Student preferences: using Grammarly to help EFL writers with paraphrasing, summarizing, and synthesizing

This study explores students’ perceptions about using Automated Writing Evaluation (AWE), Grammarly (a paid version), as a complementary instructional tool to teach and support writing from sources. Participants were second-year students (n=37) at a public university in Japan. After in-class tasks aimed at teaching paraphrasing, summarizing, and synthesizing, students completed a survey that measured their perceptions. Students had positive attitudes about Grammarly in general but had somewhat polarized opinions on how useful the tool is in teaching writing from sources and helping with plagiarism.


Introduction
The integration of secondary sources challenges many inexperienced English as a Foreign Language (EFL) academic writers (Hirvela & Du, 2013;Liu, Lo, & Wang, 2013). Undergraduate L2 students often struggle with paraphrasing, summarizing, synthesizing, as well as with appropriate attribution and referencing, which may reduce motivation in the L2 writing classroom. Research suggests it may also lead to unintentional plagiarism (Howard, 1995;Pecorari & Shaw, 2018). In the Asian context, students 'receive limited exposure' to textual borrowing strategies (in Keck, 2014). Yet, these are essential when it comes to writing academic texts. To address these issues, in this exploratory study, the authors focused on the below.
What are students' perceptions about AWE tools and their use as a complementary instructional tool to teach and support writing from sources in an academic writing class?
Grammarly was effective in reducing surface-level errors (Ghufron & Rosyida, 2018), and students had positive perceptions about the tool (O'Neill & Russell, 2019). However, Grammarly's in-built potential as a plagiarism checker 5 or textmatching software was used less in studies about writing from sources.

Method
The study was conducted during a 16-week EFL academic writing course at a Japanese university in 2019. Second-year students (n=37) were placed in three groups (majors, TOEFL ITP 6 scores) for weekly 90-minute classes. The overall goal of the course is to teach writing a 2000-word academic essay. A pre-survey measured students' interests in English and asked about previous AWE experiences (Table 1). Three in-class tasks introduced and practiced paraphrasing (Week 4), summarizing (Week 5), and synthesizing (Week 6). Each writing task was presented to students in the first 30 minutes of the class. Students were then assigned homework, based on in-class tasks, to be submitted online within one week. Due to the relatively small number of participants, Likert items were not combined into scales; thus, the response data were treated as ordinal data, and descriptive statistics were used (Lavrakas, 2008). For detecting trends in open-ended questions, KH Coder (Higuchi, 2016) was used.

Results and discussion
A snapshot of students' overall perceptions is summarized in Table 2 (n=37). The reported percentages, median, and InterQuartile Range (IQR) show students agree on all questions, except about confidence. Most students (49%) do not feel more confident in their writing after doing in-class and homework tasks. It is possible that they became more aware of the gaps in their knowledge both in terms of linguistic and rhetorical aspects of writing after using Grammarly. These answers were corroborated by analyzing open-ended questions. Figure 1 presents the results of the co-occurrence of words: patterns are detected based on the degree of modularity (Higuchi, 2016). Patterns are detected via modules, grouped circles of the same color: Grammarly is easy to use and understand and can help find mistakes and notice instances of plagiarism. Figure 1. Grammarly advantages, the co-occurrence of words, Q34. n=37, N31 D65 D.14 Table 3 (n=37, *n=36) shows perceptions about Grammarly as an instructional tool to help with writing from sources and avoiding plagiarism. Overall, students agree that Grammarly is useful for avoiding plagiarism, and it helped them with enhancing all three textual borrowing skills, although some of their answers are more scattered across the range of six possible answers.
On average, most students, 24 (65%), used Grammarly once a week. When asked about when/where and how they used this tool, students showed the highest levels of disagreement (Table 4).  Analysis of an open-ended question about the use of Grammarly as a tool for writing from sources and plagiarism ( Figure 2) shows several patterns: easy to use; plagiarism is difficult to avoid, but Grammarly can help; students can notice things that they did not notice before. Students did not read Grammarly's explanations, and only paraphrasing was mentioned. Some of the types of the student responses ( Figure 2): "I had used the same sentences without noticing";"It's difficult to avoid plagiarism perfectly but feedback helped me to understand"; and "It made me think I have to paraphrase".

Conclusions
This exploratory study investigated students' perceptions regarding Grammarly as a complementary instructional tool to teach and support writing from sources. It was the first step in the potential implementation of the use of AWE within an EFL academic writing course. As in previous research, participants had positive perceptions of Grammarly and found it useful in addressing shortcomings in their grammar knowledge, word usage, style, and mechanics of writing. Students found Grammarly to be a beneficial instructional tool that can help avoid plagiarism and writing from sources. Next is to look at how this kind of Grammarly use affected student's revisions by looking at the writing samples collected in the study. There are several implications for the practitioners: when deciding to use the tool, instructors should think about student's skill levels -if they have enough metalinguistic knowledge to understand the software output; focus on paraphrasing only; spend more in-class time demonstrating Grammarly's use; and use it in combination with the instructor's feedback.