the intervention, I aim for students to improve by at least one level in each category; for instance, a student who begins at a decent level in vocabulary and poor in grammar should progress to at least a good level in vocabulary and a decent level in grammar.
To explore students' perceptions of AI chatbots, a series of questionnaires were administered, enabling students to express their views on the chatbot's usability, engagement, and helpfulness As noted by Cohen, Manion, and Morrison (2002), questionnaires are an effective method for rapidly and efficiently gathering data from a large population, making them particularly suitable for assessing students' overall attitudes towards the chatbot.
Before I gave away this set of questionnaires for participants to fill out, I had it go through a proper process of developing and improving It was designed following the guidelines of Boynton and Greenhalgh (2004) and Payne (2012) and included two parts The first five items asked for students’ personal information The next 19 items inquired about students’ views on what aspects of their writings have been improved by ChatGPT and how much has they been improved The aspects include vocabulary (3 items), grammar (3 items), structure (3 items), spelling (2 items), idea development (3 items), creativity (3 items), and cohesion and coherence (2 items)
To assess student feedback, a 5-point Likert scale was employed, featuring responses ranging from Completely Agree (SA) to Completely Disagree (SD) This scale effectively captures varying degrees of agreement, allowing for nuanced insights into student opinions.
Table 4 Ranges of the Likert scale
After designing the questionnaires, I sought feedback from five colleagues, which led to revisions and translations The revised questionnaires were then sent to teachers for a final review, followed by necessary adjustments The finalized version was translated into Vietnamese and pilot tested with 10 non-participant students, resulting in final adjustments based on their feedback.
To gain deeper insights into the students' questionnaire responses, I opted for interviews, a method recognized for enabling researchers to delve into individuals' experiences in their natural environments (Alshenqeeti, 2014) Through one-on-one discussions with my students, I aimed to uncover richer details and perspectives.
Semi-structured interviews are particularly effective for understanding students' beliefs about the impact of ChatGPT on their writing, as they facilitate a flexible yet structured exploration of various topics (Jamshed, 2014) Unlike structured interviews, which may restrict the depth of responses, or unstructured interviews, which can yield overly diverse answers that complicate data comparison (Bryman, 2016), semi-structured interviews strike an ideal balance for gathering meaningful insights.
In the interviews conducted, each student was asked three key questions based on Phan's (2023) recommendations The first question focused on the aspects of their studies that concerned them and how ChatGPT contributed to their improvement in those areas The second question assessed their satisfaction and dissatisfaction with ChatGPT's assistance, along with their reasons for these feelings Lastly, the third question explored the challenges students encountered while using ChatGPT To facilitate open and comfortable communication, the interviews were conducted in Vietnamese, allowing students to express their experiences more naturally and accurately.
I recorded the interviews on my phone to accurately capture the conversations and ensure that no critical information was overlooked This qualitative data enriches the quantitative results from tests and questionnaires, offering a comprehensive understanding of how students interact with ChatGPT in their writing.
Data Collection
The research was divided into three stages and lasted for 12 weeks in total
Stage 1 involved using the pre-test to gather data on the students' initial writing skills The students were chosen from 2 classes that I taught at my high school In the first two weeks, I taught them writing using traditional methods, covering content of Unit 6 ("Gender Equality") In the third week, I organized a sudden pre-test to assess their initial writing performance The following steps were taken to administer the tests: (a) distribute the tests to participants, (b) explain the requirements, (c) collect the tests, and (d) provide marks and comments the following week Two experienced EFL teachers graded the tests – a university lecturer with a doctor’s degree and a high school teacher with a master’s degree Both had great English language skills and were objective in their scoring
Stage 2 included the intervention, which lasted six weeks During this time, the intervention included teaching the textbook content from Unit 7 ("Viet Nam and International Organizations"), Unit 8 ("New Ways to Learn"), and Unit 9 ("Protecting the Environment") The students in the control group continued learning with only traditional teaching method (the teacher marking student work and providing feedback) and tools such as blackboard, group writing sheet, slide, … The students in the experimental group, besides the listed methods and tools, were allowed to use phones and access ChatGPT I introduced the AI chatbot to the students and instructed them on how to use it myself
Stage 3 was comprised of the post-test, the questionnaire and the interview In the tenth week, both the control and experimental groups took a post-test to assess their English writing skills after the intervention The content for the test was a topic from Unit 8, as mentioned earlier The post-test was also a surprise, and it followed the same process as the pre-test for administration and scoring A week after that, I gave out the questionnaires to the students in the experimental group and allowed them some time to complete it I had them do it online through Google Forms for convenience In the last week, I interviewed 6 of them to get a better interpretation of the questionnaires’ answer Their codes were ES5, ES7, ES13, ES20, ES29 and ES36 I did not select students randomly, but only focused on those that stood out in the tests and the questionnaires To help readers visualize the whole data collection process, I summarized all three stages into the table below
Table 5 Stages of data collection
Stage Week Content Teaching method
AI chatbot’s assistance writing practice
6 Writing Unit 8: New Ways to Learn
8 Writing Unit 9: Protecting the Environment
Data Analysis
The three sources of data were analyzed using different methods of data analysis, which will be presented in detail below
3.6.1 Pre-test and Post-test
After evaluating the tests based on my custom rubric, I compiled the scores into Microsoft Excel 2016 and employed both independent sample T-tests and paired sample T-tests to analyze the pre-test and post-test scores This approach was chosen to effectively compare the results and assess the impact of the intervention on student performance.
Descriptive analysis simplifies extensive data sets into a more digestible format, enabling researchers to swiftly recognize overarching patterns and trends (Gravetter & Wallnau, 2016) This approach was utilized to summarize and highlight the key characteristics of the quantitative data, providing readers with a clearer understanding of the test results.
A paired sample T-test is employed to analyze the means of two related groups, assessing whether a statistically significant difference exists between them This method is particularly suitable when the same subjects are evaluated twice under varying conditions (Gravetter & Wallnau 2016) In essence, I compared the pre-test and post-test scores for each group using this test, enabling me to determine the impact of the intervention on their performance.
An independent sample T-test is utilized to compare two independent groups, where participants in one group are not related to those in the other This statistical method effectively assesses whether there is a significant difference between the means of the two groups (Field 2024, p.315) In my study, I applied this test to evaluate the post-test scores of both control and experimental groups to determine which group exhibited greater improvement.
Utilizing a combination of three analytical methods allows for a thorough examination of group differences, temporal changes, and overarching data trends, leading to a comprehensive understanding of the intervention's effects Following the analysis, I presented the T-test findings through tables and poll charts to effectively showcase the variations and improvements in student performance.
Using the data from the questionnaires, I conducted a descriptive analysis to evaluate how respondents rated each item and the variability of their answers This method revealed clear patterns and central trends in the responses, highlighting overall tendencies I utilized Google Forms, which automatically generated pie charts to summarize common responses, eliminating the need for Microsoft Excel 2016 Subsequently, I organized the processed data into tables for easier comparison based on the previously discussed aspects.
The audio recordings from semi-structured interviews were transcribed into written scripts (Appendix E1 and E2) Thematic analysis was then employed to generate a comprehensive overview of students' perceptions and identify patterns in their responses Ultimately, I compared the analyzed qualitative data with questionnaire and test results to gain a deeper insight into the quantitative findings.
Chapter Summary
This chapter outlines the research methods employed to assess the impact of ChatGPT on students' writing skills Utilizing a quasi-experimental design, the study involved a pre- and post-test comparison between two groups of 90 Grade 10 students at Phan Boi Chau High School The control group participated in traditional writing lessons, while the experimental group utilized ChatGPT as a tool to enhance their writing practice.
Over a 12-week period, data was collected in three stages to assess the impact of ChatGPT on students' writing abilities Initially, students underwent traditional instruction and a pre-test to evaluate their writing skills During the intervention phase, the experimental group utilized ChatGPT in their lessons, while the control group maintained traditional teaching methods Finally, both groups participated in a post-test to gauge improvement, and the experimental group provided feedback through questionnaires and interviews regarding ChatGPT's influence on their writing.
The research employed a mixed-methods approach, utilizing both quantitative and qualitative techniques Quantitative data from tests were examined using descriptive statistics and T-tests to uncover performance differences, while qualitative data gathered from questionnaires and interviews underwent thematic analysis, offering a deeper understanding of students' experiences This combination of methods delivered a thorough perspective on the impact of ChatGPT on students' writing skill development.
FINDINGS AND DISCUSSION
The Impact of ChatGPT on Grade 10 Students’ English Writing Skills
This section presents the analysis of pre-test and post-test scores using both paired and independent sample T-tests The paired sample T-test assesses improvements within each group, determining the effectiveness of the intervention, while the independent sample T-test compares the score changes between the two groups to identify which group experienced greater improvement.
4.1.1.1 Control Group Table 6 reports the result of the paired sample T- test done for the control group’s pre-test and post-test scores
Table 6 Control group’s paired sample T-test
In the control group, the average pre-test score was 6.39, which rose slightly to 6.6 in the post-test Although this indicates a minor improvement in students' scores after the study period, the increase is minimal and statistical analysis reveals that it is not significant enough to be deemed meaningful.
The pre-test scores exhibited a variance of 1.54, indicating a moderate spread around the average score In contrast, the post-test scores showed a slightly higher variance of 1.65, reflecting a consistent yet modest increase in student performance across the group This suggests that while individual scores improved, the overall distribution of scores remained relatively stable.
The paired sample T-test was conducted to assess the statistical significance of the difference between pre-test and post-test scores, yielding a p-value of 0.066 This p-value indicates that the observed change is not statistically significant, as it exceeds the conventional threshold of 0.05 Consequently, any increase in scores is likely attributable to chance rather than a genuine enhancement in student performance.
Figure 1 Pre-test of a student from control group
Figure 2 Post-test of the same student from control group
The T-statistic value of -1.89 indicates a minor difference between pre-test and post-test scores in relation to standard errors, suggesting that the difference is not statistically significant With 44 degrees of freedom, the precision of the results is enhanced; however, the findings still do not achieve statistical significance, reinforcing the conclusion that the observed difference is small.
The control group showed no significant improvement in scores between the pre-test and post-test, indicating that their performance remained largely unchanged The minimal variation observed could likely be attributed to chance, suggesting that the study had no substantial impact on the students in the control group.
In summary, the control group exhibited no significant changes in their scores from pre-test to post-test, demonstrating that without targeted interventions, students' performance in writing skills remained largely unchanged This finding highlights that traditional instructional methods may not be sufficient for substantial improvement, supporting previous research that emphasizes the need for specific interventions to enhance writing abilities (Cohen et al., 2013).
4.1.1.2 Experimental Group Table 7 shows the result of the paired sample T-test done for the experimental group’s pre-test and post-test scores
Table 7 Experimental group’s paired sample T-test
The experimental group demonstrated a notable improvement in performance, with pre-test scores averaging 6.48 and post-test scores rising to 8.29, resulting in an increase of 1.61 A paired T-test was conducted to evaluate the statistical significance of this change, and the results confirmed that the improvement was statistically significant.
The pre-test scores exhibited a variance of 1.64, indicating a moderate spread around the mean Following the intervention, the variance of the post-test scores dropped to 0.88, reflecting increased consistency in student performance This outcome suggests that the intervention not only enhanced average scores but also fostered more reliable performance among students.
Figure 3 Pre-test of a student from experimental group
Figure 4 Post-test of the same student from experimental group
The Pearson correlation coefficient between pre-test and post-test scores was 0.74, indicating a strong positive correlation This suggests that students who excelled in the pre-test also showed significant improvement in the post-test, highlighting that their initial abilities effectively predicted their progress following the intervention.
The T-statistic of -12.59 indicates a significant difference between pre-test and post-test scores With a p-value of 3.53E-16, which is far below the 0.05 threshold, the results confirm that the observed improvement in scores is statistically significant, suggesting that the enhancement is unlikely to have occurred by chance.
In a two-tailed test with 44 degrees of freedom, the T-critical value was found to be 2.02 The calculated T-statistic of -12.59 is significantly lower than this critical value, leading to the rejection of the null hypothesis, which posits no significant difference between pre-test and post-test scores This outcome reinforces the conclusion that the intervention resulted in a notable enhancement in student performance.
The experimental group demonstrated a significant improvement in scores following the intervention, with a notable decrease in variance and a strong positive correlation indicating enhanced consistency in performance The highly significant T-test results confirm that this improvement is attributable to the intervention rather than random chance.
In summary, the experimental group showed a notable increase in scores from pre-test to post-test, indicating that the intervention positively influenced student performance The statistically significant improvement highlights the tangible impact of the intervention, underscoring the effectiveness of AI tools like ChatGPT in delivering personalized feedback and enhancing interactive learning experiences, as supported by Ghafouri et al.
(2024) findings for the same topic
4.1.2.1 Pre-test Results Table 8 illustrates the result of the independent sample T-test conducted for the pre-test scores of the control group and the experimental group
Table 8 Pre-test results’ independent sample T-test
Group Mean Variance df t Stat p-value t Critical Control 6.39 1.54 88 -1.09 0.28 1.99
An independent sample T-test was performed to analyze the pre-test scores of both the control and experimental groups The control group recorded a mean pre-test score of 6.39, whereas the experimental group achieved a slightly higher mean score of 6.48 This indicates that the experimental group outperformed the control group on the pre-test, though the difference in scores was minimal.
Students’ Perceptions of ChatGPT’s Impact on Their Writing
This section presents the results of the questionnaires using pie charts, a feature available in Google Forms, to analyze students' responses regarding the impact of ChatGPT on various aspects of their writing.
4.2.1.1 On Vocabulary Table 10 reports three items regarding students’ perceptions of ChatGPT’s impact on their vocabulary use in writing
Table 10 Students’ perceptions of ChatGPT’s impact on vocabulary
1.1 ChatGPT helped me learn new vocabulary related to my writing topics
1.2 ChatGPT suggested similar or opposite words that enriched my writing vocabulary
1.3 I understood the meaning and usage of difficult words better after ChatGPT explained them
A recent survey revealed that 75.5% of students found ChatGPT helpful in acquiring vocabulary related to their writing topics, with 44.4% fully agreeing and 31.1% somewhat agreeing In contrast, only 13.3% remained neutral, while 11.2% expressed disagreement, indicating minimal discontent with ChatGPT's effectiveness in vocabulary acquisition Additionally, 84.5% of students felt that ChatGPT provided useful synonyms and antonyms to enhance their vocabulary, with 37.8% completely agreeing and 46.7% somewhat agreeing However, 13.3% completely disagreed with this feature's usefulness, while only 2.2% somewhat disagreed, highlighting a significant divide in opinions on its effectiveness.
A survey investigated whether students improved their understanding of difficult vocabulary through ChatGPT's explanations The findings indicated that 53.4% of respondents felt that ChatGPT enhanced their comprehension, with 35.6% somewhat agreeing and 17.8% fully agreeing However, a notable percentage remained neutral (24.4%), while 22.2% expressed disagreement, including 13.3% who somewhat disagreed and 8.9% who completely disagreed.
In summary, students demonstrated confidence in ChatGPT's ability to enhance vocabulary acquisition by providing new words and relevant synonyms, aligning with research that indicates AI tools can effectively enrich lexical knowledge (Chen, 2024) Nonetheless, a minority of students expressed indifference or dissatisfaction, suggesting that ChatGPT's effectiveness may vary based on individual user needs and preferences, as highlighted by Wang et al (2024), who noted diverse responses among learners utilizing AI for vocabulary improvement.
4.2.1.2 On Grammar Table 11 reports three items regarding students’ perceptions of ChatGPT’s impact on their grammar use in writing
Table 11 Students’ perceptions of ChatGPT’s impact on grammar
2.1 ChatGPT helped me notice and correct common grammatical mistakes in my writing
2.2 I improved my understanding of grammar points after ChatGPT explained them
2.3 ChatGPT’s grammar suggestions helped me write more accurate and complex sentences
A survey on ChatGPT's effectiveness in helping students identify and correct grammar errors revealed that 62.2% of respondents believe the tool is beneficial, with 37.8% completely agreeing and 24.4% somewhat agreeing However, 22.2% of students remained neutral, indicating that some do not see significant improvements in grammar correction Additionally, 15.6% of participants expressed disagreement with ChatGPT's utility for this purpose, with 11.1% completely disagreeing and 4.5% somewhat disagreeing.
A survey evaluating students' understanding of grammar after using ChatGPT revealed that 33.3% completely agreed that the tool provided clearer insights, while 26.7% somewhat agreed, resulting in 60% of respondents finding the explanations effective to varying extents However, 31.1% remained neutral, indicating that some students did not perceive significant benefits Additionally, 8.9% completely disagreed with the effectiveness of ChatGPT's grammar explanations, although no respondents somewhat disagreed, reflecting a minor but notable negative perception of the tool's utility.
The third question assessed students' opinions on the effectiveness of ChatGPT's grammar suggestions in enhancing sentence accuracy and complexity The findings revealed that 24.4% of respondents fully agreed and 33.3% somewhat agreed, indicating a generally positive perception, with 57.7% finding the tool beneficial However, 35.6% of participants remained neutral, reflecting indifference towards the suggestions' efficiency Only a small fraction, 4.4%, completely disagreed, and 2.3% somewhat disagreed, suggesting limited dissatisfaction with the tool.
In conclusion, students largely found that ChatGPT was beneficial for identifying common grammatical errors, clarifying complex grammar concepts, and assisting in the creation of more precise and sophisticated sentences This aligns with Kim's (2019) research findings.
AI tools like ChatGPT can aid in grammar learning by offering immediate corrections and explanations However, the mixed responses from some students indicate that the effectiveness of these tools may be constrained by their prior grammar knowledge and the complexity of the errors they are trying to correct (Nindya & Taufiqulloh, 2024).
4.2.1.3 On Structure Table 12 reports three items regarding students’ perceptions of ChatGPT’s impact on their writing’s structure
Table 12 Students’ perceptions of ChatGPT’s impact on structure
3.1 ChatGPT’s feedback helped me structure my paragraphs more effectively
3.2 I knew how to write better topic and concluding sentences thanks to ChatGPT
3.3 ChatGPT helped me arrange my ideas in a more logical order
A survey on ChatGPT's effectiveness in enhancing students' paragraph construction revealed that 35.6% of respondents completely agreed with its benefits, demonstrating a strong belief in the tool's efficiency Additionally, 24.4% somewhat agreed, while another 24.4% expressed neutrality, indicating a varied perception of its impact In contrast, only 11.1% somewhat disagreed, and a mere 4.5% completely disagreed, highlighting a low level of dissatisfaction with the feedback provided by ChatGPT.
A survey revealed that 64.4% of students felt ChatGPT improved their skills in creating effective opening and concluding sentences, with 33.3% completely agreeing and 31.1% somewhat agreeing However, 24.4% of respondents remained neutral, while 11.2% expressed disagreement, highlighting that not all students found the tool beneficial for enhancing their writing abilities.
The third question examined whether ChatGPT assisted in organizing ideas logically Both 31.1% of respondents completely agreed and remained neutral, suggesting equal feelings of clarity in organizing thoughts
Additionally, 20% somewhat agreed, while 11.1% completely disagreed and 6.7% somewhat disagreed, indicating that a minority felt ChatGPT was ineffective in this area
Students have expressed favorable views on ChatGPT’s feedback concerning paragraph organization and the formulation of topic and concluding sentences This perspective is supported by Marzuki et al (2023), who highlight the role of AI in aiding learners to organize their thoughts and enhance the coherence of their writing Additionally, ChatGPT's capability to assist students in logically sequencing their ideas further reinforces the findings of Senciana.
In 2024, research indicated that AI tools enhance the clarity of thought processes in writing However, some students expressed neutrality or disagreement, highlighting that these tools may not cater to the varied needs of all learners, especially those with distinct writing styles or varying levels of expertise (Keser).
4.2.1.4 On Spelling Table 13 reports two items regarding students’ perceptions of ChatGPT’s impact on their spelling
Table 13 Students’ perceptions of ChatGPT’s impact on spelling
4.1 ChatGPT helped me reduce spelling mistakes in my paragraphs
4.2 I paid more attention to spelling after seeing ChatGPT’s corrections
A study on the effectiveness of ChatGPT in reducing spelling errors revealed that 62.2% of participants found the tool beneficial, with 40% completely agreeing and 22.2% somewhat agreeing Conversely, 20% of respondents remained neutral, indicating indifference or a lack of noticeable impact, while a small percentage, 11.1%, completely disagreed and 6.7% somewhat disagreed, reflecting limited dissatisfaction with the tool's performance.
A significant 64.5% of students reported increased awareness of their spelling errors after receiving corrections from ChatGPT, with 37.8% fully agreeing and 26.7% somewhat agreeing Meanwhile, 26.7% remained neutral, and a small portion, comprising 8.9%, expressed disagreement regarding the effectiveness of ChatGPT's corrections in enhancing their attention to spelling mistakes.