Advancing Students’ Academic Excellence in Distance Education: Exploring the Potential of Generative AI Integration to Improve Academic Writing Skills

Perceptions of Stakeholders on Generative AI

Recent technological developments have seen an increase in integrating generative AI tools like ChatGPT, Pro Writing Aid, word processors, and Grammarly into teaching and learning environments, with many lecturers supporting this integration within university contexts (Koka et al., 2023). These tools are perceived as diagnostic instruments aiding in error identification, structural improvements, and enhancing various linguistic features of writing, connecting with the notion that generative AI promotes student responsibility in learning by providing insights and suggestions for improvement (Don, 2021; Hibert, 2019; Palermo & Wilson, 2020). However, alongside the positive impacts, scholars such as Aljohani (2021), Khalil and Er (2023), and Sevnarayan and Maphoto (2024) have raised concerns regarding generative AI’s potential role in plagiarism. These concerns have prompted proposed measures to reduce plagiarism risks, including altering assessment methods and establishing usage rules for generative AI tools (Alam & Mohanty, 2022; Lukowicz, 2023; Nguyen, 2023). Despite some hesitations, recent studies by Kiryakova and Angelova (2023), suggest a positive attitude among university lecturers toward integrating generative AI, particularly ChatGPT, into teaching practices. However, concerns persist regarding ethical usage, potential fraud, plagiarism, and privacy issues (Adiguzel et al., 2023; AIAfnan et al., 2023; Tao et al., 2019). Moreover, there is a noted gap in digital competencies among students and lecturers, which poses challenges to the effective and ethical integration of generative AI tools in HEIs (Ausat et al., 2023; Kohnke et al., 2023). Bozkurt (2024) provides a thorough analysis of the implications of integrating generative AI tools in academia and emphasises the necessity of transparent reporting for maintaining academic integrity. Bozkurt (2024) suggests practical solutions to address ethical concerns and highlights the responsibility of human authors for content integrity. Further sentiments by Bozkurt (2024) echo the importance of human oversight alongside AI assistance to uphold scholarly rigor and ethical standards.

Students’ Perceptions and Attitudes

Understanding students’ perceptions of generative AI’s role in academic writing is crucial amidst the transforming educational context. Students exhibit a range of perspectives, from optimism regarding generative AI’s potential to enhance the writing process to reservations about over-reliance and concerns regarding creativity and authenticity (Cotton et al., 2023; Dwivedi et al., 2023). These perceptions reflect the complexities in the relationship between students and generative AI, influencing their acceptance or resistance to its integration into academic writing (Chan & Hu, 2023). While students appreciate efficiency improvements and personalised learning facilitated by generative AI, they also express concerns about learning curves, data privacy, and depersonalisation (Sallam, 2023). Notably, generative AI integration has shown significant developments in writing quality, grammar, vocabulary, plagiarism reduction, creating a conducive learning context, and enhancing engagement (Zulfa et al., 2023). However, concerns persist regarding plagiarism detection and the potential impact on students’ writing authenticity (Khabib, 2022; Fitria, 2023).

Motivational Potential and Learning Enhancement

Generative AI’s integration in education provides efficiency, quality, and learning opportunities, transforming academic writing into a continuous learning process (Chang et al., 2023; Jeon, 2022). It creates personalised feedback, enhances iterative improvement, and motivates students through gamification strategies, enhancing engagement and intrinsic motivation (Bennani & Maalel, 2022; Dichey et al., 2014). Through recognising achievements and providing clear pathways for skill advancement, generative AI empowers students to enhance their writing capabilities and develop a positive attitude toward academic challenges (Alam, 2021; Chassignol et al., 2018). Moreover, generative AI tools assist in plagiarism detection and guide students in adapting their writing styles, further enhancing learning outcomes (Khabib, 2022; Fitria, 2023).

While the literature indicates various benefits of generative AI integration in academic writing modules, it is critical to address the biases and inconsistencies present in existing studies. These include concerns regarding plagiarism, ethical usage, privacy issues, and the need for digital competencies among stakeholders. In addition, further research should focus on exploring students’ perceptions in more depth and addressing their concerns about authenticity, creativity, and data privacy. Moreover, future studies should explore the long-term effects of generative AI integration on academic writing skill development and student learning outcomes in diverse educational contexts.

Research Context

This study centres on online distance learning modules offered by a South African open distance e-learning (ODeL) university, with an annual enrolment of around 500,000 students from 132 countries, including Nigeria, Namibia, Zimbabwe, India, Congo, Ethiopia, the United States of America, and China. The student body is diverse, representing various financial, linguistic, and social backgrounds, with ages ranging from 18 to 70. The Academic Writing (WRI124) module, situated within the Department of English Studies, is an undergraduate program that focuses on enhancing students’ proficiency in academic English, critical reading and writing skills. The module aims to cultivate students’ capacity to engage with academic genres across diverse disciplines and contexts. This includes honing skills in argumentative and discursive essay writing, crafting research-based essays and articles, understanding conventions for citation, and developing research skills. WRI124 caters to a diverse student body, comprising both native English speakers and those who speak English as an additional language. WRI124 is a fully online module, and all assessments are conducted online via the Moodle learning management system (LMS). Classes are conducted through Microsoft Teams, and all activities are posted on the LMS. Most students pursue part-time studies while working full-time, and they come from different ethnicities, including Black, White, Coloured, Indian, and Asian, predominantly from middle- to low-income families. Many students, residing in remote areas with limited internet access, complete assessments through cell phones or local internet cafes.

In this context, this investigation transpired in the initial phase of ChatGPT’s introduction. The researchers observed an increase in the utilisation of ChatGPT within assignments throughout the first semester of 2023. This resulted in a palpable sense of helplessness among lecturers, prompting uncertainty about how to effectively address this emergent issue. The deployment of generative AI exhibited a discernible degree of disorder, as students excessively relied on its functionality to aid them in online assessments. It is noteworthy to highlight that while students demonstrated adeptness in using this technology, a conspicuous digital divide emerged wherein some lecturers either struggled to employ it or failed to comprehend the rationale behind students’ pervasive usage. This disjunction captivated the researchers, motivating a comprehensive exploration into this particular dimension within the integration of generative AI in the WRI124 module. The generative AI system demonstrated a remarkable capacity to address virtually all questions posed on assessments by lecturers. Consequently, students adeptly manipulated this tool to their advantage, particularly in online assessments. It is imperative to highlight that lecturers refrained from utilising generative AI as a pedagogical tool; thus, this study does not constitute a pilot study. The principal aim was to meticulously examine the prospect of using generative AI to ascertain its potential to augment the development of students’ academic writing skills.

Research Method

The qualitative design adopted in this study is not merely incidental but is deemed paramount. It was selected to address the critical and multifaceted dimensions inherent in the integration of generative AI within the WRI124 in a DE context. This qualitative study adopts an exploratory approach, signifying its open-ended nature conducive to the understanding of in-depth insights (Griffin, 2016). Furthermore, it aligns itself with an interpretive study, denoting the researchers’ commitment to analysing and interpreting the meaning inherent in the collected data within its contextual framework (Walsham, 2006). The employment of discussions and interviews stands as the principal methodological choice, strategically employed to explore phenomena. This method aims to reveal rich insights from a mega-enrolment module (Ponelis, 2015) to understand if generative AI can be pedagogically used in Academic Writing modules to enhance students’ academic writing skills.

Research Design

This research study embraces a phenomenological approach to understand the essence of the phenomenon of generative AI under exploration as it is experienced by the participants. According to Wright-St Clair (2014) and Hall et al. (2016), this methodological choice enabled the researchers to understand the dynamics between students, lecturers, and markers and the potential integration of generative AI within the WRI124 module. Phenomenology, with its roots embedded in the exploration of lived experiences, positions itself as a powerful lens through which to discern the deep layers of participants’ experiences enhanced by generative AI. Through a phenomenological research design, the study aims to comprehend experiences, providing a depth of insight that surpasses mere observations.

Data Collection Instruments

Utilising a qualitative methodology, this research embarks on an exploration of the manifold approaches and applications of generative AI to enhance the pedagogical context of teaching and learning experiences (Ponelis, 2015). To address the three research questions, a triangulation of research instruments was strategically deployed, utilising the strengths of each to provide a comprehensive understanding of the potential influence between generative AI and academic writing.

The three chosen research instruments; e-mail interviews, focus group discussions (FGDs), and a WhatsApp group discussion were deliberately selected to facilitate a multifaceted exploration (Asiamah et al., 2017; Campbell et al., 2020; Hennik, 2013; Tasker & Cisneroz, 2019). E-mail interviews, designed in alignment with the framework proposed by Hunt and McHale (2007), served as an invaluable tool for asynchronous communication with lecturers. This method facilitated the systematic gathering of detailed information on technical reliability, multimedia integration, engagement strategies, communication dynamics, and collaborative aspects of incorporating generative AI within the WRI124 module.

To augment and enrich the qualitative data, an FGD and a WhatsApp group discussion were concurrently employed. These methods, under the approaches delineated by Tasker and Cisneroz (2019), enabled an in-depth exploration of the generative AI tool’s impact on academic writing skills. The deliberate choice of multiple instruments, coupled with the concurrent use of diverse platforms, not only ensured methodological triangulation but also added layers of depth and richness to the gathered data. Challenges encountered, particularly delayed responses in the asynchronous email interviews, were methodically addressed. Structured schedules were implemented, and trained observers were enlisted to enhance the efficiency of the communication process with lecturers (Dahlin, 2021). Data availability is restricted in accordance with the confidentiality guidelines established by the university’s ethics approval committee.

Population and Sampling

The study comprises a population of approximately 14,000 students enrolled in the WRI124 module. Twenty (20) lecturers who teach first-year students from the Department of English Studies were sent email interviews and were labelled ‘Lecturer 1’, ‘Lecturer 2’, ‘Lecturer 3’, ‘Lecturer 4’, and so on for anonymity (Griffin, 2016). 20 student participants were invited to take part in an FGD through the module Telegram group and were named ‘Student 1’, ‘Student 2’, ‘Student 3’, and so on. In addressing the third research question, the researchers engaged thirty (30) markers to contribute their insights on a WhatsApp group regarding how generative AI can motivate students to enhance their academic writing skills (Tasker & Cisneroz, 2019). Markers were labelled ‘Marker 1’, ‘Marker 2’, ‘Marker 3’, and so on. The purposive sampling method that was used to select lecturers, students, and marker participants is justified in a qualitative study to ensure a focused and in-depth exploration of perceptions within the target population (Campbell et al., 2020).

Research Procedure

The research adopted a multifaceted approach to explore stakeholders’ perceptions of integrating generative AI into academic writing practices at an ODeL university. Data was collected during the second semester of 2023 (November and December). To answer the first research question on lecturers’ perceptions of generative AI integration in academic writing, 20 lecturers were sent email interview questions. Of the 20 emails sent to lecturers, only 12 responded. Responses from lecturers were collected over two weeks. To answer the second research question on student perceptions, students were invited to an FGD through the module Telegram group. Of the 20 students who accepted to be part of the FGD, only 12 joined the Microsoft Teams discussion, which lasted about 45 minutes. To answer the third research question, questions were posed to markers in the module WhatsApp group. The group consisted of 30 markers; however, only 10 markers participated in the discussion. The researchers coded the data, and it was organised into the following themes:

1. Lecturers’ perceptions of generative AI use in the WRI124 module.

2. Students’ perceptions of the use of generative AI as a potential to guide them through academic writing.

3. The potential of generative AI to motivate students to improve their academic writing.

Validity and Reliability Measures

The construct validity of this study is integrated with the theoretical frameworks, which ensures a coherent discussion. This integration is complemented by an alignment with prior research studies, which positions the research within a broader academic context and strengthens the study’s theoretical underpinnings.

To enhance content validity, the research employs a multifaceted approach by incorporating diverse data collection instruments, including e-mail interviews with lecturers, an FGD with students, and a WhatsApp group discussion with markers. This choice aims to explore stakeholders’ perceptions regarding the potential of generative AI for enhancing academic writing skills. Through the employment of triangulation data collection methods, the study not only widens the scope of the explorations but also strengthens the credibility and richness of the gathered insights.

The validity of this research is highlighted by the creation and rigorous review of research questions. These questions, designed with precision, underwent a rigorous validation process led by the researchers’ adeptness in educational technology and qualitative research within the team. This approach ensures that the research questions resonate with the core objectives of the study and enhances the trustworthiness and relevance of the study.

In terms of reliability, the study adheres to stringent measures to instill confidence in the consistency and dependability of the findings. Inter-rater reliability is rigorously maintained during email interviews, emphasising uniformity in the interpretation of responses. Consistent purposive sampling across diverse stakeholders further contributes to the study’s reliability, ensuring a representative and varied perspective. Moreover, standardised data collection procedures are systematically implemented, minimising variability and enhancing the reproducibility of the study’s outcomes.

Lecturers’ Perceptions of Generative AI Use

Students’ Perceptions of the Use of Generative AI as a Potential to Guide Them through Academic Writing

The Potential of Generative AI to Motivate Students to Improve Their Academic Writing

What are lecturers’ perceptions of using generative AI to teach students academic writing skills?

In exploring lecturers’ perspectives on the integration of generative AI in the WRI124 module, an interesting divergence emerges and captures a multifaceted dialogue among lecturers. Amidst the prevalent concerns about potential misuse and the temptation of shortcuts for skill acquisition, lecturers also recognise and appreciate the inherent benefits of generative AI tools (Bansal et al., 2019; Glikson & Wooley, 2020; Lukowicz, 2023). This recognition highlights the transformative nature of the discourse and acknowledges the dual facets of generative AI as both a potential challenge and an asset in the academic writing context. The human-AI collaboration framework, a pivotal aspect of this discussion, not only emphasises the need for trust in generative AI adoption but also aligns seamlessly with the persistent call from lecturers for responsible use (Bozkurt, 2024). It positions generative AI not as a mere quick fix but as an integral learning tool, a stance echoing the principles of SCT. Within the context of Vygotsky’s ZPD, lecturers stress the responsibility of students to actively engage with generative AI as a scaffold, a support system requiring their active participation in the learning process (Allal & Ducrey, 2000; Shabani et al., 2010). This relationship between trust, responsibility, and active involvement shapes a holistic understanding of generative AI’s role in academic writing skill development. Furthermore, the integration approach advocated by lecturers aligns not only with the principles of SCT but also resonates with the concept of mixed-initiative decision-making (Bansal et al., 2019). This approach acknowledges generative AI as a supplementary tool, complementary to essential human effort rather than a replacement (Adiguzel et al., 2023; Tao et al., 2019). Lecturers emphasise the necessity of balance, encouraging generative AI utilisation for correction and comparison while safeguarding the essence of independent student writing (Iqbal et al., 2022; Seeber et al., 2020). This perspective emphasises a balance between technology and human agency and suggests an urgent need for pedagogical strategies that are tailored to address this complexity in generative AI integration within academic writing curricula.

How do students perceive the use of generative AI as a potential guide through the academic writing process?

Students’ perspectives on the utilisation of generative AI in the WRI124 module form diverse views which are influenced by skepticism, hope, and a keen awareness of authenticity concerns and echoes the cautionary sentiments expressed by researchers regarding the potential risks of overreliance that may compromise the authenticity of students’ work (Cotton et al., 2023; Dwivedi et al., 2023; Qiao & Zhao, 2023; Woithe & Filipec, 2023). These complex perspectives reflect the discourse surrounding the integration of generative AI in academic writing, where students grapple with the dual-edged nature of technological assistance. Amidst the prevalent fears of generative AI potentially altering individual writing styles, a subset of students emerges, expressing a sense of optimism about the technology’s potential to assist in overcoming specific writing challenges. This optimistic outlook resonates with scholarly discourse portraying generative AI as a supplementary tool, particularly effective in addressing issues related to grammar and structural elements in writing (Alam & Mohanty, 2022; Chan & Hu, 2023). It suggests an acceptance among students of generative AI’s role as a facilitator rather than a disruptor in the writing process. SCT provides a lens through which to understand the dynamics of students’ perceptions. Acknowledging the significant influence of students’ peer interactions, and cultural contexts, SCT provided in-depth insights into the multifaceted shaping of students’ views on generative AI. The ZPD concept within SCT notes the importance of steering clear of excessive dependence on generative AI to encourage a more self-directed learning approach (Allal & Ducrey, 2000; Chang et al., 2023; Jeon, 2022; Shabani et al., 2010). The alignment between findings and the SCT lens, coupled with insights from the human-AI collaboration framework, indicates the balance required in the collaborative relationship between students and generative AI (Jain, Garg & Khera, 2023; Lukowicz, 2023).

How can generative AI motivate students to improve their academic writing skills?

The discussions among markers concerning the integration of generative AI into academic writing provide a multifaceted view that mirrors sentiments expressed by students. While markers exhibit optimism about the strategic use of generative AI as a motivator and efficiency tool (Alam & Mohanty, 2022; Lukowicz, 2023; Zulfa et al., 2023), a critical question is: How can generative AI motivate students to improve their academic writing skills? A thorough examination reveals certain aspects that warrant careful consideration. Markers acknowledge generative AI as a significant educational tool, consistent with the SCT perspective, which positions it as scaffolding for students in academic writing development (Chang & Sun, 2009). However, the cautionary note introduced by some markers, particularly regarding the risk of verbatim copying and the emphasis on the importance of interpretation (Fitria, 2023; Khabib, 2022), introduces a counterargument. This aligns with broader concerns raised in the literature about the potential drawbacks of generative AI, including the need for students to develop critical thinking and interpretation skills (Bozkurt, 2024; Zulfa et al., 2023). While markers express optimism regarding generative AI’s role as an efficiency tool, their wary stance introduces a potential contradiction. The literature supports the idea that generative AI can enhance creativity and support learning, but the need for vigilance against verbatim copying highlights a potential tension point (Alam & Mohanty, 2022; Zulfa et al., 2023). The shared views on responsible generative AI use and motivation are in sync with the broader discourse, emphasising the importance of guiding students to use generative AI meaningfully (Alam & Mohanty, 2022; Zulfa et al., 2023). However, the potential risks associated with irresponsible use raise questions about how effectively current pedagogical strategies address these concerns.

What are lecturers, students, and markers’ perceptions towards the use of generative AI on academic writing skills in the context of ODeL?

In addressing the overarching research question on the perceptions of lecturers, students, and markers towards generative AI use in ODeL, the findings portray diverse perspectives. Lecturers recognise the potential benefits of generative AI and advocate for its integration as a supplementary learning tool within a pedagogical framework, emphasising the importance of responsible use and avoiding shortcuts in skill acquisition (Lukowicz, 2023). This aligns with their perspective on maintaining a balanced approach and encourages generative AI use for correction and comparison while preserving the essential human effort in independent student writing (Iqbal et al., 2022; Seeber et al., 2020). Students, on the other hand, exhibit diverse attitudes influenced by peer interactions, cultural contexts, and concerns about authenticity. The need for tailored approaches is evident to accommodate varied learning preferences and underlines the balance required in the collaborative relationship between students and generative AI (Chang et al., 2023; Jeon, 2022; Naidu & Sevnarayan, 2023; Shabani et al., 2010). Markers’ perspectives align with both lecturers and students and acknowledge the potential benefits of generative AI as a motivator and efficiency tool while expressing concerns about misuse and verbatim copying. The shared emphasis on responsible use and motivation resonates with broader discussions on the importance of teaching students how to use generative AI thoughtfully and supports the ongoing discourse surrounding the careful integration of generative AI into education (Alam & Mohanty, 2022; Zulfa et al., 2023). This analysis highlights the need for a critical and context-specific approach to generative AI’s integration in ODeL and emphasises the crucial elements of trust, responsible use, and continuous evaluation to optimise the positive impact on academic writing skills.