By Yousef Alafghani
Abstract
This paper examines the general perceptions of Grammarly and the Writing Center among students at the American University of Sharjah (AUS). Although other Artificial Intelligence (AI) software such as ChatGPT seem to dominate the discourse surrounding AI use in both the academic writing scene and the Writing Center, Grammarly is also an important tool in that conversation precisely because people assume that it is a benign alternative to other forms of AI. Users have a tendency to heavily rely on the program, and its feedback, despite its flaws, is taken for granted. Therefore, I make the hypothesis that students will have a much more negative perception of the Writing Center than they will with Grammarly. In order to examine the accuracy of this hypothesis, a study was conducted via an online-administered survey and an interview with a writing center tutor at AUS. Results from the survey revealed that students are aware of the faults behind Grammarly’s feedback, and have a more positive attitude towards the Writing Center. The interview with the Writing Center tutor delineated that Grammarly is not particularly perceived as a threat by other tutors at the Writing Center, but rather a tool that must be used thoughtfully. Further research on Grammarly and the Writing Center would aid in advancing the discourse surrounding the use of non-generative AI (GAI) in various academic contexts.
Keywords: Artificial Intelligence, Grammarly, Writing Center, American University of Sharjah, Academic Writing.
Introduction
Artificial intelligence (AI) has become a dominant tool in the academic writing scene, particularly with the growing influx of university students relying on generative AI (GAI) software for developing their writing, namely ChatGPT. Though there is a general tendency to synonymize both AI and GAI as interchangeable terms, their distinctions are notable. While AI refers to a set of technological processes utilized for performing tasks analogous to human intelligence, GAI is a subset of AI that is trained to learn patterns and structures from the data it is fed and produces “text, images, videos, or other forms of data” similar to that data (“Generative artificial intelligence,” 2025). GAI is a fairly recent addition to the conversation surrounding the moral and ethical issues of technology within academic writing. As technological advancements in GAI continue to grow, there has been a shift in the general perception of writing and the way in which it is being engaged with. Grammarly—an AI-powered application designed to identify errors in grammar, punctuation, sentence structure, word choice, and tone—was a prominently used program by university students before the advent of ChatGPT, mostly catering towards English-as-a-second-language (ESL) students. However, Grammarly has shown to be fundamentally fractured in its stylistic suggestions and its corrections are often misleading for its users.
Beyond these AI-powered programs, many students have the option of obtaining peer-to-peer services for written assignments. The writing center allows students to receive peer assistance and support for developing their writing skills, giving students the necessary tools to strengthen their writing proficiency. However, there seems to be a growing disregard for writing centers across various academic institutions, which either stems from a blatant negligence of such services or a reluctance to receive help from other peers out of fear of judgement or inadequacy. As a result, students gear towards AI software such as Grammarly to receive quick and automated feedback on their writing. However, the students’ sound and creativity in an assignment is just as significant as the content itself, and Grammarly may jeopardize the presence of that voice, risking the appearance of misplaced grammatical choices in their writing. Although ChatGPT has proven to be a substantially severe case of GAI misuse in academic writing, Grammarly is more passive in its use of AI, which is primarily non-generative, and is not given the same amount of caution by writing instructors and tutors as GAI programs are.
When students decide to visit the writing center with their assignments already modified by Grammarly, it can affect the ways in which tutors assist their peers and the appropriate measures they take in providing support for students to improve their writing. On top of that, Grammarly may impede the students’ ability to develop their writing skills in the long term. Based on these issues, the following research question is raised: What are the implications of Grammarly on university students’ ability to write effectively within various academic contexts? As this question is being answered, my paper will examine the issues within Grammarly’s automated feedback system, the history of AI discourse in the writing center, attitudes towards Grammarly and the writing center, and the effectiveness of Grammarly in relation to writing center services.
Literature Review
In order to understand the implications of Grammarly use among university students and what it means for attitudes regarding the writing center, it is important to observe the program that Grammarly uses for its automated feedback. Ghufron and Rosyida (2018) identify that Grammarly is effective in dealing with issues of word choice, grammar, spelling, and punctuation, but has difficulties in improving content-related and organizational issues (p. 401). What this outlining of the strengths and weaknesses of Grammarly means is that the software is imperfect, but not completely unreliable. Grammarly has shown to facilitate a basic level of assistance and learning for individuals who struggle to develop their writing skills. However, Ghufron and Rosyida do not take into account the over-reliance of students on Grammarly as a factor in curbing their ability to write effectively and independently. In fact, “The overreliance on [automated written corrective feedback (AWCF)] indicated more limited cognitive engagement which led to feedback’s blind acceptance” (Koltovskaia, 2020, abstract). This issue of “blind acceptance” becomes a considerable problem when Grammarly is being integrated into educational spaces, especially when students who may not have extensive experience in writing blindly accept Grammarly’s feedback.
With that being said, Grammarly does not necessarily represent a unique case of AI use in academic writing nor is it a newfound issue at all. As a matter of fact, Bryan (2024) identifies that “software discourse in writing center studies” stretches all the way back to 1978, and that “these earlier conversations offer current practitioners and researchers considerable opportunity to assess the present moment within the larger history of writing center discourse” (p. 23). By examining that history, writing center tutors are able to acknowledge the concept of the writing center as an on-going, developing project for the academic writing scene, while also allowing them to tackle issues of AI usage strategically and effectively. Additionally, an understanding of the history behind AI discourse (particularly Grammarly) will allow for a deeper evaluation of the more fundamental issues regarding different attitudes towards AI software and the writing center.
Looking directly at the clashes between Grammarly and the writing center reveals the possibilities and conditions of AI being integrated into various academic spaces, raising questions regarding the balancing of face-to-face tutoring and Grammarly usage. Zhang, Özer, and Bayazeed (2020) have examined the use of Grammarly in contrast to face-to-face tutoring at the writing center, revealing that participants used Grammarly more than they visited the writing center, but have used face-to-face tutoring for a “wider variety of purposes compared to Grammarly” (p. 3). It seems that Grammarly users, on a general level, have an awareness of the flaws and imperfections of the application’s corrections and feedback. In fact, students who use Grammarly and visit the writing center use both tools for “different needs in different contexts” (Zhang, Özer & Bayazeed, 2020, p. 3). Therefore, people are, to a large extent, capable of using AI-powered software cautiously and effectively within academic contexts. However, despite the fact that students have the capabilities of balancing AI software such as Grammarly with real life, face-to-face assistance with their writing, people have also been shown to be extremely reliant on such software, affecting them both on an academic and a cognitive level (Koltovskaia, 2020, discussion and conclusion, para. 2). It is important to note that the effect of AI on students’ academic and cognitive performance comes from this over-reliance on such technology, suggesting that Grammarly can still be integrated into academic writing if students are taught to use it efficiently.
There is also the issue of Grammarly as an inadequate substitute for face-to-face tutoring. Dembsey (2017) suggests that Grammarly is unable to give the “same type of support possible in a writing center consultation, such as agency, praise, individualized feedback, and assistance on issues beyond individual sentences” (p. 89). AI software such as Grammarly, beyond the errors in its feedback, lack the human qualities behind peer-tutoring that make the tutor-tutee dynamic intrinsically complex. In fact, even within its automated feedback, “[Grammarly’s] comments on grammar were too technical and inaccurate to promote learning in students from a wide range of backgrounds” (Dembsey, 2017, p. 89). Based on these findings, what Grammarly appears to lack the most is adaptation and context. Unlike the relationship between the tutor and tutee in the Writing Center—where tutors need to adapt and conform to the tutee’s needs in order to assist them effectively—Grammarly goes by a set of rules based on the algorithm it is built upon (hence the term “automated feedback”). Therefore, a strict use of Grammarly on its own deprives students of the nuanced and applied assistance they would otherwise be acquiring from other face-to-face tutoring programs.
Overall, based on the research reviewed, the conversation seems to veer towards the implications of Grammarly on the perceptions of writing center services and academic writing as a whole. Additionally, researchers have raised concerns regarding the actual effectiveness of Grammarly in giving correct and adaptable feedback to its users. Although there is research that directly discusses concerns with Grammarly in relation to the writing center (Zhang, Özer & Bayazeed, 2020. Dembsey, 2017), the existing literature on this topic is not sufficient enough to fully and holistically address the issues of Grammarly within a writing center and academic writing context. Therefore, my paper aims to further build upon the conversation surrounding the topic of Grammarly and the Writing Center, enriching and developing upon the discourse surrounding this issue. I plan to apply my knowledge on this topic—based on the literature I have reviewed—within the context of the American University of Sharjah (AUS). In doing so, I will provide a layered discussion surrounding AI software use in academic writing among university students at AUS, addressing some of the more specific issues of Grammarly use (and AI use as a whole) within the institution.
Methodology
Data was collected over a period of two weeks in October 2024 at AUS. A mix of quantitative and qualitative data was collected through two primary research methods: an online survey administered to all AUS students and an audio-recorded interview with a writing center tutor.
The Survey
An English-language online survey consisting of multiple-choice and written questions (see Appendix A) was administered and made available to all AUS students through university-oriented group chats on WhatsApp. The survey includes 30 questions separated into the following four sections:
- Demographic questions (age, gender, university standing, etc.),
- Questions about Grammarly use,
- Questions about AUS’s Writing Center,
- Two questions asking participants to compare their experiences with both Grammarly and the Writing Center.
In regards to Grammarly, the survey collected information about students’ frequency of Grammarly use, possible reasons as to why they would use it, and their general perceptions on Grammarly’s automated feedback. As for the Writing Center, the survey featured questions regarding frequency of participants’ visits to the Writing Center, their motivation for going there, and their attitudes towards the service.
A total of 43 individuals (20% male, 77.5% female, and 2.5% unspecified) participated in the survey. The majority of the participants (92.5%) were in the 18–23-year age range. Most participants were sophomores (47.5%), whereas the rest were either first-year students, juniors, seniors, or graduates.
The Interview
I conducted a semi-structured interview with a writing center tutor, Soha Abdalgawad, in the Writing Center hall. I asked Abdelgawad six questions (see Appendix B) about her encounter with Grammarly usage among her tutees, her general perceptions on the application, and her strategy of assisting students who depend on Grammarly’s corrections and feedback.
Findings
Survey Results
When asked if they were familiar with Grammarly, all 40 respondents indicated that they were, indeed, aware of the application. Thirty-six of the forty participants have used Grammarly in the past, with 63.9% of those 36 participants stating that they still use Grammarly. Additionally, 67.5% of participants reported having previously visited the Writing Center at AUS. However, 57.1% of the individuals who have not been to the Writing Center indicated that they would consider going there in the future. Among the other questions on Grammarly and the Writing Center (see Appendix A), the most important are as follows:
When asked, “On a scale from 1-5, how much do you trust Grammarly’s feedback?” (see Appendix A, part 2, question 8), 15.8% of the participants selected ‘2,’ 50% selected ‘3,’ 28% selected ‘4,’ and only 5.3% selected ‘5.’ None of the participants selected ‘1.’ In another question, when asked, “what aspects of your writing does Grammarly help you with?” (see Appendix A, part 2, question 10), most participants (89.5%) selected “grammar and punctuation.” Notably, 60.3% of participants used it for clarifying spelling mistakes, 44.7% used it for help in word choice, and 47.4% used it for sentence structure. A few respondents also used Grammarly for either paragraph structuring, style and tone, or other reasons.
Furthermore, when asked about the Writing Center’s effectiveness in improving participants’ writing (Appendix A, part 3, question 4), 48.7% of individuals thought the service was “extremely helpful.” 38.8% thought it was either “moderately helpful” or “extremely helpful.” Only 6.5% of participants selected “not helpful at all,” suggesting that the general consensus regarding the Writing Center is mostly positive. The participants’ positive attitudes towards the Writing Center was also evident when participants were asked if they encountered any limitations with the service (see Appendix A, part 3, question 6). An overwhelming 90% did not find any issues with the Writing Center, and only 6.7% encountered problems. In contrast, when asked the same question regarding Grammarly (see Appendix A, part 2, question 11), 55.3% found that there were issues and limitations with the program, and 47.4% did not encounter any issues with Grammarly.
Finally, when participants were asked which of the two services (Grammarly and the Writing Center) they found more helpful in their writing (see Appendix A, part 4, question 1), 48.8% thought the Writing Center was more helpful, with only 9.3% finding Grammarly to be more helpful for developing their writing proficiency. 25.6% believed both services were equally as helpful, and 16.3% were not sure which of the two services they favored.
Interview Results
An interview was conducted with Writing Center tutor Soha Abdalgawad. I asked Abdelgawad six questions pertaining to Grammarly usage in the Writing Center (see Appendix B), with the conversation lasting around fifteen minutes. The most important questions and answers are as follows:
When asked if Grammarly helps or hinders a student’s ability to develop their own writing skills (see Appendix B, question 1), Abdalgawad suggests that it “depends on who uses the program and how they and their knowledge of writing will help them use Grammarly better” (S. Abdalgawad, personal communication, November 18, 2024). If a student has “weak knowledge in writing,” then the program will not be used by them efficiently, and will only see it as an editing tool. For instance, Abdalgawad believes that if a student pastes their text into Grammarly, and the program picks up on multiple issues relating to commas, students may accept these edits without looking into what the feedback actually suggests and may not learn from these mistakes. Therefore, students will over-rely on Grammarly to fix their issues without taking the initiatives to develop their writing skills on their own.
Furthermore, when asked if AI software such as Grammarly influenced students’ expectations of how a tutorial session should be conducted (see Appendix B, question 3), Abdalgawad suggests that students, in her experience, come to Writing Center sessions for the “tutoring rather than fixing,” and that “they really want to learn” (S. Abdalgawad, personal communication, November 18, 2024). In fact, Abdalgawad identified that people who use Grammarly, when they see an excess amount of suggestions and underlines across their text, tend to feel overwhelmed and less confident about their writing. Thus, they accept the feedback without actually double-checking the edits Grammarly makes to their text. However, when students come to the Writing Center and see that they have the same issues with their paper, they tend to feel more compelled to develop their writing skills in the future, acknowledging the fact that these errors should be taken seriously. Therefore, Abdalgawad believes that the issue of students’ perception of the Writing Center may not solely originate from their overreliance on Grammarly, but rather from their preconceived notions or misunderstandings of how a tutoring session is typically managed.
Finally, when asked if Grammarly should be taken with the same caution as ChatGPT in regards to the future of the Writing Center (see Appendix B, question 6), Abdalgawad suggests that Grammarly is a better alternative than ChatGPT “simply because ChatGPT can write for you” (S. Abdalgawad, personal communication, November 18, 2024). Abdalgawad perceives Grammarly to be more “interactive” than ChatGPT since the program gives explanations behind the feedback it gives. She identifies that since many tutees use Grammarly instead of ChatGPT, there is more value to be extrapolated from the former. Overall, Abdalgawad suggests that there students at least have the capacity to be in a proactive position when using Grammarly since its feedback can either be taken for granted or used efficiently. Therefore, it seems that educators are more incentivized to either retaliate against ChatGPT or accentuate the discretion needed to use it responsibly because of a supposed absence of interactivity. It is precisely because of the interactive nature of Grammarly (and the contention surrounding ChatGPT) that Abdalgawad believes that “ChatGPT will die first before Grammarly.”
Discussion
The Survey
The overall results of the survey revealed a more positive attitude towards the Writing Center than Grammarly, with most participants being sceptical of the feedback and suggestions the program gives for their written texts.
In terms of Grammarly’s program, participants appear to find it more effective in aiding them with grammar, punctuation, spelling, sentence structure, and word choice (see Appendix A, part 2, question 10). In comparison, Ghufron and Rosyida (2018) identify that Grammarly works best with word choice, grammar, spelling and punctuation. Since these textual issues are minor in relation to the structural and contextual problems of a given text, students may easily overlook smaller issues such as grammar and punctuation. Therefore, Grammarly may aid in assisting students to identify technical problems with their text through a proofreading and copyediting process.
However, as I have discussed in the literature review, there is an issue of Grammarly being over-relied upon by its users, particularly the issue of blind acceptance brought up by Koltovskaia (2020). However, survey results seem to highlight a level of indifference towards Grammarly’s feedback. In fact, when participants were asked about their level of trust towards Grammarly’s feedback from a scale of 1–5 (see Appendix A, part 2, question 8), 50% of respondents selected ‘3.’ Again, these results may suggest an openness by students to use Grammarly for their writing, but are aware of the mistakes it can make in its feedback, and are thus conscious of the suggestions they accept.
With that being said, survey results also show that Grammarly lacks the human assistance a student would typically receive from the Writing Center. When asked to elaborate on their answer for which service is more helpful in improving their writing skills (see Appendix A, part 4, question 2), one of the participants believed that the “Writing center is a more personal and hands-on experience in exchange for Grammarly, which is an AI just helping you without proper explanations and responses.” This observation coincides with what Dembsey (2017) identified regarding Grammarly’s ineffectiveness in providing personalized, human feedback. However, it is not surprising that Grammarly is incapable of creating individualized feedback for its users since it is built on an algorithm with a set of rules and code to abide by. Because Grammarly already has a predetermined outline for what an ‘errorless’ piece of text should look like, it will not be able to provide suggestions based on the context of a given text.
Overall, participants maintained a neutral stance on Grammarly’s feedback and suggestions, revealing that students at AUS may have a relatively potent awareness of how faulty the program can be in its edits. In turn, respondents seemed to favor the Writing Center over Grammarly, revealing that personalized assistance may be seen as a priority over the speed and convenience of the application’s feedback.
The Interview
The interview with Abdelgawad delineated some of the issues students typically face when using Grammarly for their academic assignments. However, it has been discussed that Grammarly may not be the sole perpetrator of these problems, but rather it could be the students’ use of the program. As Abdelgawad suggests, the effectiveness of Grammarly in helping a student with their writing “depends on who uses the program and how they and their knowledge of writing will help them use Grammarly better” (S. Abdelgawad, personal communication, November 18, 2024). It appears that a good majority of Grammarly users have a general awareness of the faults behind Grammarly’s feedback, which was identified by the research conducted by Zhang and Özer (2020), showing that students use Grammarly and the writing center for separate reasons. Therefore, it seems that AUS students who use Grammarly possess more of an awareness of its limited feedback than the average Grammarly user.
Furthermore, Abdelgawad suggests that Grammarly is a safer, more accessible alternative to ChatGPT “simply because ChatGPT can write for you” (S. Abdalgawad, personal communication, November 18, 2024). I believe that because of users’ ability to control the edits and suggestions Grammarly gives for their texts, granting them more agency with their writing, it may be easier for tutors and other educators to confront the use of Grammarly in contrast to ChatGPT. However, as Bryan (2024) emphasizes, AI discourse in the writing center has existed for a long period of time—before the advent of AI software such as Grammarly and ChatGPT. Abdalgawad predicts that ChatGPT is most likely going to meet its demise before Grammarly. With that being said, given the persistence of AI discourse in the writing center and the ubiquity of GAI technology in economic, social, artistic, and academic spaces, It may be difficult to foresee the collapse of such pervasive technological advancements. In fact, there seems to be more of an incentive for creative industries to further utilize these technologies in hopes that “Future GAI will surpass traditional single-media formats, integrating text, images, audio, and even video to create richer and more personalized creative experiences” (Zhang et al., 2025).
These descriptions of “personalized creative experiences,” while utopian and optimistic in nature, appear to be promoting a vision of an artistic field of creativity where art is treated as a product to be consumed (and therefore, purely a commodity) rather than a meaningful process of creation. I would even argue that this integration of “text, images, audio, and even video” into these generative models is fundamentally neglectful of that process.
This issue has evidently sept into academic spaces where GAI software seem to safeguard a quick, efficient, and automated completion of coursework, essays, homework, and various other academic tasks. Unfortunately, despite its initial function as a non-generative AI-powered application, Grammarly has been more than keen to adapt these generative models into their program as of late. A built-in GAI service is currently available for users to utilize, allowing them to generate and automatically rewrite entire blocks of text via a simple prompt (Grammarly, 2024). This service not only amplifies a severe reprimanding for university students to effectively develop their academic writing skills, but it further exacerbates the pre-existing problems that stem from Grammarly’s supposed non-generative feedback. Although Abdelgawad suggests that there is a foundational difference between Grammarly and ChatGPT “simply because ChatGPT can write for you” (S. Abdelgawad personal communication, November 18, 2024), the utilization of GAI services among Grammarly users could challenge that distinction.
Conclusion
This paper has examined the perceptions regarding Grammarly and the Writing Center by focusing on Grammarly’s feedback system, the significance of observing AI discourse in academic writing history, and the effectiveness of Grammarly and the Writing Center in relation to one another. Survey and interview findings have revealed that students at AUS were generally aware of some of the issues with Grammarly’s feedback. Furthermore, students have shown a higher favorability towards the Writing Center in comparison to Grammarly, revealing some of the existing scepticism regarding the inadequacies of the program. Another significant topic in relation to this paper, however, is the use of Grammarly by Writing Center tutors. Therefore, by further examining the implications of Grammarly on tutees in the Writing Center, a more in-depth examination of the use of Grammarly among tutors and educators could fill the existing gap in this topic.
Implementing more AI-related workshops/training sessions for tutors in the Writing Center would allow for an elevated awareness and understanding of how Grammarly and other non-generative AI software operate on an algorithmic level and what tutors can do to educate tutees on the productive uses of these programs. However, given the implementation of GAI models in Grammarly’s program (and perhaps the following comment stems from a place of scepticism), one may have to question whether writing centers should be promoting these AI-powered services at all.
With that being said, there are limitations regarding the research behind this paper. Although the primary research conducted for the paper was thorough and detailed, the data collection for both the survey and interview results spanned the course of two weeks only. If there had been more time dedicated to collecting the data, more diverse results could have been utilized for the research. Furthermore, conducting interviews with multiple Writing Center tutors instead of one would have allowed for more diverse qualitative data. Finally, although the survey was designed to be as detailed and considerate of the paper’s goals as possible, results are based on participants’ self-reports, and may not reflect respondents’ true perceptions and/or usage of Grammarly and the Writing Center.
References
Bryan, M.D. (2024). Bringing AI to the center: What historical writing center discourse can teach us about responses to artificial intelligence-based writing tools. Computers & Writing 2023. https://doi.org/10.37514/PCW-B.2024.2296.2.02
Dembsey, J. M. (2017). Closing the Grammarly gaps: A study of claims and feedback from an online grammar program. The Writing Center Journal, 36(1), 63-100. https://www.jstor.org/stable/44252638
Generative Artificial Intelligence. (2025, August 16). In Wikipedia. https://en.wikipedia.org/wiki/Generative_artificial_intelligence
Ghufron, M. A. Rosyida, F. (2018). The role of Grammarly in assessing English as a foreign language (EFL) writing. Lingua Cultura, 12(4), 395-403. http://dx.doi.org/10.21512/lc.v12i4.4582
Grammarly. (2024, July 25). Generative AI directly where you write | Write your best work with Grammarly [Video]. YouTube. https://www.youtube.com/watch?v=Fw2JcR9wtkI&t=33s
Koltovskaia, S. (2020). Student engagement with automated written corrective feedback (AWCF) provided by Grammarly: A multiple case study. Elsevier, 44. https://doi.org/10.1016/j.asw.2020.100450
Zhang, J. Özer, H. Z. Bayazeed, R. (2020). Grammarly vs. face-to-face tutoring at the writing center: ESL student writers’ perceptions. Praxis: A Writing Center Journal, 17(2), 3-17. https://www.researchgate.net/publication/341849218_Grammarly_VS_Face-to-face_Tutoring_at_the_Writing_Center_ESL_Student_Writers’_Perceptions
Zhang, Z., Zhang, J., Zhang, X., & Mai, W. (2025). A comprehensive overview of Generative AI (GAI): Technologies, applications, and challenges. Elsevier, 632. https://doi.org/10.1016/j.neucom.2025.129645
Appendix
Please refer to pages 100 to 105 of the digital issue for all appendices.