LLMs and Academic Writing in Practice: Exploring Participants’ Utilization of GPT during an AI-Assisted Course on Writing Research Papers

LLMs and Academic Writing in Practice: Exploring Participants’ Utilization of GPT during an AI-Assisted Course on Writing Research Papers | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article LLMs and Academic Writing in Practice: Exploring Participants’ Utilization of GPT during an AI-Assisted Course on Writing Research Papers Esmaeel Ali Salimi, Mahdi Hajinia This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5534554/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract This study investigates the integration of Generative Pre-trained Transformer (GPT) models in scholarly writing instruction, aiming to understand their perceived positive and negative impacts on academic writing. The research employs a phenomenographic design to study the practical implications of such models in real classrooms. The course design involves five sessions, each focusing on specific sections of a research paper, with activities using GPT integrated into classes and home projects. The methodology comprehensively investigates the impact of GPT-4 on scholarly writing, with a phenomenographic perspective towards participant selection, data collection, and research design. The study reveals the multifaceted ways in which GPT models contribute to the productivity and efficiency of writing research papers, highlighting both the advantages and challenges associated with their use. The findings underscore the transformative potential of AI in education, emphasizing the need for educators and researchers to understand and maximize the benefits of GPT models. AI LLMs GPT Academic Writing Teaching English Figures Figure 1 Figure 2 Figure 3 1. Introduction In the field of education, the ongoing quest for innovative teaching methodologies persists among educators, paralleled by students' perennial interest in adopting new pedagogical tools (Firmin, 2013 ). The emergence of Artificial Intelligence (AI), specifically within Artificial Intelligence in Education (AIEd), has transformed the educational landscape (Chen et al., 2020 ; Zhang & Aslan, 2021 ). In this regard, Large Language Models (LLMs), at the intersection of AI and linguistics, allow computers to process and generate human languages (Khurana, 2023), and provide language learners with opportunities towards betterment of language learning process. As a category of AIs stemmed from LLMs, Generative Pre-trained Processors (GPT), exemplified by GPT 4, demonstrate proficiency in generating human-like languages (Qin et al., 2023 ) and has recently come to the attention of many researchers across various disciplines. In line with the research available on application of LLMs in education (e.g., Derakhshan & Ghiasvand, 2024 ), this research focuses on their impacts, particularly GPT-4 for its ease of access and popularity, on scholarly writing in researchers’ L2 (i.e., English), specifically in research paper composition. While existing literature explores broader dimensions of GPT in education, this study addresses a gap by examining its integration into real classrooms, particularly in scholarly writing instruction. The study aims to elucidate the dynamics of GPT within education, considering its effects on educators' and students' perspectives. It emphasizes identifying challenges and how these models improve researchers’ productivity and efficiency, anticipating a potential paradigm specifically in the field of academic writing. This research highlights GPT’s role in education, influencing teaching, learning, and policy, with practical outcomes for all academic stakeholders. It aims to provide a nuanced understanding of GPT's implications in contemporary education, specifically in scholarly writing instruction. The research questions to be answered by this study are: RQ1. What are participants’ perceptions (positive, negative, and mixed) towards applying GPT models in classrooms of scholarly writing after utilizing such models during a course? RQ2. How introducing GPT to classrooms would increase productivity and efficiency of researchers in their L2 academic writing? RQ3. What challenges and limitations would participants deal with while using such models? 2. Literature Review Natural Language Processing (NLP) stands as a multifaceted domain within Artificial Intelligence (AI), converging elements from linguistics, computer science, and mathematics (Jiang & Lu, 2020 ). Originating in the 1950s, NLP employs computational methodologies to analyze and represent human language, with the overarching objective of achieving language processing akin to human capabilities (Jurafsky & Martin, 2019 ; Liddy, 2001 ). Manning and Schütze ( 1999 ) elaborate, defining NLP as an exploration of algorithms enabling computers to comprehend human language. Scholars broadly define NLP as an inquiry into computer handling of written or spoken language for practical purposes (Johri et al., 2021 ). The historical trajectory of NLP models (which directly influenced the emergence of LLMs) unfolds with the development of the first machine translation (MT) system in 1954, marking an early milestone (Hutchins, 1986 ). However, significant strides in NLP techniques gained momentum only in the 1980s and 1990s. Early MT systems primarily relied on dictionary lookups, emphasizing vocabulary and word order distinctions between languages (Liddy, 2001 ). It was Chomsky's introduction of generative grammar in "Syntactic Structures" (1957) which provided a deeper understanding of linguistics, fostering the emergence of additional NLP facets like speech recognition (Liddy, 2001 ). After Chomsky’s contributions, the 1960s witnessed substantial linguistic advancements, prompting investigations into the computational representation of meaning (Wilks, 1978 ). Prototype systems such as ELIZA, SHRDLU, PARRY, and LUNAR contributed to theoretical progress of this idea (Weizenbaum, 1966 ; Cerf, 1973 ; Woods, 1970 ). As a significant advancement in those years, the 1980s prompted a shift towards comprehensive NLP programs, addressing shortcomings of isolated solutions (Liddy, 2001 ). From the 1990s onwards, NLP entered a prosperous era, marked by significant growth fueled by technological advancements, including the internet, high-speed computers, and greater access to large volumes of electronic text (Jiang & Lu, 2020 ). The availability of vast web data has propelled deep learning models to unprecedented precision in NLP benchmarks (Brown et al., 2020 ). As models rooted in NLP, LLMs such as GPT emerge as prominent deep learning models, boasting 175 billion parameters in GPT-3 (Brown et al., 2020 ). Although GPT models benefit from various technologies such as deep learning, machine learning, and so on, in this study, due to its more relevance to the field of applied linguistics, NLP has been covered and reviewed as one of the significant contributors of GPT models. GPT's pre-training process encompasses learning grammar, syntactic knowledge, common sense, and world knowledge (Liu et al., 2021 ; Hewitt and Manning, 2019 ). Categorized as a generative model, GPT-3's capabilities transcend language generation, incorporating transfer learning principles (Pavlik, 2023 ). Pre-trained language models, including GPT-3, fall into three categories: unidirectional for natural language generation, bidirectional for understanding, and hybrid models combining both (Liu et al., 2021 ). GPT-3's impact on machine learning suggests unidirectional models can excel in both natural language understanding and generation (Brown et al., 2020 ). Its success has led to applications in diverse fields, facilitated by transfer learning principles (King, 2022 ). GPT's merits include its ability to generate high-quality language, adapt language level and tone, and facilitate specific writing tasks (Lund & Wang, 2023 ; Bishop, 2023 ). However, limitations such as potential biases, context comprehension issues, and high resource demands, and ethical issues have been identified (Zhou et al., 2021 ; Floridi & Chiriatti, 2020 ). GPT has attracted attention in various disciplines, offering potential benefits in writing research papers (Biswas, 2023 ; Lund and Wang, 2023 ). Researchers have explored its applications in enhancing productivity, imitating writing styles, and addressing specific parts of papers like literature reviews and abstracts (Chen, 2023 ; Gao et al., 2022; Wang et al., 2023 ). While acknowledging its utility, concerns about ethical and legal issues persist, urging responsible use and consideration of potential consequences (Bishop, 2023 ; Lund and Wang, 2023 ). The research landscape indicates a need for further exploration, particularly in understanding the nuanced impact of GPT on scholarly writing in diverse contexts. To borrow examples from fields other than applied linguistics, in the medical field, ChatGPT shows promise in assisting healthcare professionals by providing information, generating reports, and aiding in documentation (O’Connor & ChatGPT, 2023; D’amico et al., 2022 ). However, the story of GPT models is not unanimously believed to be optimistic, and there are downsides to it as well. The integration of GPT into journalism raises questions about the authenticity and objectivity of machine-generated content. While GPT can assist in drafting articles and summarizing information, the potential for biased reporting and the need for human oversight are crucial considerations (Pavlik, 2023 ). In the educational landscape, ChatGPT has demonstrated its value as a teaching assistant, aiding educators in various tasks (Baidoo-Anu & Owusu Ansah, 2023 ). Its ability to generate practice questions, provide explanations, and offer feedback on assignments has the potential to revolutionize the learning experience. However, concerns about the ethical implications of relying heavily on AI in education have been raised, questioning issues of accountability and the potential impact on critical thinking skills (Williams, 2023 ). ChatGPT’s potential in education and other fields is recognized, despite mixed feelings among educators (Baidoo-Anu & Owusu Ansah, 2023 ). Its utility has been explored in various domains including education (Williams, 2023 ; Tate, 2023), engineering (Qadir, 2022 ), journalism (Pavlik, 2023 ), and medicine (Nisar & Aslam, 2023 ; O’Connor & ChatGPT, 2023; D’amico et al., 2022 ; Ali et al., 2023 ). It aids researchers in writing papers (Alkaissi & McFarlane, 2023 ), with a study by Biswas ( 2023 ) focusing on its influence on medical writing. Lund and Wang ( 2023 ) found it beneficial for information management and content creation, enhancing search services, aiding reference services, facilitating cataloging, and contributing to content creation. Chen ( 2023 ) highlighted its role in collaborative writing, generating ideas, offering suggestions, and emulating writing styles. This fosters interdisciplinary collaboration and innovation. However, ethical use concerns persist, with Bishop ( 2023 ) emphasizing responsible AI use, especially where machine-generated content can sway public opinion. Biases, context comprehension issues, and resource demands call for careful deployment. Despite extensive research on GPT models’ uses, more qualitative studies are needed, particularly in different areas of English language teaching. Current research lacks conclusive findings on the benefits for teachers in academic writing instruction. This study aims to fill this gap, exploring changes in participants’ perceptions and identifying potential opportunities and challenges through rigorous qualitative analysis, and study various implications such models would have in specifically teaching academic writing and writing research papers. 3. Methodology 3.1 Research Design In this study, we investigated the application of AI, specifically LLMs, in academic writing classrooms. The research employed phenomenographic qualitative research design to systematically gather and analyze data obtained from participants in the study. This design has also been adopted by other researchers in the field in order to investigate participants’ perceptions regarding application of LLMs in teaching and learning English (Derakhshan & Ghiasvand, 2024). Given the lack of empirical work on the impacts of LLMs in scholarly writing, the study aims to study participants’ perceptions towards these models following a phenomenographic point of view due to its potential and use in investigating how people perceive real world phenomena following a naturalistic and explorative approach (Hajar, 2021; Sin, 2010). In addition to that, Creswell's recommendation (2013) supports the use of a qualitative lens for transformative outcomes in unexplored research areas. Given the diverse perspectives and understandings that researchers may hold regarding LLMs and AI in education, this study employed a qualitative research design to explore participants’ views on the potential benefits and limitations of such models. 3.2 Participants A total of 20 (11 male and 9 female) university students from 2 different universities with advanced English proficiency (IELTS score>7.5) were selected from 53 volunteers based on the inclusion criteria mentioned in Table 1 below following a criterin sampling approach. The process, detailed in Figure 1, involves screening volunteers to ensure language proficiency and academic writing criteria are met. Additionally, 5 experienced ELT instructors (3 male and 2 female) (Figure 2), were chosen through convenience sampling (Mackey & Gass, 2016) in order to teach scholarly writing classes. Table 1. Inclusion Criteria for Students From the initial 53 volunteers, 20 students meeting the inclusion criteria were selected for the study. The inclusion and exclusion criteria aim to ensure a homogeneous group with a consistent background for a focused investigation. To hold classes, 5 ELT instructors with experience in scholarly writing instruction at a university level were selected. The instructors, all possessing related university degrees, were chosen based on their qualifications and willingness to participate in the study as will be detailed in Figure 2 below. All ELT instructors have been teaching at higher education level for more than 10 years and have run various courses related to academic writing and writing research papers. The recruitment process is detailed in Figure 2. 3.3 Instruments and Materials 3.3.1 Classroom Observations To better understand how the process of the treatment goes on, and to guarantee that instructors adhered to the course design, all classes were observed by two observers. Moreover, to increase the quality of qualitative data obtained and to analyze the classes in the next phases of the study, after receiving consent from teachers, classes were audio recorded. These classroom observations provided researchers with the opportunities to gain genuine first-hand insights from classrooms (Cohen et al., 2013). Ethical considerations in this step, including obtaining informed consent and providing the right to withdraw from the study, were all considered and discussed with participants. To ensure reliability of classroom observations, inter-rater consistency was assessed through the re-analysis of randomly selected sessions and in all instances, two observers had consensus according to taking notes and completing the checklists. These classroom observations helped us to attain relative degrees of objectivity in data collection and analysis. Although the nature of this study posits a perception seeking inquiry, direct field observations helped us to add some objectivity to the results of the study and increased the validity of results of the study. 3.3.2 Semi-structured Interviews Semi-structured interviews (Mackey & Gass, 2016) are used in this study to investigate participants’ perceptions after using GPT during the course of study. A set of questions, refined based on expert feedback, was prepared by researchers for these interviews. They explored participants’ experiences with GPT, their perceptions of its use, and their views on AI-assisted teaching and learning. Following steps proposed by Creswell (2013) in conducting qualitative research, interview questions in compliance with research objectives were developed by researchers. These questions were delivered to three experts with experience in conducting qualitative research and their comments and feedbacks on questions’ content validity in terms of language suitability, pertinence, and clarity were considered and incorporated. Interviews with students were held online using Skype platform. Instructor interviews were face to face and interactive and were recorded on an iPhone. 3.3.3 Audio-journals Audio-journals were employed to capture participants' experiences during the treatment. After each session, all participants, teachers and students, were required to record an audio-journal and send it to the researcher to be transcribed and analyzed. These journals offered rich insights and aligned with Breakwell's recognition of their value in data collection. As Breakwell (2012) highlighted, audio journals are a great source of obtaining valuable information particularly when it comes to investigating the process of one treatment. They allow participants to freely express themselves and make comments more accurately. This instrument is invaluable in qualitative research since provides participants with opportunities to keep record of the process and gives them sufficient time to elaborate on their thoughts (Breakwell, 2012). Since the course was run for 5 sessions, each participant had to send 5 files as their audio journal, and after the course a number of 125 files (100 from students and 25 from instructors) were received. 3.3.4 Focus Group Discussions Post-research Skype focus group discussions (FGDs) involving teachers and students were conducted. In these FGDs, participants discussed their GPT models’ utilization experiences and challenges, and all discussions were recorded for data analysis. In qualitative studies, this type of discussions provides researchers with methodological triangulation, helping them to enhance their data reliability and add new insights to their research (Denzin, 1973; Creswell, 2013). As for the questions used in this stage, we focused on the semi-structured interview questions and tried to elaborate on them by discussing these questions in a less structured format with participants. 3.4 Course Design The course comprised five sessions, each focusing on a research paper section. While running classes, instructors followed a standard plan, integrating GPT based activities into classes and homework, and encouraging students to use GPT in order to complete their tasks. The first session covered introductions and literature reviews, the second focused on methodology, the third on data analysis and collection, the fourth on discussion and conclusion, and the final session reviewed course content and focused on abstract writing. Each session included in-class GPT based activities and homework to reinforce students’ understanding and encourage them to use GPT in order to write a section of a research paper. In each of the five sessions of the course, the instructor gave a lecture on the content (e.g., how to write a literature review) and asked students to use GPT in the classroom to practice writing that section with the aid of GPT. It was students’ decision how to adopt GPT and how to use it in the classroom. Then, as data for this study, students kept an audio journal of how they used GPT to write that specific section of an article in their classes. Homework was also given to students and they had to use GPT at home in order to complete required tasks (which were generally product-oriented). At the same time, teachers monitored students’ activities in class and outside the class and they kept their own audio journals as well. Besides, similarly to students, teachers were asked to use GPT while preparing for their lectures, designing tasks, and giving homework to students. All classes were observed by researchers, so that we made sure instructors adhered to the plan, and participants’ perceptions were generally affected by the course content and design. 3.5 Data Collection Procedure This study used semi-structured interviews, observations, audio journals, and FGDs to collect data. Before the study, interviews with teachers and students (participants were interviewed individually) ensured equal background knowledge and if participants are fully aware of course design and the treatment process. Interviews were piloted and improved based on feedback from 2 interviewees and 3 applied linguistics experts. After conducting these initial interviews, 5 classes of 4 students were formed. Instructors taught 5 sessions of scholarly writing using GPT-4. To ensure reliability, a 90-minute workshop was held before the courses began to align instructors with the research objectives. Both teachers and students were trained to keep audio journals during the treatment, and classes were observed and recorded for reliability of field notes obtained from these observations. Post-treatment, participants attended the main interviews and FGDs, so that researchers studied their perceptions and experiences after using GPT. As for FGDs, two sessions each for teachers and students were held on Skype to discuss their opinions on the course and GPT’s contribution to improving research paper writing. All interviews and discussions were recorded and transcribed by both researchers, and coded for analysis. 4. Data Analysis This study uses a phenomenographic approach towards qualitative data analysis as it complies with the objectives of this study. In this study, we analyzed obtained data through content and thematic analysis to attain an understanding of participants’ perceptions. As has been emphasized by Braun and Clarke (2006), content analysis deals with the presence and frequency of certain words, themes, or concepts within qualitative data. Thematic analysis then helps researchers to find themes and patterns across data and investigate relationships between these themes. Moreover, to analyze data systematically, a framework for phenomenographic data analysis (PDA) proposed by Stenfors-Hayes et al. (2013) adopted. This model includes seven steps of data familiarization, condensation, comparing, contrasting, grouping, articulating, and final labeling. Data analysis was conducted using MAXQDA software for systematic coding and theme organization in this study. Four data sets: semi-structured interviews, observations, audio-journals, and focus-group discussions were transcribed, coded, and inputted into MAXQDA for pattern identification and thematic analysis. Due to the qualitative nature of this study, and as has been emphasized by Lincoln and Guba (1985) in their seminal book, principles of trustworthiness and naturalistic investigation were considered. The first principle employed was member checking, wherein participants were requested to review their transcripts and our interpretations of their responses. Secondly, inter-coder agreement was calculated based on the independent coding of a randomly selected twenty percent of the data by the second coder. Results showed that there was a 97% agreement between coders. The third principle, confirmability, was maintained by asking an experienced L2 researcher meticulously review the data analysis phase. As another important principle, we tried to establish credibility through interactive interviews, peer debriefing, orientation sessions before treatment, and the bracketing of personal experiences and perceptions. To enhance the dependability of the findings as another principle, the researchers conducted content and thematic analyses, iterative data transcription and coding, and provided a coherent description of the methodological steps undertaken. Finally, the transferability of the findings was bolstered by providing comprehensive descriptive data, enabling future researchers to replicate and recontextualize the study. Regarding researcher positionality, it is important to acknowledge that while the researchers strived to maintain neutrality in this qualitative study, complete objectivity is inherently challenging in such designs. 5. Findings In the exploration of teachers and students’ perceptions towards the integration of LLMs in academic writing classrooms, a systematic approach categorized responses of all participants into positive, negative, and mixed viewpoints. Quantitative analysis, as illustrated in Table 2 below, provided a comprehensive overview of perceptions, with 51.43% positive, 28.57% negative, and 20.00% mixed responses throughout the 70 relevant codes. Table 2. Post-treatment Positive, Negative, and Mixed Perceptions Themes Frequency Percentage Positive Perceptions ( 51.43%) Ad-Free Environment 7 19.45 Using Technology Is Good 6 16.67 Engaging Interface and Unique Interaction Pattern 6 16.67 Interactive Nature Of GPT 5 13.89 Reducing Workload 4 11.11 Welcoming It as An Advancement 3 8.34 Helps In Many Different Ways 2 5.56 Always Available 2 5.56 Improves Communication Skills 1 2.78 TOTAL 36 100.00 Negative Perceptions ( 28.57%) More Research Needed on Its Effectiveness and Impacts 10 50.00 Changes the Whole Educational System and Undermines Teachers and Educators 2 10.00 Not Feeling Comfortable with Technology 2 10.00 No Training Accessible for Majority of People 3 15.00 Makes People Lazy 2 10.00 Just Do Not Like It 1 5.00 TOTAL 20 100.00 Mixed Perceptions ( 20.00%) It Can Be Both Good and Evil 6 42.86 Too Complicated 4 28.57 Technology Use in Schools 2 14.29 More Time Needed to Decide 2 14.29 TOTAL 14 100.00 The breakdown revealed that 51.43% of participants held positive perceptions, much appreciating GPT for its ad-free nature (e.g., Student 3 mentioned “… I also hate ads which come with the websites that google introduces ”) and engaging qualities (e.g., Teacher 2 mentioned “if students see that it really helps them, they would become more engaged in the process of learning” ). Noteworthy sentiments included participants welcoming GPT with open arms (e.g., Teacher 1 highlighted “… I welcome it with open arms” ) and recognizing its benefits in both teaching and learning. Conversely, 28.57% expressed negative perceptions, citing concerns about changes in educational system and the potential erosion of teachers' authority (e.g., Teacher 3 highlighted, “there might not be a need for teachers anymore if the uses of AI are more prevalent in future” ), discomfort with technology (e.g., Teacher 1 mentioned “it is not still prevalent and not many teachers use it (they might be afraid of it)” ), and reservations about GPT's impact on students' skill development (e.g., Teacher 4 with reference to limitations and challenges mentioned: “it makes students at the end of the term lack necessary skills for writing due to the overreliance on GPT” ). However, half of the negative perceptions were coming from the perception that it is a relatively new field and must be researched in more depth (e.g., Student 6 mentioned “we do not know if it really works” ). Mixed perceptions, accounting for 20.00% of responses, reflected indecisiveness among participants. Ambiguities centered on the multifaceted nature of GPT's impact (e.g., student 8 mentioned “GPT is everywhere and can do anything for us. I don’t know if it’s good or bad”), uncertainty about its role in society since it is a mixed blessings for them, and considerations related to national rules on technology use in schools (as student 14 mentioned “I can’t take my laptop or mobile phone to school with myself, I like it but I have problems with it too” ). In the investigation of LLMs and their impact on researchers' productivity and efficiency in writing papers, a meticulous open coding process identified 10 main categories. Considering the common and non-technical meaning of efficiency and productivity, in this study, we consider them mainly as advantages GPT brings about into classrooms that would help researchers to write academic papers with more ease, more effectively, and improve their papers’ quality. Table 3 provides a comprehensive overview of these categories, highlighting the frequency distribution of codes within each distinct aspect. The high frequency of codes in this category (180) showed that participants deem GPT to significantly boost productivity and efficiency of researchers while writing academic papers. Table 3 . Efficiency and Productivity Themes Frequency Percentage Boosts the Speed of Writing 54 30.00 Immediate Feedback on Writing 36 20.00 Improves Moves and Steps in a Section 24 13.33 Suggestions on Academic Language 23 12.77 Categorization of Writing 12 6.67 Organizing, Planning, and Outlining 9 5.00 Ideas for Getting Started and How to Plan 9 5.00 Extracts Keywords 4 2.22 An Asset in Systematic Work 3 1.67 Helps to Better Find Gaps in Research 3 1.67 Meeting the Guidelines of a Journal 2 1.11 Putting Ideas in Sentences 1 0.55 TOTAL 180 100.00 The study revealed that GPT has the potential to significantly enhance researchers' productivity and efficiency across various dimensions. The most prominent category, with a high frequency of 30.00%, centered on how GPT increases the speed of tasks (e.g., when Teacher 2 was asked what was significant about GPT and why you liked it, she mentioned in a decisive tone, “it saves your time” ). This table further delineates the breakdown of categories, emphasizing the multifaceted contributions of GPT. Providing researchers with immediate feedback on their writing (20.00%) (e.g., “you can receive feedback from GPT on your wiring, and it helps you to improve your writing” (Participant 11)), improving moves and steps in a writing (13.33%) (e.g., “it helped me to better tailor my text for the audience and know how to write different sections after each other” (Participant 5)), and enhancing writing style in terms of being more academic (12.77%) (e.g., “… or consulting with it while I want to improve my style” (Participant 9)) emerged as significant areas where GPT showcased its efficacy. In the next phase of data analysis, the examination shifted towards the challenges and limitations inherent in leveraging GPT models as the primary instructional aid for teaching academic writing. The qualitative analysis brought forth 11 distinct categories of challenges, as depicted in Table 4, illuminating the potential hurdles faced by both teachers and students. Table 4. Challenges and Limitations Themes Frequency Percentage Overreliance on AI 19 20.21 Accurate Prompting Must be Learnt 18 19.15 Not everyone Can Use GPT Effectively 13 13.82 Must Always be Doublechecked in Terms of Scientific Accuracy 12 12.77 Fabricating Non-Existing Information 10 10.64 Still Not Complete and Accurate 6 6.38 High-Tech Plagiarism 6 6.38 Patchwriting 4 4.26 Not Sensitive Towards Different cultures 3 3.19 Data from Non-Existing Sources 2 2.13 Generates Accurate Data Only in English 1 1.06 TOTAL 94 100.00 Among the identified challenges, the most significant was the issue of overreliance on AI (20.21%). Participants encountered difficulties due to an excessive dependence on GPT models, emphasizing the necessity for a balanced approach in utilizing these tools effectively (e.g., “I believe it to be so addictive which makes you lazy” (Teacher 4)). Prompting emerged as another crucial factor, with 19.15% of segments underscoring its pivotal role and difficulties related to it as limiting (e.g., “we should be hyper precise while talking to GPT” (Participant 14)). Concerns about repeated answers underscored the challenges faced by teachers in generating precise prompts. Education on working with AI (13.82%) and the requirement for always being double-checked (12.77%) emerged as significant challenges, indicating a perceived need for more comprehensive training to effectively utilize GPT models. Concerns related to the fabrication of information (10.64%) and the risk of high-tech plagiarism (term borrowed from Chomsky (2023)) (6.38%) highlighted apprehensions about the accuracy and reliability of information generated by GPT models. Issues related to the lack of focus on cultural-linguistic points of the language (3.19%) indicated a desire for a more nuanced approach to language instruction (e.g., “I suppose GPT does not care about me in a way that it does not take cultural differences into account (Participant 5)). Additionally, the limitation of generating accurate data only in English (1.06%) (e.g., “I tried to get help from GPT to write Persian texts but my problem is that GPT can only generate accurate data in English, not in any other languages” (Participant 16)) and the existence of non-existing sources (2.13%) (e.g., “… it referred to sth which did not exist at all” (Teacher 3)) underscored the current limitations of GPT models, suggesting areas for improvement. 5. Discussion AI models such as GPT have recently come to the attention of many researchers and their impacts on various educational contexts have been investigated. However, a gap remained regarding how such models impact writing research papers after completing a course in which GPT was the main focus of the instruction. As could be expected, after completion of the course, the majority of participants held positive perceptions towards such models and this could explain why FGDs in this study were mainly revolving around the positive aspects of such models. Since positive perceptions predominated, in this study, we investigated in which areas participants (instructors and students) found GPT models to be effective and increase their productivity and proficiency. 10 major themes were extracted with deep support in available literature as will be mentioned below. However, it is nearly impossible to find one teaching method all meritorious without challenges and limitations. 11 distinct categories of challenges and limitations were extracted in this study, and we found that in other areas of language teaching such themes have been studied and supported. To answer our first research question regarding participants’ attitudes, this study revealed participants’ predominantly positive perceptions of GPT models, influenced by factors such as ad-free usage and extended engagement which are supported by existing literature as well (Munoz et al., 2023; Diwan et al., 2023 ). Material development and interaction patterns with models contributed to this positive view. Provided participants held predominantly positive perceptions towards applying such models, a close relationship can be found between their positive perceptions and our second research question regarding impacts of such models on participants productivity and efficiency in writing research papers. This relationship can be justified by the Technology Acceptance Model (TAM), first introduced by Davis et al. in 1989, which examines individual behavior towards technology adoption within various information system frameworks. Rooted in social psychology, TAM explores the interplay between cognitive and affective factors and how they influence users' technology usage. Technology acceptance unfolds in three stages: initial external factors lead to cognitive responses (like perceived usefulness and ease of use); these then shape an affective response (user's intention or attitude towards using the technology), which ultimately influences the user's behavior. The model and relationships are illustrated in Fig. 3 below. However, as could be expected, negative perceptions exist too, and many of them have been mentioned by other researchers. For example, Wang et al.'s ( 2023 ) research on GPT’s trustworthiness and concerns about data leakage supported participants’ negative perceptions in this regard. Besides, some participants were concerned whether such models inhibit students’ skill development. Concerns about these models replacing teachers and disrupting traditional teaching methods were prevalent in results, aligning with previous studies in the field (Fuchs, ccd2023; Farazouli et al., 2023 ). All in all, we sought participants’ perceptions to make sure how utilizing LLMs in classrooms would impact their perceptions, and as results confirmed, in line with previous research available, their perceptions were of a more positive nature and this is promising for the future of AI-Assisted instruction. Possible explanations have been proposed by researchers such as Derakhshan and Ghiasvand ( 2024 ) as they emphasized the interactive and engaging nature of these models. However, as expected, addressing these concerns requires proper training and support (Unser, 2017 ), and requires more attention to application of such models from scholars’ side. Although previous research emphasizing participants’ positive perceptions exist, this current study differs in terms of offering an experimental course utilizing GPT as the main aid for delivering and learning content. Results of this study answering to the first research question, in addition to previous research, support their findings particularly in the case of scholarly writing through a real classroom experimental design. In line with our second research question concerning how these models help improve researchers’ productivity and efficiency, this study highlights GPT models’ multifaceted contributions to writing research papers marking a paradigm shift in academic research. To further classify themes extracted, we identified two areas in which GPT helped researchers increase productivity and efficiency: (1) in terms of language and linguistic accuracy and (2) in terms of adhering to academic conventions and helping with completing a research paper. Many of the themes extracted have supports in available literature. For example, in terms of linguistic aids, GPT accelerates writing by increasing the speed of writing, allowing researchers to focus on conceptualization and analysis (Buruk, 2023 ) rather than focusing on linguistic aspects of writing. It also aids researchers in text editing and feedback provision, ensuring high-quality academic writing (Carlson, 2023; Shidiq, 2023 ; Straume & Anson, 2022 ). In terms of completing research papers, GPT improves research paper discourse and style which helps researchers to adhere to academic conventions (de Rivero et al., 2023 ) more appropriately. Moreover, it supports researchers in planning and structuring papers helping them with moves and steps in writing different sections of a paper. Besides, it contributes to the quality of thematic analysis and categorization in qualitative data analysis (Zhang et al., 2023 ; Gamieldien et al., 2023 ; Koch, 2023 ) and as was mentioned by one participant of the study, “ … it can be a good source for asking my research questions ”. GPT’s keyword generation feature optimizes paper visibility and impact, and it also helps researchers in literature gap identification (Firoozeh et al., 2020 ; Ezzelding & El-Dakhakhni, 2020). And as a very important area in which GPT helps significantly, it empowers researchers to better adhere to academic writing guidelines, ensuring proper formatting and citation (Zhao, 2022 ; Salvagno et al., 2022) and therefore, increasing researchers’ chance of getting published in academic journals. This could be interpreted in line with the dominance of positive perceptions in answer to the first research question. Participants of the study believe GPT to revolutionize the field of scholarly writing since it brings about unprecedented tools to increase researchers’ productivity and efficiency. GPT models have revolutionized academic writing, enhancing efficiency and productivity. However, as per our third research question, challenges exist. Overreliance on AI could undermine research exploration (Abd-Alrazaq et al., 2023 ). The effectiveness of GPT depends on well-formulated prompts (Liu et al., 2023 ; Alivanistos et al., 2023) that requires proper training and effort. So, education and prior knowledge are essential for effective GPT use (Min et al., 2023 ). GPT’s information fabrication is a limitation, raising ethical concerns (Walters & Wilder, 2023 ; Mosca, 2023) and challenging research papers validity. Plagiarism risk is another challenge (Dehouche, 2021 ) for researchers and publishers which must be investigated more in-depth. GPT’s lack of focus on cultural aspects of languages poses challenges and must be investigated through a sociolinguistic point of view (Quian et al., 2021; Johnson et al., 2022 ). The generation of non-existing sources and limitations in generating accurate data in languages other than English were additional challenges participants faced during the course of study. These limitations and challenges are still to be researched by scholars in the field and their multifaceted impacts should be brought to attention of academics. However, with all its limitations and challenges, participants of this current study held a more positive perception towards the application of LLMs in classrooms. 6. Conclusion This study has provided a comprehensive exploration of the impacts of LLMs on scholarly writing. It has shed light on the transformative potential of these models in enhancing researchers’ productivity and efficiency, and has offered a nuanced understanding of their implications in contemporary education. The study indicates that GPT models can enhance research paper writing, potentially causing a shift in academic writing. However, it also highlights the challenges of integrating these models into classrooms. The study’s implications are significant for educators, syllabus designers, students, and researchers, suggesting GPT models can improve teaching methods, anticipate pedagogical changes, impact language skills, and provide a basis for interventions. More importantly, this research would be influential for EAP and ESP instructors and course designers, and help them to better understand the nature of integrating AI in courses tailored for the needs of a specific group of students. Despite progress in understanding GPT’s impact on academic writing, more research is needed to understand its full effects. Future research could include longitudinal studies on GPT use in different educational settings, exploring the impacts of various GPT models, and studying GPT’s long-term effects on language skills. As AI evolves, continuous research is essential to keep up with developments and their educational implications. Declarations Ethics Approval and Consent to Participate The need for ethical approval was waived off by the nature of the study and the ethical board at Allameh Tabatabei University, Department of Literature and Languages. However, considering the nature of the study, informed consent was received from all participants with right to withdraw from the study at any moment during the process of the research. Availability of Data and Materials The datasets used and analyzed during the current study are available from the corresponding author on reasonable request. Competing interests The authors declare that they have no competing interests. Authors' contributions All authors contributed to the process of researching, analyzing, and reporting this paper. However, MH was mainly in charge of code analysis and working with MAXQDA due to his expertise in the field, and EAS designed the method and approach of the study. Authors' Biography Esmaeel Ali Salimi is a faculty member of Allameh Tabataei university who has published many artcicle in the fields of Applied Linguistics, Learner Factors, Assessment, Innovation in Teaching, and Discourse Analysis in esteemed journals such as Learning and Motivation, Teaching English Language, Journal of Language and Education, and Teflin journal. He is currently working as an associate professor supervising PhD and MA students’ research, and instructs courses related to Research, Innovations in TEFL, and Discourse Studies. References Abd-Alrazaq, A., AlSaad, R., Alhuwail, D., Ahmed, A., Healy, P. M., Latifi, S., … & Sheikh, J. (2023). Large Language Models in Medical Education: Opportunities, Challenges, and Future Directions. JMIR Medical Education , 9(1), e48291. https://doi.org/10.2196/48291 Ali, J. K. M., Shamsan, M. A. A., Hezam, T. A., & Mohammed, A. A. (2023). Impact of ChatGPT on Learning Motivation: Teachers and Students’ Voices. Journal of English Studies in Arabia Felix , 2(1), 41-49. https://doi.org/10.56540/jesaf.v2i1.51 Alivanistos, D., Santamaría, S. B., Cochez, M., Kalo, J. C., van Krieken, E., & Thanapalasingam, T. (2022). Prompting as probing: Using language models for knowledge base construction. arXiv preprint arXiv:2208.11057. Alkaissi, H., & McFarlane, S. I. (2023). Artificial hallucinations in ChatGPT: implications in scientific writing. Cureus , 15(2). https://doi.org/10.7759/cureus.35179 Baidoo-Anu, D., & Owusu Ansah, L. (2023). Education in the Era of Generative Artificial Intelligence (AI): Understanding the Potential Benefits of ChatGPT in Promoting Teaching and Learning. Available at SSRN 4337484. https://doi.org/10.2139/ssrn.4337484 Bishop, L. (2023). A Computer Wrote This Paper: What ChatGPT Means for Education, Research, and Writing. Research, and Writing (January 26, 2023). https://doi.org/10.2139/ssrn.4338981 Biswas, S. (2023). ChatGPT and the Future of Medical Writing. Radiology , 223312. https://doi.org/10.1148/radiol.223312 Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology , 3(2), 77–101. https://doi.org/10.1191/1478088706qp063oa Breakwell, G. M. (2012). Diary and Narrative Methods. In G. Breakwell, J. Smith, & D. Wright (Eds.), Research Methods in Psychology (4th ed., pp. 111392-410). Sage. https://doi.org/10.1002/9780470776278 Brown, T., Mann, B., Ryder, N., Subbiah, M., Kaplan, J. D., Dhariwal, P., … & Amodei, D. (2020). Language Models Are Few-Shot Learners. Advances in Neural Information Processing Systems , 33, 1877-1901. https://doi.org/10.18653/v1/2020.emnlp-main.375 Buruk, O. O. (2023, October). Academic Writing with GPT-3.5 (ChatGPT): Reflections on Practices, Efficacy and Transparency. In Proceedings of the 26th International Academic Mindtrek Conference (pp. 144-153). https://doi.org/10.1145/3616961.3616992 Carlson, M., Pack, A., & Escalante, J. (2023). Utilizing OpenAI’s GPT‐4 for Written Feedback. TESOL Journal , e759. Cerf, V. (1973). Parry Encounters the Doctor (No. rfc439). https://doi.org/10.17487/rfc0439 Chen, L., Chen, P., & Lin, Z. (2020). Artificial Intelligence in Education: A Review. IEEE Access , 8, 75264-75278. https://doi.org/10.1109 Chen, T. J. (2023). ChatGPT and Other Artificial Intelligence Applications Speed Up Scientific Writing. Journal of the Chinese Medical Association . https://doi.org/10.1097/jcma.0000000000000900 Chomsky, N. (1957). Syntactic Structures . The Hague: Mouton. https://doi.org/10.1515/9783112316009 Chomsky N. Noam Chomsky on ChatGPT: It’s “Basically High-Tech Plagiarism” and “a Way of Avoiding Learning”. Open Culture. 2023 February 10th, 2023. [Available from: https://www.openculture.com/2023/02/noam-chomsky-on-chatgpt.html. Cohen, L., Manion, L., & Morrison, K. (2013). Research methods in education . Routledge. https://doi.org/10.4324/9780203720967 Creswell, J. W. (2013). Research design. Qualitative, quantitative and mixed method approaches (3rd ed.). Sage. https://doi.org/10.1086/602876 D’Amico, R. S., White, T. G., Shah, H. A., & Langer, D. J. (2022). I asked a ChatGPT to write an editorial about how we can incorporate chatbots into neurosurgical research and patient care…. Neurosurgery , 10-1227. https://doi.org/10.1227/neu.0000000000002414 Davis, F.D., R.P. Bagozzi, and P.R. Warshaw. 1989. User acceptance of computer technology: A comparison of twotheoretical models. Management Science 35: 982 1003. (PDF) The impact of Perceived Usefulness and Perceived Ease of Use on Online Purchases: A comparison of Buyers and Non-Buyers Perceptions . Available from: https://www.researchgate.net/publication/278676371_The_impact_of_Perceived_Usefulness_and_Perceived_Ease_of_Use_on_Online_Purchases_A_comparison_of_Buyers_and_Non-Buyers_Perceptions [accessed Dec 06 2024]. de Rivero, M., Tirado, C., & Ugarte, W. (2023). Formalstyler: GPT-Based Model for Formal Style Transfer with Meaning Preservation. SN Computer Science , 4(6), 739. https://doi.org/10.1007/s42979-023-02110-7 Dehouche, N. (2021). Plagiarism in the age of massive Generative Pre-trained Transformers (GPT-3). Ethics in Science and Environmental Politics , 21, 17-23. https://doi.org/10.3354/esep00195 Denzin, N. K. (1973). The research act: A theoretical introduction to sociological methods . Transaction Publishers. https://doi.org/10.4324/9781315134543 Derakhshan, A., & Ghiasvand, F. (2024). Is ChatGPT an evil or an angel for second language education and research? A phenomenographic study of research‐active EFL teachers’ perceptions. International Journal of Applied Linguistics . https://doi.org/10.1111/ijal.12561 Diwan, C., Srinivasa, S., Suri, G., Agarwal, S., & Ram, P. (2023). AI-based learning content generation and learning pathway augmentation to increase learner engagement. Computers and Education: Artificial Intelligence , 4, 100110. https://doi.org/10.1016/j.caeai.2022.100110 Ezzeldin, M., & El-Dakhakhni, W. (2020). Metaresearching structural engineering using text mining: Trend identifications and knowledge gap discoveries. Journal of Structural Engineering , 146(5), 04020061. https://doi.org/10.1061/(asce)st.1943-541x.0002523 Farazouli, A., Cerratto-Pargman, T., Bolander-Laksov, K., & McGrath, C. (2023). Hello GPT! Goodbye home examination? An exploratory study of AI chatbots’ impact on university teachers’ assessment practices. Assessment & Evaluation in Higher Education , 1-13. https://doi.org/10.1080/02602938.2023.2241676 Firmin, M. W., & Genesi, D. J. (2013). History and implementation of classroom technology. Procedia-Social and Behavioral Sciences , 93, 1603-1617. https://doi.org/10.1016/j.sbspro.2013.10.089 Firoozeh, N., Nazarenko, A., Alizon, F., & Daille, B. (2020). Keyword extraction: Issues and methods. Natural Language Engineering , 26(3), 259-291. https://doi.org/10.1017/s1351324919000457 Floridi, L., & Chiriatti, M. (2020). GPT-3: Its nature, scope, limits, and consequences. Minds and Machines , 30, 681-694. https://doi.org/10.1007/s11023-020-09548-1 Fuchs, K. (2023, May). Exploring the opportunities and challenges of NLP models in higher education: is Chat GPT a blessing or a curse?. In Frontiers in Education (Vol. 8, p. 1166682). Frontiers. https://doi.org/10.3389/feduc.2023.1166682 Gamieldien, Y., Case, J. M., & Katz, A. (2023). Advancing Qualitative Analysis: An Exploration of the Potential of Generative AI and NLP in Thematic Coding. Available at SSRN 4487768. https://doi.org/10.2139/ssrn.4487768 Hajar, A. (2021). Theoretical foundations of phenomenography: A critical review. Higher Education Research & Development, 40(7), 1421–1436. https://doi.org/10.1080/07294360.2020.1833844 Hewitt, J., & Manning, C. D. (2019). A structural probe for finding syntax in word representations. In North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL) . Association for Computational Linguistics. Hutchins W. J. (1986). Machine translation: Past, present, future. In Machine Translation (pp. 3-13). Springer. https://doi.org/10.2307/328480 Jiang, K., & Lu, X. (2020, November). Natural language processing and its applications in machine translation: A diachronic review. In 2020 IEEE 3rd International Conference of Safe Production and Informatization (IICSPI) (pp. 210-214). IEEE. https://doi.org/10.1109/iicspi51290.2020.9332458 Johnson, R. L., Pistilli, G., Menédez-González, N., Duran, L. D. D., Panai, E., Kalpokiene, J., & Bertulfo, D. J. (2022). The Ghost in the Machine has an American accent: value conflict in GPT-3. arXiv preprint arXiv:2203.07785. Johri, P., Khatri, S. K., Al-Taani, A. T., Sabharwal, M., Suvanov, S., & Kumar, A. (2021). Natural language processing: History, evolution, application, and future work. In Proceedings of 3rd International Conference on Computing Informatics and Networks: ICCIN 2020 (pp. 365-375). Springer Singapore. https://doi.org/10.1007/978-981-15-9712-1_31 Jurafsky, D., & Martin, J. H. (2019). Speech and Language Processing (3rd ed.). Pearson. Khurana, D., Koli, A., Khatter, K., & Singh, S. (2023). Natural language processing: State of the art, current trends and challenges. Multimedia tools and applications , 82(3), 3713-3744. https://doi.org/10.1007/s11042-022-13428-4 King, M.R. (2022), “The future of AI in medicine: a perspective from a Chatbot”, Annals of Biomedical Engineering , https://doi.org/10.1007/s10439-022-03121-w Koch, M. A. (2023). Turning Chaos into Meaning: A Chat GPT-Assisted Exploration of COVID-19 Narratives (Master’s thesis, University of Twente). Liddy, E. D. (2001). Natural language processing. In Encyclopedia of Library and Information Science, 2nd Ed . NY. Marcel Decker, Inc. Lincoln, Y. S., Guba, E. G., & Pilotta, J. J. (1985). Naturalistic inquiry. International Journal of Intercultural Relations, 9(4), 438–439. https://doi.org/10.1016/0147-1767(85)90062-8 Liu, P., Yuan, W., Fu, J., Jiang, Z., Hayashi, H., & Neubig, G. (2023). Pre-train, prompt, and predict: A systematic survey of prompting methods in natural language processing. ACM Computing Surveys , 55(9), 1-35. https://doi.org/10.1145/3560815 Liu, X., Zheng, Y., Du, Z., Ding, M., Qian, Y., Yang, Z., & Tang, J. (2021). GPT understands, too. arXiv preprint arXiv:2103.10385. https://doi.org/10.1016/j.aiopen.2023.08.012 Lund, B. D., Wang, T. (2023). Chatting about ChatGPT: how may AI and GPT impact academia and libraries? Library Hi Tech News . https://doi.org/10.1108/lhtn-01-2023-0009 Lund, B. D., Wang, T., Mannuru, N. R., Nie, B., Shimray, S., & Wang, Z. (2023). ChatGPT and a new academic reality: Artificial Intelligence‐written research papers and the ethics of the large language models in scholarly publishing. Journal of the Association for Information Science and Technology , 74(5), 570-581. https://doi.org/10.1002/asi.24750 Mackey, A., & Gass, S.M. (2016). Second language research methodology and design (2nd ed.). Rutledge. https://doi.org/10.4324/9781003188414 Manning, C. D., & Schütze, H. (1999). Foundations of Statistical Natural Language Processing . ISBN-13 978-0-521-86571-5. MIT Press. Min, B., Ross, H., Sulem, E., Veyseh, A. P. B., Nguyen, T. H., Sainz, O., … & Roth, D. (2023). Recent advances in natural language processing via large pre-trained language models: A survey. ACM Computing Surveys , 56(2), 1-40. https://doi.org/10.1145/3605943 Mosca, E., Abdalla, M. H. I., Basso, P., Musumeci, M., & Groh, G. (2023, July). Distinguishing Fact from Fiction: A Benchmark Dataset for Identifying Machine-Generated Scientific Papers in the LLM Era. In Proceedings of the 3rd Workshop on Trustworthy Natural Language Processing (TrustNLP 2023) (pp. 190-207). https://doi.org/10.18653/v1/2023.trustnlp-1.17 Muñoz, S. A. S., Gayoso, G. G., Huambo, A. C., Tapia, R. D. C., Incaluque, J. L., Aguila, O. E. P., … & Arias-Gonzáles, J. L. (2023). Examining the Impacts of ChatGPT on Student Motivation and Engagement. Social Space , 23(1), 1-27. Nisar, S., & Aslam, M. S. (2023). Is ChatGPT a Good Tool for T&CM Students in Studying Pharmacology? Available at SSRN 4324310. https://doi.org/10.2139/ssrn.4324310 O’Connor, S., & ChatGPT. (2023). Open artificial intelligence platforms in nursing education: Tools for academic progress or abuse? Nurse Education in Practice , 66, 103-537. https://doi.org/10.1016/j.nepr.2022.103537 Pavlik, J. V. (2023). Collaborating With ChatGPT: Considering the Implications of Generative Artificial Intelligence for Journalism and Media Education. Journalism & Mass Communication Educator , 0(0)*. https://doi.org/10.1177/10776958221149577 Qadir, Junaid (2022): Engineering Education in the Era of ChatGPT: Promise and Pitfalls of Generative AI for Education. TechRxiv. Preprint. https://doi.org/10.36227/techrxiv.21789434.v1. Qian, M., Newton, C., & Qian, D. (2021). Cultural Understanding Using In-context Learning and Masked Language Modeling. In HCI International 2021-Late Breaking Papers: Multimodality, eXtended Reality, and Artificial Intelligence: 23rd HCI International Conference, HCII 2021, Virtual Event, July 24–29, 2021, Proceedings 23 (pp. 500-508). Springer International Publishing. https://doi.org/10.1007/978-3-030-90963-5_38 Qin, C., Zhang, A., Zhang, Z., Chen, J., Yasunaga, M., & Yang, D. (2023). Is chatgpt a general-purpose natural language processing task solver?. arXiv preprint arXiv:2302.06476. https://doi.org/10.18653/v1/2023.emnlp-main.85 Salvagno, M., Taccone, F. S., & Gerli, A. G. (2023). Can artificial intelligence help for scientific writing? Critical care , 27(1), 1-5. https://doi.org/10.1186/s13054-023-04380-2 Sin, S. (2010). Considerations of quality in phenomenographic research. International Journal of Qualitative Methods, 9(4), 305– 319. https://doi.org/10.1177/160940691000900401 Shidiq, M. (2023, May). The use of artificial intelligence-based chat-gpt and its challenges for the world of education; from the viewpoint of the development of creative writing skills. In Proceeding of International Conference on Education, Society and Humanity (Vol. 1, No. 1), 353-357. Stenfors-Hayes, T., Hult, H., & Dahlgren, M. A. (2013). A phenomenographic approach to research in medical education. Medical Education, 47(3), 261–270. https://doi.org/10.1111/medu.12101 Straume, I., & Anson, C. (2022). Amazement and trepidation: Implications of AI-based natural language production for the teaching of writing. Journal of Academic Writing , 12(1), 1-9. https://doi.org/10.18552/joaw.v12i1.820 Tate, T. P., Doroudi, S., Ritchie, D., Xu, Y., & Uci, m. w. (2023, January 10). Educational Research and AI-Generated Writing: Confronting the Coming Tsunami. https://doi.org/10.35542/osf.io/4mec3 Unser, C. E. (2017). A Study on the positives and negatives of using technology in the classroom. Undergraduate Honors College Theses 2016-. 18. https://digitalcommons.liu.edu/post_honors_theses/18 Walters, W. H., & Wilder, E. I. (2023). Fabrication and errors in the bibliographic citations generated by ChatGPT. Scientific Reports , 13(1), 14045. https://doi.org/10.1038/s41598-023-41032-5 Wang, B., Chen, W., Pei, H., Xie, C., Kang, M., Zhang, C., … & Li, B. (2023). DecodingTrust: A Comprehensive Assessment of Trustworthiness in GPT Models. arXiv preprint arXiv:2306.11698. Weizenbaum, J. (1966). ELIZA—a computer program for the study of natural language communication between man and machine. Communications of the ACM , 9(1), 36-45. https://doi.org/10.1145/365153.365168 Wilks, Y. (1978). Making preferences more active. Artificial intelligence, 11(3) , 197-223. https://doi.org/10.1016/0004-3702(78)90001-2 Williams, C. (2023). Hype, or the future of learning and teaching? 3 Limits to AI's ability to write student essays. London School of Economics internet blog . https://kar.kent.ac.uk/99505/. Woods, W. A. (1970). Transition network grammars for natural language analysis. Communications of the ACM, 13(10) , 591-606. https://doi.org/10.1145/355598.362773 Zhang, H., Wu, C., Xie, J., Kim, C., & Carroll, J. M. (2023). QualiGPT: GPT as an easy-to-use tool for qualitative coding. arXiv preprint arXiv:2310.07061 . Zhang, K., & Aslan, A. B. (2021). AI technologies for education: Recent research & future directions. Computers and Education: Artificial Intelligence, 2 , 100025. https://doi.org/10.1016/j.caeai.2021.100025 Zhao, X. (2022). Leveraging artificial intelligence (AI) technology for English writing: Introducing Wordtune as a digital writing assistant for EFL writers. RELC Journal . https://doi.org/10.1177/00336882221094089 Zhou, X., Chen, Z., Jin, X. and Wang, W.Y. (2021), “HULK: an energy efficiency benchmark platform for responsible natural language processing”, Proceedings of the Conference of the European Chapter of the Association for Computational Linguistics: System Demonstrations, Vol. 16 , pp. 329-336. https://doi.org/10.18653/v1/2021.eacl-demos.39 Additional Declarations No competing interests reported. Supplementary Files Appendices.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5534554","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":392255556,"identity":"9c4bc0de-ad6a-4380-bee0-fdc02174f723","order_by":0,"name":"Esmaeel Ali Salimi","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAxElEQVRIiWNgGAWjYLCCBwYMPPwwDhtRWhKAWiQbmEnSAsQGB5iJdBM/+/GHHxIKtskYHz9/TIKhxo6BT/oAfi2SPTnGEgkGt3nMziSzSTAcS2Zg40vAr8XgQA4DRMsBkBa2AwxsPAQcZn/++eMfIC3G/Y+BWv4RocVAIsEMbIuBBNAWxjYitEjceGNmAdIiceOxsUViXzIPQS38/emPb3z4c9uevz/x4Y0P3+zk5HsIaEEFCQwMhOwYBaNgFIyCUUAMAABCpjg5aMXEuAAAAABJRU5ErkJggg==","orcid":"","institution":"Allameh Tabataba'i University","correspondingAuthor":true,"prefix":"","firstName":"Esmaeel","middleName":"Ali","lastName":"Salimi","suffix":""},{"id":392255557,"identity":"e5670685-eeec-4ddd-b8fa-49cd818b78b5","order_by":1,"name":"Mahdi Hajinia","email":"","orcid":"","institution":"Allameh Tabataba'i University","correspondingAuthor":false,"prefix":"","firstName":"Mahdi","middleName":"","lastName":"Hajinia","suffix":""}],"badges":[],"createdAt":"2024-11-27 10:38:25","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5534554/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5534554/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":72335654,"identity":"30cc40f5-fe1a-453b-bf46-19b722555519","added_by":"auto","created_at":"2024-12-25 15:42:11","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":17781,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eThe Process of Inclusion and Exclusion of Students\u003c/em\u003e\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-5534554/v1/9e074399a0634eaa1bac81d0.png"},{"id":72335653,"identity":"d3fe2fbc-5025-4eb9-b568-4e39bca21c84","added_by":"auto","created_at":"2024-12-25 15:42:11","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":20549,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eFlowchart of Instructors’ Participation in the Study\u003c/em\u003e\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-5534554/v1/7af1483844cdfe5b1843863e.png"},{"id":72336705,"identity":"3f46fb79-e70d-49f2-868b-5d4cc85c45f9","added_by":"auto","created_at":"2024-12-25 15:50:11","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":38431,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eTechnology Acceptance Model (TAM), Taken from Davis et al. (1989)\u003c/em\u003e\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-5534554/v1/b6b612e80684253f7e534133.png"},{"id":82812791,"identity":"2c28ab1e-969f-4dbb-a564-fcc53c62f666","added_by":"auto","created_at":"2025-05-15 13:47:03","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":860940,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5534554/v1/e5f67998-2278-4797-9e8d-f04abdca667b.pdf"},{"id":72335661,"identity":"a108e6fb-6097-41da-a5b5-d47bfde9bc7f","added_by":"auto","created_at":"2024-12-25 15:42:11","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":15183,"visible":true,"origin":"","legend":"","description":"","filename":"Appendices.docx","url":"https://assets-eu.researchsquare.com/files/rs-5534554/v1/0f314c7088244e1efd796474.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"\u003cp\u003eLLMs and Academic Writing in Practice: Exploring Participants’ Utilization of GPT during an AI-Assisted Course on Writing Research Papers\u003c/p\u003e","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eIn the field of education, the ongoing quest for innovative teaching methodologies persists among educators, paralleled by students' perennial interest in adopting new pedagogical tools (Firmin, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). The emergence of Artificial Intelligence (AI), specifically within Artificial Intelligence in Education (AIEd), has transformed the educational landscape (Chen et al., \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Zhang \u0026amp; Aslan, \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). In this regard, Large Language Models (LLMs), at the intersection of AI and linguistics, allow computers to process and generate human languages (Khurana, 2023), and provide language learners with opportunities towards betterment of language learning process. As a category of AIs stemmed from LLMs, Generative Pre-trained Processors (GPT), exemplified by GPT 4, demonstrate proficiency in generating human-like languages (Qin et al., \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) and has recently come to the attention of many researchers across various disciplines.\u003c/p\u003e \u003cp\u003eIn line with the research available on application of LLMs in education (e.g., Derakhshan \u0026amp; Ghiasvand, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), this research focuses on their impacts, particularly GPT-4 for its ease of access and popularity, on scholarly writing in researchers\u0026rsquo; L2 (i.e., English), specifically in research paper composition. While existing literature explores broader dimensions of GPT in education, this study addresses a gap by examining its integration into real classrooms, particularly in scholarly writing instruction. The study aims to elucidate the dynamics of GPT within education, considering its effects on educators' and students' perspectives. It emphasizes identifying challenges and how these models improve researchers\u0026rsquo; productivity and efficiency, anticipating a potential paradigm specifically in the field of academic writing. This research highlights GPT\u0026rsquo;s role in education, influencing teaching, learning, and policy, with practical outcomes for all academic stakeholders. It aims to provide a nuanced understanding of GPT's implications in contemporary education, specifically in scholarly writing instruction.\u003c/p\u003e \u003cp\u003eThe research questions to be answered by this study are:\u003c/p\u003e \u003cp\u003e \u003cb\u003eRQ1.\u003c/b\u003e What are participants\u0026rsquo; perceptions (positive, negative, and mixed) towards applying GPT models in classrooms of scholarly writing after utilizing such models during a course?\u003c/p\u003e \u003cp\u003e \u003cb\u003eRQ2.\u003c/b\u003e How introducing GPT to classrooms would increase productivity and efficiency of researchers in their L2 academic writing?\u003c/p\u003e \u003cp\u003e \u003cb\u003eRQ3.\u003c/b\u003e What challenges and limitations would participants deal with while using such models?\u003c/p\u003e"},{"header":"2. Literature Review","content":"\u003cp\u003eNatural Language Processing (NLP) stands as a multifaceted domain within Artificial Intelligence (AI), converging elements from linguistics, computer science, and mathematics (Jiang \u0026amp; Lu, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Originating in the 1950s, NLP employs computational methodologies to analyze and represent human language, with the overarching objective of achieving language processing akin to human capabilities (Jurafsky \u0026amp; Martin, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Liddy, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2001\u003c/span\u003e). Manning and Sch\u0026uuml;tze (\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e1999\u003c/span\u003e) elaborate, defining NLP as an exploration of algorithms enabling computers to comprehend human language. Scholars broadly define NLP as an inquiry into computer handling of written or spoken language for practical purposes (Johri et al., \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). The historical trajectory of NLP models (which directly influenced the emergence of LLMs) unfolds with the development of the first machine translation (MT) system in 1954, marking an early milestone (Hutchins, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e1986\u003c/span\u003e). However, significant strides in NLP techniques gained momentum only in the 1980s and 1990s. Early MT systems primarily relied on dictionary lookups, emphasizing vocabulary and word order distinctions between languages (Liddy, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2001\u003c/span\u003e). It was Chomsky's introduction of generative grammar in \"Syntactic Structures\" (1957) which provided a deeper understanding of linguistics, fostering the emergence of additional NLP facets like speech recognition (Liddy, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2001\u003c/span\u003e). After Chomsky\u0026rsquo;s contributions, the 1960s witnessed substantial linguistic advancements, prompting investigations into the computational representation of meaning (Wilks, \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e1978\u003c/span\u003e). Prototype systems such as ELIZA, SHRDLU, PARRY, and LUNAR contributed to theoretical progress of this idea (Weizenbaum, \u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e1966\u003c/span\u003e; Cerf, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e1973\u003c/span\u003e; Woods, \u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e1970\u003c/span\u003e). As a significant advancement in those years, the 1980s prompted a shift towards comprehensive NLP programs, addressing shortcomings of isolated solutions (Liddy, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2001\u003c/span\u003e). From the 1990s onwards, NLP entered a prosperous era, marked by significant growth fueled by technological advancements, including the internet, high-speed computers, and greater access to large volumes of electronic text (Jiang \u0026amp; Lu, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). The availability of vast web data has propelled deep learning models to unprecedented precision in NLP benchmarks (Brown et al., \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAs models rooted in NLP, LLMs such as GPT emerge as prominent deep learning models, boasting 175\u0026nbsp;billion parameters in GPT-3 (Brown et al., \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Although GPT models benefit from various technologies such as deep learning, machine learning, and so on, in this study, due to its more relevance to the field of applied linguistics, NLP has been covered and reviewed as one of the significant contributors of GPT models. GPT's pre-training process encompasses learning grammar, syntactic knowledge, common sense, and world knowledge (Liu et al., \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Hewitt and Manning, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Categorized as a generative model, GPT-3's capabilities transcend language generation, incorporating transfer learning principles (Pavlik, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Pre-trained language models, including GPT-3, fall into three categories: unidirectional for natural language generation, bidirectional for understanding, and hybrid models combining both (Liu et al., \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). GPT-3's impact on machine learning suggests unidirectional models can excel in both natural language understanding and generation (Brown et al., \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Its success has led to applications in diverse fields, facilitated by transfer learning principles (King, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). GPT's merits include its ability to generate high-quality language, adapt language level and tone, and facilitate specific writing tasks (Lund \u0026amp; Wang, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Bishop, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). However, limitations such as potential biases, context comprehension issues, and high resource demands, and ethical issues have been identified (Zhou et al., \u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Floridi \u0026amp; Chiriatti, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eGPT has attracted attention in various disciplines, offering potential benefits in writing research papers (Biswas, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Lund and Wang, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Researchers have explored its applications in enhancing productivity, imitating writing styles, and addressing specific parts of papers like literature reviews and abstracts (Chen, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Gao et al., 2022; Wang et al., \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). While acknowledging its utility, concerns about ethical and legal issues persist, urging responsible use and consideration of potential consequences (Bishop, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Lund and Wang, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). The research landscape indicates a need for further exploration, particularly in understanding the nuanced impact of GPT on scholarly writing in diverse contexts. To borrow examples from fields other than applied linguistics, in the medical field, ChatGPT shows promise in assisting healthcare professionals by providing information, generating reports, and aiding in documentation (O\u0026rsquo;Connor \u0026amp; ChatGPT, 2023; D\u0026rsquo;amico et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). However, the story of GPT models is not unanimously believed to be optimistic, and there are downsides to it as well. The integration of GPT into journalism raises questions about the authenticity and objectivity of machine-generated content. While GPT can assist in drafting articles and summarizing information, the potential for biased reporting and the need for human oversight are crucial considerations (Pavlik, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn the educational landscape, ChatGPT has demonstrated its value as a teaching assistant, aiding educators in various tasks (Baidoo-Anu \u0026amp; Owusu Ansah, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Its ability to generate practice questions, provide explanations, and offer feedback on assignments has the potential to revolutionize the learning experience. However, concerns about the ethical implications of relying heavily on AI in education have been raised, questioning issues of accountability and the potential impact on critical thinking skills (Williams, \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eChatGPT\u0026rsquo;s potential in education and other fields is recognized, despite mixed feelings among educators (Baidoo-Anu \u0026amp; Owusu Ansah, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Its utility has been explored in various domains including education (Williams, \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Tate, 2023), engineering (Qadir, \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), journalism (Pavlik, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), and medicine (Nisar \u0026amp; Aslam, \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; O\u0026rsquo;Connor \u0026amp; ChatGPT, 2023; D\u0026rsquo;amico et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Ali et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). It aids researchers in writing papers (Alkaissi \u0026amp; McFarlane, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), with a study by Biswas (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) focusing on its influence on medical writing. Lund and Wang (\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) found it beneficial for information management and content creation, enhancing search services, aiding reference services, facilitating cataloging, and contributing to content creation. Chen (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) highlighted its role in collaborative writing, generating ideas, offering suggestions, and emulating writing styles. This fosters interdisciplinary collaboration and innovation. However, ethical use concerns persist, with Bishop (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) emphasizing responsible AI use, especially where machine-generated content can sway public opinion. Biases, context comprehension issues, and resource demands call for careful deployment.\u003c/p\u003e \u003cp\u003eDespite extensive research on GPT models\u0026rsquo; uses, more qualitative studies are needed, particularly in different areas of English language teaching. Current research lacks conclusive findings on the benefits for teachers in academic writing instruction. This study aims to fill this gap, exploring changes in participants\u0026rsquo; perceptions and identifying potential opportunities and challenges through rigorous qualitative analysis, and study various implications such models would have in specifically teaching academic writing and writing research papers.\u003c/p\u003e"},{"header":"3. Methodology","content":"\u003cp\u003e\u003cstrong\u003e3.1 \u0026nbsp; \u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eResearch Design\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn this study, we investigated the application of AI, specifically LLMs, in academic writing classrooms. The research employed phenomenographic qualitative research design to systematically gather and analyze data obtained from participants in the study. This design has also been adopted by other researchers in the field in order to investigate participants’ perceptions regarding application of LLMs in teaching and learning English (Derakhshan \u0026amp; Ghiasvand, 2024). Given the lack of empirical work on the impacts of LLMs in scholarly writing, the study aims to study participants’ perceptions towards these models following a phenomenographic point of view due to its potential and use in investigating how people perceive real world phenomena following a naturalistic and explorative approach (Hajar, 2021; Sin, 2010). In addition to that, Creswell's recommendation (2013) supports the use of a qualitative lens for transformative outcomes in unexplored research areas. Given the diverse perspectives and understandings that researchers may hold regarding LLMs and AI in education, this study employed a qualitative research design to explore participants’ views on the potential benefits and limitations of such models.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.2 \u0026nbsp; \u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eParticipants\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA total of 20 (11 male and 9 female) university students from 2 different universities with advanced English proficiency (IELTS score\u0026gt;7.5) were selected from 53 volunteers based on the inclusion criteria mentioned in Table 1 below following a criterin sampling approach. The process, detailed in Figure 1, involves screening volunteers to ensure language proficiency and academic writing criteria are met. Additionally, 5 experienced ELT instructors (3 male and 2 female) (Figure 2), were chosen through convenience sampling (Mackey \u0026amp; Gass, 2016) in order to teach scholarly writing classes.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 1.\u003c/strong\u003e \u003cem\u003eInclusion Criteria for Students\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u003cimg src=\"data:image/png;base64,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\" width=\"842\" height=\"231\"\u003e\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eFrom the initial 53 volunteers, 20 students meeting the inclusion criteria were selected for the study. The inclusion and exclusion criteria aim to ensure a homogeneous group with a consistent background for a focused investigation.\u003c/p\u003e\n\u003cp\u003eTo hold classes, 5 ELT instructors with experience in scholarly writing instruction at a university level were selected. The instructors, all possessing related university degrees, were chosen based on their qualifications and willingness to participate in the study as will be detailed in Figure 2 below.\u003c/p\u003e\n\u003cp\u003eAll ELT instructors have been teaching at higher education level for more than 10 years and have run various courses related to academic writing and writing research papers. The recruitment process is detailed in Figure 2.\u003c/p\u003e\n\u003cp\u003e3.3 \u003cstrong\u003eInstruments and Materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003e3.3.1\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e\u003cem\u003eClassroom Observations\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo better understand how the process of the treatment goes on, and to guarantee that instructors adhered to the course design, all classes were observed by two observers. Moreover, to increase the quality of qualitative data obtained and to analyze the classes in the next phases of the study, after receiving consent from teachers, classes were audio recorded. These classroom observations provided researchers with the opportunities to gain genuine first-hand insights from classrooms (Cohen et al., 2013). Ethical considerations in this step, including obtaining informed consent and providing the right to withdraw from the study, were all considered and discussed with participants. To ensure reliability of classroom observations, inter-rater consistency was assessed through the re-analysis of randomly selected sessions and in all instances, two observers had consensus according to taking notes and completing the checklists. These classroom observations helped us to attain relative degrees of objectivity in data collection and analysis. Although the nature of this study posits a perception seeking inquiry, direct field observations helped us to add some objectivity to the results of the study and increased the validity of results of the study.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e3.3.2\u0026nbsp;\u003c/em\u003e\u003cstrong\u003e\u003cem\u003eSemi-structured Interviews\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSemi-structured interviews (Mackey \u0026amp; Gass, 2016) are used in this study to investigate participants’ perceptions after using GPT during the course of study. A set of questions, refined based on expert feedback, was prepared by researchers for these interviews. They explored participants’ experiences with GPT, their perceptions of its use, and their views on AI-assisted teaching and learning. Following steps proposed by Creswell (2013) in conducting qualitative research, interview questions in compliance with research objectives were developed by researchers. These questions were delivered to three experts with experience in conducting qualitative research and their comments and feedbacks on questions’ content validity in terms of language suitability, pertinence, and clarity were considered and incorporated. Interviews with students were held online using Skype platform. Instructor interviews were face to face and interactive and were recorded on an iPhone.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e3.3.3\u0026nbsp;\u003c/em\u003e\u003cstrong\u003e\u003cem\u003eAudio-journals\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAudio-journals were employed to capture participants' experiences during the treatment. After each session, all participants, teachers and students, were required to record an audio-journal and send it to the researcher to be transcribed and analyzed. These journals offered rich insights and aligned with Breakwell's recognition of their value in data collection. As Breakwell (2012) highlighted, audio journals are a great source of obtaining valuable information particularly when it comes to investigating the process of one treatment. They allow participants to freely express themselves and make comments more accurately. This instrument is invaluable in qualitative research since provides participants with opportunities to keep record of the process and gives them sufficient time to elaborate on their thoughts (Breakwell, 2012). Since the course was run for 5 sessions, each participant had to send 5 files as their audio journal, and after the course a number of 125 files (100 from students and 25 from instructors) were received.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e3.3.4\u0026nbsp;\u003c/em\u003e\u003cstrong\u003e\u003cem\u003eFocus Group Discussions\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePost-research Skype focus group discussions (FGDs) involving teachers and students were conducted. In these FGDs, participants discussed their GPT models’ utilization experiences and challenges, and all discussions were recorded for data analysis. In qualitative studies, this type of discussions provides researchers with methodological triangulation, helping them to enhance their data reliability and add new insights to their research (Denzin, 1973; Creswell, 2013).\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003eAs for the questions used in this stage, we focused on the semi-structured interview questions and tried to elaborate on them by discussing these questions in a less structured format with participants.\u003c/p\u003e\n\u003cp\u003e3.4 \u003cstrong\u003eCourse Design\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe course comprised five sessions, each focusing on a research paper section. While running classes, instructors followed a standard plan, integrating GPT based activities into classes and homework, and encouraging students to use GPT in order to complete their tasks. The first session covered introductions and literature reviews, the second focused on methodology, the third on data analysis and collection, the fourth on discussion and conclusion, and the final session reviewed course content and focused on abstract writing. Each session included in-class GPT based activities and homework to reinforce students’ understanding and encourage them to use GPT in order to write a section of a research paper. In each of the five sessions of the course, the instructor gave a lecture on the content (e.g., how to write a literature review) and asked students to use GPT in the classroom to practice writing that section with the aid of GPT. It was students’ decision how to adopt GPT and how to use it in the classroom. Then, as data for this study, students kept an audio journal of how they used GPT to write that specific section of an article in their classes. Homework was also given to students and they had to use GPT at home in order to complete required tasks (which were generally product-oriented). At the same time, teachers monitored students’ activities in class and outside the class and they kept their own audio journals as well. Besides, similarly to students, teachers were asked to use GPT while preparing for their lectures, designing tasks, and giving homework to students. All classes were observed by researchers, so that we made sure instructors adhered to the plan, and participants’ perceptions were generally affected by the course content and design.\u003c/p\u003e\n\u003cp\u003e3.5 \u003cstrong\u003eData Collection Procedure\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study used semi-structured interviews, observations, audio journals, and FGDs to collect data. Before the study, interviews with teachers and students (participants were interviewed individually) ensured equal background knowledge and if participants are fully aware of course design and the treatment process. Interviews were piloted and improved based on feedback from 2 interviewees and 3 applied linguistics experts. After conducting these initial interviews, 5 classes of 4 students were formed. Instructors taught 5 sessions of scholarly writing using GPT-4. To ensure reliability, a 90-minute workshop was held before the courses began to align instructors with the research objectives. Both teachers and students were trained to keep audio journals during the treatment, and classes were observed and recorded for reliability of field notes obtained from these observations. Post-treatment, participants attended the main interviews and FGDs, so that researchers studied their perceptions and experiences after using GPT. As for FGDs, two sessions each for teachers and students were held on Skype to discuss their opinions on the course and GPT’s contribution to improving research paper writing. All interviews and discussions were recorded and transcribed by both researchers, and coded for analysis.\u003c/p\u003e\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n"},{"header":"4. Data Analysis","content":"\u003cp\u003eThis study uses a phenomenographic approach towards qualitative data analysis as it complies with the objectives of this study. In this study, we analyzed obtained data through content and thematic analysis to attain an understanding of participants’ perceptions. As has been emphasized by Braun and Clarke (2006), content analysis deals with the presence and frequency of certain words, themes, or concepts within qualitative data. Thematic analysis then helps researchers to find themes and patterns across data and investigate relationships between these themes. Moreover, to analyze data systematically, a framework for phenomenographic data analysis (PDA) proposed by Stenfors-Hayes et al. (2013) adopted. This model includes seven steps of data familiarization, condensation, comparing, contrasting, grouping, articulating, and final labeling.\u003c/p\u003e\u003cp\u003eData analysis was conducted using MAXQDA software for systematic coding and theme organization in this study. Four data sets: semi-structured interviews, observations, audio-journals, and focus-group discussions were transcribed, coded, and inputted into MAXQDA for pattern identification and thematic analysis. Due to the qualitative nature of this study, and as has been emphasized by Lincoln and Guba (1985) in their seminal book, principles of trustworthiness and naturalistic investigation were considered. The first principle employed was member checking, wherein participants were requested to review their transcripts and our interpretations of their responses. Secondly, inter-coder agreement was calculated based on the independent coding of a randomly selected twenty percent of the data by the second coder. Results showed that there was a 97% agreement between coders. The third principle, confirmability, was maintained by asking an experienced L2 researcher meticulously review the data analysis phase. As another important principle, we tried to establish credibility through interactive interviews, peer debriefing, orientation sessions before treatment, and the bracketing of personal experiences and perceptions. To enhance the dependability of the findings as another principle, the researchers conducted content and thematic analyses, iterative data transcription and coding, and provided a coherent description of the methodological steps undertaken. Finally, the transferability of the findings was bolstered by providing comprehensive descriptive data, enabling future researchers to replicate and recontextualize the study. Regarding researcher positionality, it is important to acknowledge that while the researchers strived to maintain neutrality in this qualitative study, complete objectivity is inherently challenging in such designs.\u003c/p\u003e\u003cp\u003e5. \u003cstrong\u003eFindings\u003c/strong\u003e\u003c/p\u003e\u003cp\u003eIn the exploration of teachers and students’ perceptions towards the integration of LLMs in academic writing classrooms, a systematic approach categorized responses of all participants into positive, negative, and mixed viewpoints. Quantitative analysis, as illustrated in Table 2 below, provided a comprehensive overview of perceptions, with 51.43% positive, 28.57% negative, and 20.00% mixed responses throughout the 70 relevant codes.\u003c/p\u003e\u003cp\u003e\u003cstrong\u003eTable 2.\u003c/strong\u003e \u003cem\u003ePost-treatment Positive, Negative, and Mixed Perceptions\u003c/em\u003e\u003c/p\u003e\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"100%\"\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 13px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003eThemes\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003eFrequency\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003ePercentage\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd rowspan=\"11\" valign=\"top\" style=\"width: 13px;\"\u003e\n \u003cp\u003e\u003cem\u003ePositive\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n \u003cp\u003e\u003cem\u003ePerceptions (\u003c/em\u003e\u003cem\u003e51.43%)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003eAd-Free Environment\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e7\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e19.45\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003eUsing Technology Is Good\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e6\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e16.67\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003eEngaging Interface and Unique Interaction Pattern\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e6\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e16.67\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003eInteractive Nature Of GPT\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e5\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e13.89\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003eReducing Workload\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e11.11\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003eWelcoming It as An Advancement\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e8.34\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003eHelps In Many Different Ways\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e5.56\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003eAlways Available\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e5.56\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003eImproves Communication Skills\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e2.78\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTOTAL\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e36\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e100.00\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd rowspan=\"8\" valign=\"top\" style=\"width: 13px;\"\u003e\n \u003cp\u003e\u003cem\u003eNegative Perceptions\u003c/em\u003e\u003c/p\u003e\n \u003cp\u003e\u003cem\u003e(\u003c/em\u003e\u003cem\u003e28.57%)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003eMore Research Needed on Its Effectiveness and Impacts\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e10\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e50.00\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003eChanges the Whole Educational System and Undermines Teachers and Educators\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e10.00\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003eNot Feeling Comfortable with Technology\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e10.00\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003eNo Training Accessible for Majority of People\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e15.00\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003eMakes People Lazy\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e10.00\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003eJust Do Not Like It\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e5.00\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTOTAL\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e20\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e100.00\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd rowspan=\"5\" valign=\"top\" style=\"width: 13px;\"\u003e\n \u003cp\u003e\u003cem\u003eMixed Perceptions\u003c/em\u003e\u003c/p\u003e\n \u003cp\u003e\u003cem\u003e(\u003c/em\u003e\u003cem\u003e20.00%)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003eIt Can Be Both Good and Evil\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e6\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e42.86\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003eToo Complicated\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e28.57\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003eTechnology Use in Schools\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e14.29\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003eMore Time Needed to Decide\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e14.29\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTOTAL\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e14\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e100.00\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/table\u003e\u003cp\u003eThe breakdown revealed that 51.43% of participants held positive perceptions, much appreciating GPT for its ad-free nature (e.g., Student 3 mentioned \u003cem\u003e“… I also hate ads which come with the websites that google introduces\u003c/em\u003e”) and engaging qualities (e.g., Teacher 2 mentioned \u003cem\u003e“if students see that it really helps them, they would become more engaged in the process of learning”\u003c/em\u003e). Noteworthy sentiments included participants welcoming GPT with open arms (e.g., Teacher 1 highlighted \u003cem\u003e“… I welcome it with open arms”\u003c/em\u003e) and recognizing its benefits in both teaching and learning.\u003c/p\u003e\u003cp\u003eConversely, 28.57% expressed negative perceptions, citing concerns about changes in educational system and the potential erosion of teachers' authority (e.g., Teacher 3 highlighted, \u003cem\u003e“there might not be a need for teachers anymore if the uses of AI are more prevalent in future”\u003c/em\u003e), discomfort with technology (e.g., Teacher 1 mentioned \u003cem\u003e“it is not still prevalent and not many teachers use it (they might be afraid of it)”\u003c/em\u003e), and reservations about GPT's impact on students' skill development (e.g., Teacher 4 with reference to limitations and challenges mentioned: \u003cem\u003e“it makes students at the end of the term lack necessary skills for writing due to the overreliance on GPT”\u003c/em\u003e). However, half of the negative perceptions were coming from the perception that it is a relatively new field and must be researched in more depth (e.g., Student 6 mentioned \u003cem\u003e“we do not know if it really works”\u003c/em\u003e).\u003c/p\u003e\u003cp\u003eMixed perceptions, accounting for 20.00% of responses, reflected indecisiveness among participants. Ambiguities centered on the multifaceted nature of GPT's impact (e.g., student 8 mentioned \u003cem\u003e“GPT is everywhere and can do anything for us. I don’t know if it’s good or bad”),\u0026nbsp;\u003c/em\u003euncertainty about its role in society since it is a mixed blessings for them, and considerations related to national rules on technology use in schools (as student 14 mentioned \u003cem\u003e“I can’t take my laptop or mobile phone to school with myself, I like it but I have problems with it too”\u003c/em\u003e).\u003c/p\u003e\u003cp\u003eIn the investigation of LLMs and their impact on researchers' productivity and efficiency in writing papers, a meticulous open coding process identified 10 main categories. Considering the common and non-technical meaning of efficiency and productivity, in this study, we consider them mainly as advantages GPT brings about into classrooms that would help researchers to write academic papers with more ease, more effectively, and improve their papers’ quality.\u003c/p\u003e\u003cp\u003eTable 3 provides a comprehensive overview of these categories, highlighting the frequency distribution of codes within each distinct aspect. The high frequency of codes in this category (180) showed that participants deem GPT to significantly boost productivity and efficiency of researchers while writing academic papers.\u003c/p\u003e\u003cp\u003e\u003cstrong\u003eTable 3\u003c/strong\u003e\u003cstrong\u003e\u003cem\u003e.\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003cem\u003eEfficiency and Productivity\u003c/em\u003e\u003c/p\u003e\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"84%\"\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003e\u0026nbsp;Themes\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 18px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003eFrequency\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 21px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003ePercentage\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u003cem\u003eBoosts the Speed of Writing\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 18px;\"\u003e\n \u003cp\u003e54\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 21px;\"\u003e\n \u003cp\u003e30.00\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u003cem\u003eImmediate Feedback on Writing\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 18px;\"\u003e\n \u003cp\u003e36\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 21px;\"\u003e\n \u003cp\u003e20.00\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u003cem\u003eImproves Moves and Steps in a Section\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 18px;\"\u003e\n \u003cp\u003e24\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 21px;\"\u003e\n \u003cp\u003e13.33\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u003cem\u003eSuggestions on Academic Language\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 18px;\"\u003e\n \u003cp\u003e23\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 21px;\"\u003e\n \u003cp\u003e12.77\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u003cem\u003eCategorization of Writing\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 18px;\"\u003e\n \u003cp\u003e12\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 21px;\"\u003e\n \u003cp\u003e6.67\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u003cem\u003eOrganizing, Planning, and Outlining\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 18px;\"\u003e\n \u003cp\u003e9\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 21px;\"\u003e\n \u003cp\u003e5.00\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u003cem\u003eIdeas for Getting Started and How to Plan\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 18px;\"\u003e\n \u003cp\u003e9\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 21px;\"\u003e\n \u003cp\u003e5.00\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u003cem\u003eExtracts Keywords\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 18px;\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 21px;\"\u003e\n \u003cp\u003e2.22\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u003cem\u003eAn Asset in Systematic Work\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 18px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 21px;\"\u003e\n \u003cp\u003e1.67\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u003cem\u003eHelps to Better Find Gaps in Research\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 18px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 21px;\"\u003e\n \u003cp\u003e1.67\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u003cem\u003eMeeting the Guidelines of a Journal\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 18px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 21px;\"\u003e\n \u003cp\u003e1.11\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u003cem\u003ePutting Ideas in Sentences\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 18px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 21px;\"\u003e\n \u003cp\u003e0.55\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003eTOTAL\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 18px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e180\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 21px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e100.00\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/table\u003e\u003cp\u003eThe study revealed that GPT has the potential to significantly enhance researchers' productivity and efficiency across various dimensions. The most prominent category, with a high frequency of 30.00%, centered on how GPT increases the speed of tasks (e.g., when Teacher 2 was asked what was significant about GPT and why you liked it, she mentioned in a decisive tone, \u003cem\u003e“it saves your time”\u003c/em\u003e). This table further delineates the breakdown of categories, emphasizing the multifaceted contributions of GPT. Providing researchers with immediate feedback on their writing (20.00%) (e.g., \u003cem\u003e“you can receive feedback from GPT on your wiring, and it helps you to improve your writing”\u003c/em\u003e (Participant 11)), improving moves and steps in a writing (13.33%) (e.g., \u003cem\u003e“it helped me to better tailor my text for the audience and know how to write different sections after each other”\u0026nbsp;\u003c/em\u003e(Participant 5)), and enhancing writing style in terms of being more academic (12.77%) (e.g., \u003cem\u003e“… or consulting with it while I want to improve my style”\u003c/em\u003e (Participant 9)) emerged as significant areas where GPT showcased its efficacy.\u003c/p\u003e\u003cp\u003eIn the next phase of data analysis, the examination shifted towards the challenges and limitations inherent in leveraging GPT models as the primary instructional aid for teaching academic writing. The qualitative analysis brought forth 11 distinct categories of challenges, as depicted in Table 4, illuminating the potential hurdles faced by both teachers and students.\u003c/p\u003e\u003cp\u003e\u003cstrong\u003eTable 4.\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cem\u003eChallenges and Limitations\u003c/em\u003e\u003c/p\u003e\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"93%\"\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 67px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003e\u0026nbsp;Themes\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003eFrequency\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003ePercentage\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 67px;\"\u003e\n \u003cp\u003e\u003cem\u003eOverreliance on AI\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e19\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e20.21\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 67px;\"\u003e\n \u003cp\u003e\u003cem\u003eAccurate Prompting Must be Learnt\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e18\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e19.15\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 67px;\"\u003e\n \u003cp\u003e\u003cem\u003eNot everyone Can Use GPT Effectively\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e13\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e13.82\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 67px;\"\u003e\n \u003cp\u003e\u003cem\u003eMust Always be Doublechecked in Terms of Scientific Accuracy\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e12\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e12.77\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 67px;\"\u003e\n \u003cp\u003e\u003cem\u003eFabricating Non-Existing Information\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e10\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e10.64\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 67px;\"\u003e\n \u003cp\u003e\u003cem\u003eStill Not Complete and Accurate\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e6\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e6.38\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 67px;\"\u003e\n \u003cp\u003e\u003cem\u003eHigh-Tech Plagiarism\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e6\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e6.38\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 67px;\"\u003e\n \u003cp\u003e\u003cem\u003ePatchwriting\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e4.26\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 67px;\"\u003e\n \u003cp\u003e\u003cem\u003eNot Sensitive Towards Different cultures\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e3.19\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 67px;\"\u003e\n \u003cp\u003e\u003cem\u003eData from Non-Existing Sources\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e2.13\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 67px;\"\u003e\n \u003cp\u003e\u003cem\u003eGenerates Accurate Data Only in English\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e1.06\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd valign=\"top\" style=\"width: 67px;\"\u003e\n \u003cp\u003e\u003cem\u003eTOTAL\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e94\u003c/p\u003e\n \u003c/td\u003e\u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e100.00\u003c/p\u003e\n \u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/table\u003e\u003cp\u003eAmong the identified challenges, the most significant was the issue of overreliance on AI (20.21%). Participants encountered difficulties due to an excessive dependence on GPT models, emphasizing the necessity for a balanced approach in utilizing these tools effectively (e.g., \u003cem\u003e“I believe it to be so addictive which makes you lazy”\u003c/em\u003e (Teacher 4)). Prompting emerged as another crucial factor, with 19.15% of segments underscoring its pivotal role and difficulties related to it as limiting (e.g., \u003cem\u003e“we should be hyper precise while talking to GPT”\u0026nbsp;\u003c/em\u003e(Participant 14)). Concerns about repeated answers underscored the challenges faced by teachers in generating precise prompts. Education on working with AI (13.82%) and the requirement for always being double-checked (12.77%) emerged as significant challenges, indicating a perceived need for more comprehensive training to effectively utilize GPT models. Concerns related to the fabrication of information (10.64%) and the risk of high-tech plagiarism (term borrowed from Chomsky (2023)) (6.38%) highlighted apprehensions about the accuracy and reliability of information generated by GPT models.\u003c/p\u003e\u003cp\u003eIssues related to the lack of focus on cultural-linguistic points of the language (3.19%) indicated a desire for a more nuanced approach to language instruction (e.g., “I suppose GPT does not care about me in a way that it does not take cultural differences into account (Participant 5)). Additionally, the limitation of generating accurate data only in English (1.06%) (e.g., “I tried to get help from GPT to write Persian texts but my problem is that GPT can only generate accurate data in English, not in any other languages” (Participant 16)) and the existence of non-existing sources (2.13%) (e.g., “… it referred to sth which did not exist at all” (Teacher 3)) underscored the current limitations of GPT models, suggesting areas for improvement.\u003c/p\u003e"},{"header":"5. Discussion","content":"\u003cp\u003eAI models such as GPT have recently come to the attention of many researchers and their impacts on various educational contexts have been investigated. However, a gap remained regarding how such models impact writing research papers after completing a course in which GPT was the main focus of the instruction. As could be expected, after completion of the course, the majority of participants held positive perceptions towards such models and this could explain why FGDs in this study were mainly revolving around the positive aspects of such models. Since positive perceptions predominated, in this study, we investigated in which areas participants (instructors and students) found GPT models to be effective and increase their productivity and proficiency. 10 major themes were extracted with deep support in available literature as will be mentioned below. However, it is nearly impossible to find one teaching method all meritorious without challenges and limitations. 11 distinct categories of challenges and limitations were extracted in this study, and we found that in other areas of language teaching such themes have been studied and supported.\u003c/p\u003e \u003cp\u003eTo answer our first research question regarding participants\u0026rsquo; attitudes, this study revealed participants\u0026rsquo; predominantly positive perceptions of GPT models, influenced by factors such as ad-free usage and extended engagement which are supported by existing literature as well (Munoz et al., 2023; Diwan et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Material development and interaction patterns with models contributed to this positive view. Provided participants held predominantly positive perceptions towards applying such models, a close relationship can be found between their positive perceptions and our second research question regarding impacts of such models on participants productivity and efficiency in writing research papers. This relationship can be justified by the Technology Acceptance Model (TAM), first introduced by Davis et al. in 1989, which examines individual behavior towards technology adoption within various information system frameworks. Rooted in social psychology, TAM explores the interplay between cognitive and affective factors and how they influence users' technology usage. Technology acceptance unfolds in three stages: initial external factors lead to cognitive responses (like perceived usefulness and ease of use); these then shape an affective response (user's intention or attitude towards using the technology), which ultimately influences the user's behavior. The model and relationships are illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e below.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eHowever, as could be expected, negative perceptions exist too, and many of them have been mentioned by other researchers. For example, Wang et al.'s (\u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) research on GPT\u0026rsquo;s trustworthiness and concerns about data leakage supported participants\u0026rsquo; negative perceptions in this regard. Besides, some participants were concerned whether such models inhibit students\u0026rsquo; skill development. Concerns about these models replacing teachers and disrupting traditional teaching methods were prevalent in results, aligning with previous studies in the field (Fuchs, ccd2023; Farazouli et al., \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). All in all, we sought participants\u0026rsquo; perceptions to make sure how utilizing LLMs in classrooms would impact their perceptions, and as results confirmed, in line with previous research available, their perceptions were of a more positive nature and this is promising for the future of AI-Assisted instruction. Possible explanations have been proposed by researchers such as Derakhshan and Ghiasvand (\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) as they emphasized the interactive and engaging nature of these models. However, as expected, addressing these concerns requires proper training and support (Unser, \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), and requires more attention to application of such models from scholars\u0026rsquo; side. Although previous research emphasizing participants\u0026rsquo; positive perceptions exist, this current study differs in terms of offering an experimental course utilizing GPT as the main aid for delivering and learning content. Results of this study answering to the first research question, in addition to previous research, support their findings particularly in the case of scholarly writing through a real classroom experimental design.\u003c/p\u003e \u003cp\u003eIn line with our second research question concerning how these models help improve researchers\u0026rsquo; productivity and efficiency, this study highlights GPT models\u0026rsquo; multifaceted contributions to writing research papers marking a paradigm shift in academic research. To further classify themes extracted, we identified two areas in which GPT helped researchers increase productivity and efficiency: (1) in terms of language and linguistic accuracy and (2) in terms of adhering to academic conventions and helping with completing a research paper. Many of the themes extracted have supports in available literature. For example, in terms of linguistic aids, GPT accelerates writing by increasing the speed of writing, allowing researchers to focus on conceptualization and analysis (Buruk, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) rather than focusing on linguistic aspects of writing. It also aids researchers in text editing and feedback provision, ensuring high-quality academic writing (Carlson, 2023; Shidiq, \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Straume \u0026amp; Anson, \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). In terms of completing research papers, GPT improves research paper discourse and style which helps researchers to adhere to academic conventions (de Rivero et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) more appropriately. Moreover, it supports researchers in planning and structuring papers helping them with moves and steps in writing different sections of a paper. Besides, it contributes to the quality of thematic analysis and categorization in qualitative data analysis (Zhang et al., \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Gamieldien et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Koch, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) and as was mentioned by one participant of the study, \u0026ldquo;\u003cem\u003e\u0026hellip; it can be a good source for asking my research questions\u003c/em\u003e\u0026rdquo;. GPT\u0026rsquo;s keyword generation feature optimizes paper visibility and impact, and it also helps researchers in literature gap identification (Firoozeh et al., \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Ezzelding \u0026amp; El-Dakhakhni, 2020). And as a very important area in which GPT helps significantly, it empowers researchers to better adhere to academic writing guidelines, ensuring proper formatting and citation (Zhao, \u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Salvagno et al., 2022) and therefore, increasing researchers\u0026rsquo; chance of getting published in academic journals. This could be interpreted in line with the dominance of positive perceptions in answer to the first research question. Participants of the study believe GPT to revolutionize the field of scholarly writing since it brings about unprecedented tools to increase researchers\u0026rsquo; productivity and efficiency.\u003c/p\u003e \u003cp\u003eGPT models have revolutionized academic writing, enhancing efficiency and productivity. However, as per our third research question, challenges exist. Overreliance on AI could undermine research exploration (Abd-Alrazaq et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). The effectiveness of GPT depends on well-formulated prompts (Liu et al., \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Alivanistos et al., 2023) that requires proper training and effort. So, education and prior knowledge are essential for effective GPT use (Min et al., \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). GPT\u0026rsquo;s information fabrication is a limitation, raising ethical concerns (Walters \u0026amp; Wilder, \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Mosca, 2023) and challenging research papers validity. Plagiarism risk is another challenge (Dehouche, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) for researchers and publishers which must be investigated more in-depth. GPT\u0026rsquo;s lack of focus on cultural aspects of languages poses challenges and must be investigated through a sociolinguistic point of view (Quian et al., 2021; Johnson et al., \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). The generation of non-existing sources and limitations in generating accurate data in languages other than English were additional challenges participants faced during the course of study. These limitations and challenges are still to be researched by scholars in the field and their multifaceted impacts should be brought to attention of academics. However, with all its limitations and challenges, participants of this current study held a more positive perception towards the application of LLMs in classrooms.\u003c/p\u003e"},{"header":"6. Conclusion","content":"\u003cp\u003eThis study has provided a comprehensive exploration of the impacts of LLMs on scholarly writing. It has shed light on the transformative potential of these models in enhancing researchers\u0026rsquo; productivity and efficiency, and has offered a nuanced understanding of their implications in contemporary education. The study indicates that GPT models can enhance research paper writing, potentially causing a shift in academic writing. However, it also highlights the challenges of integrating these models into classrooms. The study\u0026rsquo;s implications are significant for educators, syllabus designers, students, and researchers, suggesting GPT models can improve teaching methods, anticipate pedagogical changes, impact language skills, and provide a basis for interventions. More importantly, this research would be influential for EAP and ESP instructors and course designers, and help them to better understand the nature of integrating AI in courses tailored for the needs of a specific group of students. Despite progress in understanding GPT\u0026rsquo;s impact on academic writing, more research is needed to understand its full effects. Future research could include longitudinal studies on GPT use in different educational settings, exploring the impacts of various GPT models, and studying GPT\u0026rsquo;s long-term effects on language skills. As AI evolves, continuous research is essential to keep up with developments and their educational implications.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics Approval and Consent to Participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe need for ethical approval was waived off by the nature of the study and the ethical board at Allameh Tabatabei University, Department of Literature and Languages. However, considering the nature of the study, informed consent was received from all participants with right to withdraw from the study at any moment during the process of the research.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of Data and Materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets used and analyzed during the current study are available from the corresponding author on reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors' contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll authors contributed to the process of researching, analyzing, and reporting this paper. However, MH was mainly in charge of code analysis and working with MAXQDA due to his expertise in the field, and EAS designed the method and approach of the study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors' Biography\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eEsmaeel Ali Salimi is a faculty member of Allameh Tabataei university who has published many artcicle in the fields of Applied Linguistics, Learner Factors, Assessment, Innovation in Teaching, and Discourse Analysis in esteemed journals such as Learning and Motivation, Teaching English Language, Journal of Language and Education, and Teflin journal. He is currently working as an associate professor supervising PhD and MA students’ research, and instructs courses related to Research, Innovations in TEFL, and Discourse Studies.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAbd-Alrazaq, A., AlSaad, R., Alhuwail, D., Ahmed, A., Healy, P. M., Latifi, S., \u0026hellip; \u0026amp; Sheikh, J. (2023). Large Language Models in Medical Education: Opportunities, Challenges, and Future Directions. \u003cem\u003eJMIR Medical Education\u003c/em\u003e, 9(1), e48291. https://doi.org/10.2196/48291 \u003c/li\u003e\n\u003cli\u003eAli, J. K. M., Shamsan, M. A. A., Hezam, T. A., \u0026amp; Mohammed, A. A. (2023). Impact of ChatGPT on Learning Motivation: Teachers and Students\u0026rsquo; Voices. \u003cem\u003eJournal of English Studies in Arabia Felix\u003c/em\u003e, 2(1), 41-49. https://doi.org/10.56540/jesaf.v2i1.51 \u003c/li\u003e\n\u003cli\u003eAlivanistos, D., Santamar\u0026iacute;a, S. B., Cochez, M., Kalo, J. C., van Krieken, E., \u0026amp; Thanapalasingam, T. (2022). Prompting as probing: Using language models for knowledge base construction. arXiv preprint arXiv:2208.11057.\u003c/li\u003e\n\u003cli\u003eAlkaissi, H., \u0026amp; McFarlane, S. I. (2023). Artificial hallucinations in ChatGPT: implications in scientific writing. \u003cem\u003eCureus\u003c/em\u003e, 15(2). https://doi.org/10.7759/cureus.35179 \u003c/li\u003e\n\u003cli\u003eBaidoo-Anu, D., \u0026amp; Owusu Ansah, L. (2023). Education in the Era of Generative Artificial Intelligence (AI): Understanding the Potential Benefits of ChatGPT in Promoting Teaching and Learning. Available at SSRN 4337484. https://doi.org/10.2139/ssrn.4337484 \u003c/li\u003e\n\u003cli\u003eBishop, L. (2023). A Computer Wrote This Paper: What ChatGPT Means for Education, Research, and Writing. \u003cem\u003eResearch, and Writing\u003c/em\u003e (January 26, 2023). https://doi.org/10.2139/ssrn.4338981 \u003c/li\u003e\n\u003cli\u003eBiswas, S. (2023). ChatGPT and the Future of Medical Writing. \u003cem\u003eRadiology\u003c/em\u003e, 223312. https://doi.org/10.1148/radiol.223312\u003c/li\u003e\n\u003cli\u003eBraun, V., \u0026amp; Clarke, V. (2006). Using thematic analysis in psychology. \u003cem\u003eQualitative Research in Psychology\u003c/em\u003e, 3(2), 77\u0026ndash;101. https://doi.org/10.1191/1478088706qp063oa \u003c/li\u003e\n\u003cli\u003eBreakwell, G. M. (2012). Diary and Narrative Methods. In G. Breakwell, J. Smith, \u0026amp; D. Wright (Eds.), \u003cem\u003eResearch Methods in Psychology\u003c/em\u003e (4th ed., pp. 111392-410). Sage. https://doi.org/10.1002/9780470776278 \u003c/li\u003e\n\u003cli\u003eBrown, T., Mann, B., Ryder, N., Subbiah, M., Kaplan, J. D., Dhariwal, P., \u0026hellip; \u0026amp; Amodei, D. (2020). Language Models Are Few-Shot Learners. \u003cem\u003eAdvances in Neural Information Processing Systems\u003c/em\u003e, 33, 1877-1901. https://doi.org/10.18653/v1/2020.emnlp-main.375 \u003c/li\u003e\n\u003cli\u003eBuruk, O. O. (2023, October). Academic Writing with GPT-3.5 (ChatGPT): Reflections on Practices, Efficacy and Transparency. In \u003cem\u003eProceedings of the 26th International Academic Mindtrek Conference\u003c/em\u003e (pp. 144-153). https://doi.org/10.1145/3616961.3616992 \u003c/li\u003e\n\u003cli\u003eCarlson, M., Pack, A., \u0026amp; Escalante, J. (2023). Utilizing OpenAI\u0026rsquo;s GPT‐4 for Written Feedback. \u003cem\u003eTESOL Journal\u003c/em\u003e, e759.\u003c/li\u003e\n\u003cli\u003eCerf, V. (1973). Parry Encounters the Doctor (No. rfc439). https://doi.org/10.17487/rfc0439 \u003c/li\u003e\n\u003cli\u003eChen, L., Chen, P., \u0026amp; Lin, Z. (2020). Artificial Intelligence in Education: A Review. \u003cem\u003eIEEE Access\u003c/em\u003e, 8, 75264-75278. https://doi.org/10.1109 \u003c/li\u003e\n\u003cli\u003eChen, T. J. (2023). ChatGPT and Other Artificial Intelligence Applications Speed Up Scientific Writing. \u003cem\u003eJournal of the Chinese Medical Association\u003c/em\u003e. https://doi.org/10.1097/jcma.0000000000000900 \u003c/li\u003e\n\u003cli\u003eChomsky, N. (1957). \u003cem\u003eSyntactic Structures\u003c/em\u003e. The Hague: Mouton. https://doi.org/10.1515/9783112316009\u003c/li\u003e\n\u003cli\u003eChomsky N. Noam Chomsky on ChatGPT: It\u0026rsquo;s \u0026ldquo;Basically High-Tech Plagiarism\u0026rdquo; and \u0026ldquo;a Way of Avoiding Learning\u0026rdquo;. Open Culture. 2023 February 10th, 2023. [Available from: https://www.openculture.com/2023/02/noam-chomsky-on-chatgpt.html. \u003c/li\u003e\n\u003cli\u003eCohen, L., Manion, L., \u0026amp; Morrison, K. (2013). \u003cem\u003eResearch methods in education\u003c/em\u003e. Routledge. https://doi.org/10.4324/9780203720967 \u003c/li\u003e\n\u003cli\u003eCreswell, J. W. (2013). \u003cem\u003eResearch design. Qualitative, quantitative and mixed method approaches\u003c/em\u003e (3rd ed.). Sage. https://doi.org/10.1086/602876 \u003c/li\u003e\n\u003cli\u003eD\u0026rsquo;Amico, R. S., White, T. G., Shah, H. A., \u0026amp; Langer, D. J. (2022). I asked a ChatGPT to write an editorial about how we can incorporate chatbots into neurosurgical research and patient care\u0026hellip;. \u003cem\u003eNeurosurgery\u003c/em\u003e, 10-1227. https://doi.org/10.1227/neu.0000000000002414\u003c/li\u003e\n\u003cli\u003eDavis, F.D., R.P. Bagozzi, and P.R. Warshaw. 1989. User acceptance of computer technology: A comparison of twotheoretical models. Management Science 35: 982 1003. \u003cbr\u003e\u003cem\u003e(PDF) The impact of Perceived Usefulness and Perceived Ease of Use on Online Purchases: A comparison of Buyers and Non-Buyers Perceptions\u003c/em\u003e. Available from: https://www.researchgate.net/publication/278676371_The_impact_of_Perceived_Usefulness_and_Perceived_Ease_of_Use_on_Online_Purchases_A_comparison_of_Buyers_and_Non-Buyers_Perceptions [accessed Dec 06 2024].\u003c/li\u003e\n\u003cli\u003ede Rivero, M., Tirado, C., \u0026amp; Ugarte, W. (2023). Formalstyler: GPT-Based Model for Formal Style Transfer with Meaning Preservation. \u003cem\u003eSN Computer Science\u003c/em\u003e, 4(6), 739. https://doi.org/10.1007/s42979-023-02110-7 \u003c/li\u003e\n\u003cli\u003eDehouche, N. (2021). Plagiarism in the age of massive Generative Pre-trained Transformers (GPT-3). \u003cem\u003eEthics in Science and Environmental Politics\u003c/em\u003e, 21, 17-23. https://doi.org/10.3354/esep00195 \u003c/li\u003e\n\u003cli\u003eDenzin, N. K. (1973). \u003cem\u003eThe research act: A theoretical introduction to sociological methods\u003c/em\u003e. Transaction Publishers. https://doi.org/10.4324/9781315134543\u003c/li\u003e\n\u003cli\u003eDerakhshan, A., \u0026amp; Ghiasvand, F. (2024). Is ChatGPT an evil or an angel for second language education and research? A phenomenographic study of research‐active EFL teachers\u0026rsquo; perceptions. \u003cem\u003eInternational Journal of Applied Linguistics\u003c/em\u003e. https://doi.org/10.1111/ijal.12561 \u003c/li\u003e\n\u003cli\u003eDiwan, C., Srinivasa, S., Suri, G., Agarwal, S., \u0026amp; Ram, P. (2023). AI-based learning content generation and learning pathway augmentation to increase learner engagement. \u003cem\u003eComputers and Education: Artificial Intelligence\u003c/em\u003e, 4, 100110. https://doi.org/10.1016/j.caeai.2022.100110 \u003c/li\u003e\n\u003cli\u003eEzzeldin, M., \u0026amp; El-Dakhakhni, W. (2020). Metaresearching structural engineering using text mining: Trend identifications and knowledge gap discoveries. \u003cem\u003eJournal of Structural Engineering\u003c/em\u003e, 146(5), 04020061. https://doi.org/10.1061/(asce)st.1943-541x.0002523 \u003c/li\u003e\n\u003cli\u003eFarazouli, A., Cerratto-Pargman, T., Bolander-Laksov, K., \u0026amp; McGrath, C. (2023). Hello GPT! Goodbye home examination? An exploratory study of AI chatbots\u0026rsquo; impact on university teachers\u0026rsquo; assessment practices. \u003cem\u003eAssessment \u0026amp; Evaluation in Higher Education\u003c/em\u003e, 1-13. https://doi.org/10.1080/02602938.2023.2241676 \u003c/li\u003e\n\u003cli\u003eFirmin, M. W., \u0026amp; Genesi, D. J. (2013). History and implementation of classroom technology. \u003cem\u003eProcedia-Social and Behavioral Sciences\u003c/em\u003e, 93, 1603-1617. https://doi.org/10.1016/j.sbspro.2013.10.089 \u003c/li\u003e\n\u003cli\u003eFiroozeh, N., Nazarenko, A., Alizon, F., \u0026amp; Daille, B. (2020). Keyword extraction: Issues and methods. \u003cem\u003eNatural Language Engineering\u003c/em\u003e, 26(3), 259-291. https://doi.org/10.1017/s1351324919000457 \u003c/li\u003e\n\u003cli\u003eFloridi, L., \u0026amp; Chiriatti, M. (2020). GPT-3: Its nature, scope, limits, and consequences. \u003cem\u003eMinds and Machines\u003c/em\u003e, 30, 681-694. https://doi.org/10.1007/s11023-020-09548-1 \u003c/li\u003e\n\u003cli\u003eFuchs, K. (2023, May). Exploring the opportunities and challenges of NLP models in higher education: is Chat GPT a blessing or a curse?. In \u003cem\u003eFrontiers in Education\u003c/em\u003e (Vol. 8, p. 1166682). Frontiers. https://doi.org/10.3389/feduc.2023.1166682 \u003c/li\u003e\n\u003cli\u003eGamieldien, Y., Case, J. M., \u0026amp; Katz, A. (2023). Advancing Qualitative Analysis: An Exploration of the Potential of Generative AI and NLP in Thematic Coding. Available at SSRN 4487768. https://doi.org/10.2139/ssrn.4487768 \u003c/li\u003e\n\u003cli\u003eHajar, A. (2021). Theoretical foundations of phenomenography: A critical review. Higher Education Research \u0026amp; Development, 40(7), 1421\u0026ndash;1436. https://doi.org/10.1080/07294360.2020.1833844 \u003c/li\u003e\n\u003cli\u003eHewitt, J., \u0026amp; Manning, C. D. (2019). A structural probe for finding syntax in word representations. In \u003cem\u003eNorth American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL)\u003c/em\u003e. Association for Computational Linguistics. \u003c/li\u003e\n\u003cli\u003eHutchins W. J. (1986). Machine translation: Past, present, future. In \u003cem\u003eMachine Translation\u003c/em\u003e (pp. 3-13). Springer. https://doi.org/10.2307/328480 \u003c/li\u003e\n\u003cli\u003eJiang, K., \u0026amp; Lu, X. (2020, November). Natural language processing and its applications in machine translation: A diachronic review. In \u003cem\u003e2020 IEEE 3rd International Conference of Safe Production and Informatization (IICSPI)\u003c/em\u003e (pp. 210-214). IEEE. https://doi.org/10.1109/iicspi51290.2020.9332458 \u003c/li\u003e\n\u003cli\u003eJohnson, R. L., Pistilli, G., Men\u0026eacute;dez-Gonz\u0026aacute;lez, N., Duran, L. D. D., Panai, E., Kalpokiene, J., \u0026amp; Bertulfo, D. J. (2022). The Ghost in the Machine has an American accent: value conflict in GPT-3. arXiv preprint arXiv:2203.07785. \u003c/li\u003e\n\u003cli\u003eJohri, P., Khatri, S. K., Al-Taani, A. T., Sabharwal, M., Suvanov, S., \u0026amp; Kumar, A. (2021). Natural language processing: History, evolution, application, and future work. In \u003cem\u003eProceedings of 3rd International Conference on Computing Informatics and Networks: ICCIN 2020\u003c/em\u003e (pp. 365-375). Springer Singapore. https://doi.org/10.1007/978-981-15-9712-1_31 \u003c/li\u003e\n\u003cli\u003eJurafsky, D., \u0026amp; Martin, J. H. (2019). \u003cem\u003eSpeech and Language Processing\u003c/em\u003e (3rd ed.). Pearson. \u003c/li\u003e\n\u003cli\u003eKhurana, D., Koli, A., Khatter, K., \u0026amp; Singh, S. (2023). Natural language processing: State of the art, current trends and challenges. \u003cem\u003eMultimedia tools and applications\u003c/em\u003e, 82(3), 3713-3744. https://doi.org/10.1007/s11042-022-13428-4 \u003c/li\u003e\n\u003cli\u003eKing, M.R. (2022), \u0026ldquo;The future of AI in medicine: a perspective from a Chatbot\u0026rdquo;, \u003cem\u003eAnnals of Biomedical Engineering\u003c/em\u003e, https://doi.org/10.1007/s10439-022-03121-w \u003c/li\u003e\n\u003cli\u003eKoch, M. A. (2023). Turning Chaos into Meaning: A Chat GPT-Assisted Exploration of COVID-19 Narratives (Master\u0026rsquo;s thesis, University of Twente).\u003c/li\u003e\n\u003cli\u003eLiddy, E. D. (2001). Natural language processing. In \u003cem\u003eEncyclopedia of Library and Information Science, 2nd Ed\u003c/em\u003e. NY. Marcel Decker, Inc.\u003c/li\u003e\n\u003cli\u003eLincoln, Y. S., Guba, E. G., \u0026amp; Pilotta, J. J. (1985). Naturalistic inquiry. International Journal of Intercultural Relations, 9(4), 438\u0026ndash;439. https://doi.org/10.1016/0147-1767(85)90062-8 \u003c/li\u003e\n\u003cli\u003eLiu, P., Yuan, W., Fu, J., Jiang, Z., Hayashi, H., \u0026amp; Neubig, G. (2023). Pre-train, prompt, and predict: A systematic survey of prompting methods in natural language processing. \u003cem\u003eACM Computing Surveys\u003c/em\u003e, 55(9), 1-35. https://doi.org/10.1145/3560815 \u003c/li\u003e\n\u003cli\u003eLiu, X., Zheng, Y., Du, Z., Ding, M., Qian, Y., Yang, Z., \u0026amp; Tang, J. (2021). GPT understands, too. arXiv preprint arXiv:2103.10385. https://doi.org/10.1016/j.aiopen.2023.08.012 \u003c/li\u003e\n\u003cli\u003eLund, B. D., Wang, T. (2023). Chatting about ChatGPT: how may AI and GPT impact academia and libraries? \u003cem\u003eLibrary Hi Tech News\u003c/em\u003e. https://doi.org/10.1108/lhtn-01-2023-0009 \u003c/li\u003e\n\u003cli\u003eLund, B. D., Wang, T., Mannuru, N. R., Nie, B., Shimray, S., \u0026amp; Wang, Z. (2023). ChatGPT and a new academic reality: Artificial Intelligence‐written research papers and the ethics of the large language models in scholarly publishing. \u003cem\u003eJournal of the Association for Information Science and Technology\u003c/em\u003e, 74(5), 570-581. https://doi.org/10.1002/asi.24750 \u003c/li\u003e\n\u003cli\u003eMackey, A., \u0026amp; Gass, S.M. (2016). \u003cem\u003eSecond language research methodology and design\u003c/em\u003e (2nd ed.). Rutledge. https://doi.org/10.4324/9781003188414 \u003c/li\u003e\n\u003cli\u003eManning, C. D., \u0026amp; Sch\u0026uuml;tze, H. (1999). \u003cem\u003eFoundations of Statistical Natural Language Processing\u003c/em\u003e. ISBN-13 978-0-521-86571-5. MIT Press.\u003c/li\u003e\n\u003cli\u003eMin, B., Ross, H., Sulem, E., Veyseh, A. P. B., Nguyen, T. H., Sainz, O., \u0026hellip; \u0026amp; Roth, D. (2023). Recent advances in natural language processing via large pre-trained language models: A survey. \u003cem\u003eACM Computing Surveys\u003c/em\u003e, 56(2), 1-40. https://doi.org/10.1145/3605943 \u003c/li\u003e\n\u003cli\u003eMosca, E., Abdalla, M. H. I., Basso, P., Musumeci, M., \u0026amp; Groh, G. (2023, July). Distinguishing Fact from Fiction: A Benchmark Dataset for Identifying Machine-Generated Scientific Papers in the LLM Era. In \u003cem\u003eProceedings of the 3rd Workshop on Trustworthy Natural Language Processing (TrustNLP 2023)\u003c/em\u003e (pp. 190-207). https://doi.org/10.18653/v1/2023.trustnlp-1.17 \u003c/li\u003e\n\u003cli\u003eMu\u0026ntilde;oz, S. A. S., Gayoso, G. G., Huambo, A. C., Tapia, R. D. C., Incaluque, J. L., Aguila, O. E. P., \u0026hellip; \u0026amp; Arias-Gonz\u0026aacute;les, J. L. (2023). Examining the Impacts of ChatGPT on Student Motivation and Engagement. \u003cem\u003eSocial Space\u003c/em\u003e, 23(1), 1-27.\u003c/li\u003e\n\u003cli\u003eNisar, S., \u0026amp; Aslam, M. S. (2023). Is ChatGPT a Good Tool for T\u0026amp;CM Students in Studying Pharmacology? Available at SSRN 4324310. https://doi.org/10.2139/ssrn.4324310 \u003c/li\u003e\n\u003cli\u003eO\u0026rsquo;Connor, S., \u0026amp; ChatGPT. (2023). Open artificial intelligence platforms in nursing education: Tools for academic progress or abuse? \u003cem\u003eNurse Education in Practice\u003c/em\u003e, 66, 103-537. https://doi.org/10.1016/j.nepr.2022.103537 \u003c/li\u003e\n\u003cli\u003ePavlik, J. V. (2023). Collaborating With ChatGPT: Considering the Implications of Generative Artificial Intelligence for Journalism and Media Education. \u003cem\u003eJournalism \u0026amp; Mass Communication Educator\u003c/em\u003e, 0(0)*. https://doi.org/10.1177/10776958221149577 \u003c/li\u003e\n\u003cli\u003eQadir, Junaid (2022): Engineering Education in the Era of ChatGPT: Promise and Pitfalls of Generative AI for Education. TechRxiv. Preprint. https://doi.org/10.36227/techrxiv.21789434.v1. \u003c/li\u003e\n\u003cli\u003eQian, M., Newton, C., \u0026amp; Qian, D. (2021). Cultural Understanding Using In-context Learning and Masked Language Modeling. In \u003cem\u003eHCI International 2021-Late Breaking Papers: Multimodality, eXtended Reality, and Artificial Intelligence: 23rd HCI International Conference, HCII 2021, Virtual Event, July 24\u0026ndash;29, 2021, Proceedings 23\u003c/em\u003e (pp. 500-508). Springer International Publishing. https://doi.org/10.1007/978-3-030-90963-5_38 \u003c/li\u003e\n\u003cli\u003eQin, C., Zhang, A., Zhang, Z., Chen, J., Yasunaga, M., \u0026amp; Yang, D. (2023). Is chatgpt a general-purpose natural language processing task solver?. arXiv preprint arXiv:2302.06476. https://doi.org/10.18653/v1/2023.emnlp-main.85 \u003c/li\u003e\n\u003cli\u003eSalvagno, M., Taccone, F. S., \u0026amp; Gerli, A. G. (2023). Can artificial intelligence help for scientific writing? \u003cem\u003eCritical care\u003c/em\u003e, 27(1), 1-5. https://doi.org/10.1186/s13054-023-04380-2\u003c/li\u003e\n\u003cli\u003eSin, S. (2010). Considerations of quality in phenomenographic research. International Journal of Qualitative Methods, 9(4), 305\u0026ndash; 319. https://doi.org/10.1177/160940691000900401 \u003c/li\u003e\n\u003cli\u003eShidiq, M. (2023, May). The use of artificial intelligence-based chat-gpt and its challenges for the world of education; from the viewpoint of the development of creative writing skills. In \u003cem\u003eProceeding of International Conference on Education, Society and Humanity\u003c/em\u003e (Vol. 1, No. 1), 353-357. \u003c/li\u003e\n\u003cli\u003eStenfors-Hayes, T., Hult, H., \u0026amp; Dahlgren, M. A. (2013). A phenomenographic approach to research in medical education. Medical Education, 47(3), 261\u0026ndash;270. https://doi.org/10.1111/medu.12101 \u003c/li\u003e\n\u003cli\u003eStraume, I., \u0026amp; Anson, C. (2022). Amazement and trepidation: Implications of AI-based natural language production for the teaching of writing. \u003cem\u003eJournal of Academic Writing\u003c/em\u003e, 12(1), 1-9. https://doi.org/10.18552/joaw.v12i1.820 \u003c/li\u003e\n\u003cli\u003eTate, T. P., Doroudi, S., Ritchie, D., Xu, Y., \u0026amp; Uci, m. w. (2023, January 10). Educational Research and AI-Generated Writing: Confronting the Coming Tsunami. https://doi.org/10.35542/osf.io/4mec3 \u003c/li\u003e\n\u003cli\u003eUnser, C. E. (2017). A Study on the positives and negatives of using technology in the classroom. Undergraduate Honors College Theses 2016-. 18. https://digitalcommons.liu.edu/post_honors_theses/18 \u003c/li\u003e\n\u003cli\u003eWalters, W. H., \u0026amp; Wilder, E. I. (2023). Fabrication and errors in the bibliographic citations generated by ChatGPT. \u003cem\u003eScientific Reports\u003c/em\u003e, 13(1), 14045. https://doi.org/10.1038/s41598-023-41032-5 \u003c/li\u003e\n\u003cli\u003eWang, B., Chen, W., Pei, H., Xie, C., Kang, M., Zhang, C., \u0026hellip; \u0026amp; Li, B. (2023). DecodingTrust: A Comprehensive Assessment of Trustworthiness in GPT Models. arXiv preprint arXiv:2306.11698.\u003c/li\u003e\n\u003cli\u003eWeizenbaum, J. (1966). ELIZA\u0026mdash;a computer program for the study of natural language communication between man and machine. \u003cem\u003eCommunications of the ACM\u003c/em\u003e, 9(1), 36-45. https://doi.org/10.1145/365153.365168 \u003c/li\u003e\n\u003cli\u003eWilks, Y. (1978). Making preferences more active. \u003cem\u003eArtificial intelligence, 11(3)\u003c/em\u003e, 197-223. https://doi.org/10.1016/0004-3702(78)90001-2\u003c/li\u003e\n\u003cli\u003eWilliams, C. (2023). Hype, or the future of learning and teaching? 3 Limits to AI\u0026apos;s ability to write student essays. \u003cem\u003eLondon School of Economics internet blog\u003c/em\u003e. https://kar.kent.ac.uk/99505/.\u003c/li\u003e\n\u003cli\u003eWoods, W. A. (1970). Transition network grammars for natural language analysis. \u003cem\u003eCommunications of the ACM, 13(10)\u003c/em\u003e, 591-606. https://doi.org/10.1145/355598.362773\u003c/li\u003e\n\u003cli\u003eZhang, H., Wu, C., Xie, J., Kim, C., \u0026amp; Carroll, J. M. (2023). QualiGPT: GPT as an easy-to-use tool for qualitative coding. \u003cem\u003earXiv preprint arXiv:2310.07061\u003c/em\u003e. \u003c/li\u003e\n\u003cli\u003eZhang, K., \u0026amp; Aslan, A. B. (2021). AI technologies for education: Recent research \u0026amp; future directions. \u003cem\u003eComputers and Education: Artificial Intelligence, 2\u003c/em\u003e, 100025. https://doi.org/10.1016/j.caeai.2021.100025\u003c/li\u003e\n\u003cli\u003eZhao, X. (2022). Leveraging artificial intelligence (AI) technology for English writing: Introducing Wordtune as a digital writing assistant for EFL writers. \u003cem\u003eRELC Journal\u003c/em\u003e. https://doi.org/10.1177/00336882221094089 \u003c/li\u003e\n\u003cli\u003eZhou, X., Chen, Z., Jin, X. and Wang, W.Y. (2021), \u0026ldquo;HULK: an energy efficiency benchmark platform for responsible natural language processing\u0026rdquo;, \u003cem\u003eProceedings of the Conference of the European Chapter of the Association for Computational Linguistics: System Demonstrations, Vol. 16\u003c/em\u003e, pp. 329-336. https://doi.org/10.18653/v1/2021.eacl-demos.39\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"AI, LLMs, GPT, Academic Writing, Teaching English","lastPublishedDoi":"10.21203/rs.3.rs-5534554/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5534554/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThis study investigates the integration of Generative Pre-trained Transformer (GPT) models in scholarly writing instruction, aiming to understand their perceived positive and negative impacts on academic writing. The research employs a phenomenographic design to study the practical implications of such models in real classrooms. The course design involves five sessions, each focusing on specific sections of a research paper, with activities using GPT integrated into classes and home projects. The methodology comprehensively investigates the impact of GPT-4 on scholarly writing, with a phenomenographic perspective towards participant selection, data collection, and research design. The study reveals the multifaceted ways in which GPT models contribute to the productivity and efficiency of writing research papers, highlighting both the advantages and challenges associated with their use. The findings underscore the transformative potential of AI in education, emphasizing the need for educators and researchers to understand and maximize the benefits of GPT models.\u003c/p\u003e","manuscriptTitle":"LLMs and Academic Writing in Practice: Exploring Participants’ Utilization of GPT during an AI-Assisted Course on Writing Research Papers","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-12-25 15:42:06","doi":"10.21203/rs.3.rs-5534554/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"530c61e0-3c7b-461e-aafd-693379dc6cf5","owner":[],"postedDate":"December 25th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-05-15T13:38:50+00:00","versionOfRecord":[],"versionCreatedAt":"2024-12-25 15:42:06","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-5534554","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5534554","identity":"rs-5534554","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}