The Influence of Grammarly towards Indonesian EFL Students’ First -Degree Thesis Writing Confidence

Abstract


Introduction
Thesis writing completion as a requirement to graduate from the study was commonly perceived hindrance for the students. Nevertheless, thesis writing completion is mandatory according to university curricula. The objective is to provide an opportunity and experience for graduates to be able to do the research and report the result of the research. The thesis writing report for first-degree students in English education study programs is even harder, compared to master-degree because it demands academic and scientifically correct English senses.
It was often suggested by examiners and discussants in thesis proposal seminars and examinations that the proposal and research report writing could be attempted for its betterment using an application, Grammarly. To follow up on the suggestion, the researcher was interested in investigating how the suggestions have been consciously responded to by the students, especially the ones enrolling in the last semesters.
Elsya : Journal of English Language Studies https://journal.unilak.ac.id/index.php/elsya 55 The Influence of Grammarly towards Indonesian EFL Students' First-Degree Thesis Writing Confidence As one of the most productive skills, writing is considered a difficult skill to master. It concerns several aspects that might affect language use (Senjahari et al., 2021;Zainab, 2021). Moreover, writing is a thinking process to succeed in the subject matter (de Laurido et al., 2018). Since writing is a thinking process and challenging (Cunningham, 2020;Medvedeva, 2016), its achievement is influenced by one of the psychological factors, that is, self-confidence (Olasina, 2019).
According to Martinez & Villa (2017), "self-confidence is a specific area of psychology related to the language learning process development" (p. 24). This self-confidence in learning is one breakdown of the affective factors. Other breakdowns are emotion, motivation, attitude, anxiety, and personality (Kavitha, 2020). Moreover, this research is on Grammarly use and its qualitative influences on self-confidence in thesis writing.
As an English teacher to be, the students of the English Education Study Program of Tanjungpura University are prepared to master writing skills, especially in terms of sentence grammar or structure, and one way to do it is by using an online grammar checker application, Grammarly. Grammar checkers in writing is beneficial for identifying mistakes and other language errors in writing. This checker is helpful in teaching grammar, improving students' writing, and aiding in the learning process by introducing grammatical terms and rules (Cavaleri & Dianati, 2016). Grammarly, as stated by O'Neill & Russell (2019), is a good and easy online tool for providing grammar feedback.
Several studies have investigated the influence of Grammarly on writing. Some results of the investigation were that Grammarly helped and improved students' understanding of grammar rules and confidence in writing (Ventayen & Orlanda-Ventayen, 2018). Another investigation finding on Grammarly had a significant impact on students' writing quality, developing students' writing accuracy and attitude toward Computer Assisted Language Learning (CALL), and learning English writing (Fadhilah et al., 2018;M. Ghufron, 2019). Moreover, the results of some studies also show that Grammarly can motivate students to learn writing, encourage students to be self-independent learners, and make them enthusiastic to revise their work by reducing the error such as grammar, punctuation, spelling, sentence structure, style, and vocabulary enhancement (Ghufron & Rosyida, 2018;Yulianti, 2018).
This current research is on students' engagement with written corrective feedback, that is based on Han & Hyland (2015), and Grammarly as a tutoring application to write (Zhang et al., 2020). It is a way for foreign language acquisition learning (Lucas, 2022). By type, this research is a case study to investigate the influence of Grammarly on students' writing thesis confidence.
The questions raised up to answer through this research are: 1) What were research participants' engagements in writing thesis by using Grammarly?
2) What were research participants' Grammarly influence towards thesis writing self-confidence?
Meanwhile, the purposes of this research are: 1) To qualitatively describe participants' engagement in terms of behavioural, cognitive and affective engagement.
2) To qualitatively describe participants' Grammarly influence towards thesis writing self-confidence in terms of feedback given, affective filters and indicators of writing confidence.

Writing First-Degree University Thesis
In writing a first-degree university thesis, the students of the English education study program encounter two issues. They are thesis organization or structure and English language proficiency. Since the language is the EFL, in the context of thesis writing, the students have to produce the English language in written The Influence of Grammarly towards Indonesian EFL Students' First-Degree Thesis Writing Confidence form. It is the process of putting down thoughts and ideas and transforming them into words (Lund, 2021) for the research work on paper. Moreover, the students writing the thesis should understand the aspects of writing to make a well-organized text. The aspects comprise content, organization, vocabulary, language use, and mechanics (Meyntjens, 2021). Additionally, writing itself is a difficult task to succeed for it has the aspect of linguistics, physiology-reader's point on the writing, and cognitive-writing instruction (Novospasskaya & Lazareva, 2021). To help write a first-degree university thesis that fits organization or structure and English language proficiency, a specific tool like Grammarly is required.

Self-Confidence in Writing
Self-confidence is described as how one's behaviour and understanding of the influence of particular things to succeed (Crisfield, 2020). As an influential variable with a social psychological approach to foreign or second language learning, self-confidence is the causal factor that directly affected L2 or target language achievement (Alenezi, 2020). This self-confidence has the element of cognitive-knowledge of abilities, performance-ability to do, and comfort of the former two elements. From the three elements, it is concluded that self-confidence affects cognitive or knowledge, performance or behaviour, and affective or emotion (En-Chong, 2022). Hence, self-confidence in writing is writing knowledge, writing behaviour, and writing emotion.
In language learning, especially writing skills, the affective filters play a role in the formation of one's self-confidence. Affective filter is anxiety, motivation, and self-confidence variables. It is a critical mediating role between linguistic input available in the educational environment and the ability of students to learn and to do the writing (Waluyo & Bakoko, 2022). Moreover, self-confidence in English writing includes the function of planning or drafting the writing, editing, and correcting. The specific condition, as in (Irawati, 2022), for this function is: 1) Sufficient time at disposal to edit, correct or revise. It takes sufficient time to produce good writing.
2) To focus on a form of good and brief English sentences to make people understand what is being conveyed.
3) To know the rules of grammar as the requirement in language learning.
Additionally, self-confidence and motivation in writing, specifically thesis writing, are influenced by the corrective feedback given (Junaidi & Hadi, 2020;Yu et al., 2020). Based on the three issues listed and the corrective feedback given, it is vivid that the writing confidence by the EFL thesis writers is elevated when time to edit, correct, revise is sufficient, when English sentence is understandable, and when grammar requirement is fulfilled.

Students' Engagement in Writing
Students' engagement in writing refers to investment or commitment to their learning. It embraces a complex factor to texts and attitudes to writing and responding. In writing, engagement comprises dimensions of behavioural, emotional, and cognitive (Anjarwati & Sa'adah, 2021;Z. V. Zhang, 2017). Moreover, Selcuk et al. (2021) added that the dimensions also include corrective feedback (CF), a cognitive perspective that focuses on how students attend to the CF they receive. A behavioural perspective focuses on how students accept or revise their written texts. Meanwhile, an emotional perspective involves students' affective responses to CF, that is, anxiety and attitude, that is, dislike. Furthermore, the dimensions of students' engagement in writing and feedback in the L2 writing model, as updated by Han and Hyland (2015), is: 1) Behavioural engagement concerns students' reactions to feedback, including revision actions, revision strategies, and time spent on revision.
2) Cognitive engagement concerns how deeply students understand the feedback information and monitor the revision process and metacognitive use. By the EFL thesis writing context, the engagements are thesis writer's behavioural reactions to feedback given, understanding of feedback information, and immediate emotional reactions to feedback. In addition, there are five ways how students dispose of their continuum engagement in school-related writing tasks and activities (Hiromori, 2021;Ku et al., 2022) to be contextualized in this research, namely: 1) Authentic engagement-high attention, and commitment to a writing task.
2) Strategic compliance-high attention to the writing task but having a low commitment at the same time.
3) Ritual compliance-low attention to the writing task, and no commitment to do it. 4) Retreatism-have no attention or commitment to the task. 5) Rebellion-refuse to do the writing task and have no commitment to do it.
The engagement in school-related writing tasks and activities, which in this case is thesis writing revision and submission, are guaranteed to be successful only when the thesis writers have an authentic engagement. This authentic engagement is high attention and commitment to thesis writing obligation that is not optional but conditional to graduate the study. The strategic and ritual compliance, retreatism and rebellion, as sequenced in the above order, are negative disposes to fail the thesis writing success.

Grammarly in ELT Context
Grammarly, in the EFL context, is a helpful online website and application to correct errors in writing. This online grammar checker is popular with the various features provided to help students, specifically in writing. As cited in Villar (2018), Grammarly automatically detects potential grammar, spelling, punctuation, word choice, and style mistakes in writing. It is also functional for online proofreading to scan documents for grammar mistakes, spelling and punctuation correction, vocabulary or word choice suggestions, and plagiarism detection (Gain et al., 2019). By the EFL context, a positive attitude towards learning English writing can be developed through the use of Grammarly (Karyuatry, 2018). Moreover, this Grammarly has two versions−free and premium. The features of the free and premium versions are in the following Table 1.

Features
Free Premium

Correctness
Grammar and spelling checks ✓ ✓

Clarity
Check punctuation, grammar, context, and sentence structure X ✓

Engagement
Vocabulary enhancement X ✓

Delivery
Genre-specific writing style checks X ✓ Plagiarism detector X ✓ Based on the above table, this research deals with the premium Grammarly, and therefore, there were only three students working on their thesis writing selected as a case. They have a premium Grammarly account to completely engage with their thesis writing and revising.

The Influence of Grammarly towards Indonesian EFL Students'
First-Degree Thesis Writing Confidence

Method
As to refer to research methodology types (Creswell, 2012), this study applied a qualitative case study approach to investigate the influence of Grammarly on students' thesis writing confidence. The expression of confidence was collected from students of the English Education Study Program of Tanjungpura University working on their thesis completion. The questionnaire, adopting a questionnaire from Grammarly (2017), was distributed to 40 students enrolling in the ninth semester of the academic year 2020/2021. The questionnaire was closed-ended (Iwaniec, 2020), and the objective of administering them was to find out students' attitudes toward Grammarly and to determine the correct research participants.
The result of the questionnaire identified that 29 students used Grammarly to help write their writing subjects, and from the identification, there were three students working on their thesis writing completion chosen as the case to expose. The criteria employed to select the three participants for this research were on the basis of (1) working on their thesis writing completion, (2) possessing a premium Grammarly account, and (3) having a positive attitude towards Grammarly. The initials of the participants are IW, UA, and AA.
The data for this research is primary, that is, first-hand data (Glen, 2022). The data, following minor taxonomy adaptation (Ferris, 2012), was in the form of error types and numbers collected from reviewed participants' thesis, and its revision using premium Grammarly. Moreover, the data in the form of opinions or expressions were collected by interviewing participants. The interview guideline was a one-on-one interview with semi-structured questions (Ahmad, 2012). The questions asked in the interview dealt with participants' behavioural, cognitive and affective engagements, and on writing self-confidence in terms of feedback given, affective filters and indicators of writing confidence. The data, which in this case is nonparametric (Winter, 2019), were analysed by calculating the percentage of the error types based on their number. Meanwhile, the participants' opinions or expressions to enrich the research finding and discussion, adapting Naess (2020), were presented by excerpting the result of the interview.

The Participants' Engagement with Grammarly
The first result of the research was descriptive statistics or non-parametric data. The data was tabulated by errors types and quantity listed from each participant. The participants' engagement with Grammarly in terms of error types and numbers and revisions is presented in Table 2. The participants' engagement with Grammarly, as shown in Table 2 is furthermore broken down into subtypes of behavioural, cognitive, and affective engagements, and presented below consecutively.

Behavioral engagement
Behavioural engagement with Grammarly involved revision actions and strategies to enhance the accuracy of writing and time spent on revision. Table 2 illustrated that each participant had a different revision action. Firstly, IW's revision actions in response to Grammarly feedback showed 545 errors in his 8,800 words of text with fifteen error types. From those error types, articles were the most identified by Grammarly. Of 545 received Grammarly feedback, of which 512 were accurate, and 33 were inaccurate. IW correctly accepted 363, incorrectly rejected 138, incorrectly substituted 11, incorrectly accepted 15, and correctly rejected 18, thus fixing 66.6 % of his total errors. This suggested that IW made moderate changes to his writing. Secondly, UA's revision actions showed 220 errors in her 7,781word text, with fourteen error types identified in her thesis writing. The most frequent error in UA's thesis writing was phrase structure. From 220 received Grammarly feedback, which was accurate, UA correctly accepted 148, incorrectly rejected 72, and fixed 67.2 % of her total errors. This suggested that UA made moderate changes to his writing, like IW. Lastly, AA's revision actions showed that AA made 206 errors in her 5,724 words of text for which she received Grammarly feedback. There were thirteen error types identified in her writing and the most identified were phrases and idioms. From 206 received Grammarly feedback, of which 123 were accurate, AA correctly accepted 123, and incorrectly rejected 83. Thus, also fixing 59 % of her total errors. This suggested that AA made slight changes to her writing, different from IW and UA.
Concerning revision strategies and time spent, IW asserted that he used the internet as an extra assistance to verify the accuracy of the feedback. However, it differed from UA and AA in that they used their intention and limited understanding by being careful in analysing the feedback. Interestingly, they spent a little time revising each part of their writing continuously despite having full time in revision. It was because they needed to convince themselves before compiling his revisions for the supervisors.

Cognitive engagement
Cognitive engagement is concerned with how deeply students understand the feedback information and monitor the revision process. Regarding the monitor level, IW reported that Grammarly feedback was conspicuous in that there are some colours that signed the error IW made in his writing. Since he read the feedback that is explicit and implicit in nature, he liked to recognize the corrective intent of key feedback because it was easy to notice and understand the errors. Therefore, he found several problems in his writing, including 275 alerts of correctness, 260 related to clarity, 5 related to vocabulary engagement, and 5 related to his writing style. He realized that Grammarly gave him a lot of information.
Unlike IW, who encountered many problems, the participants with initial UA, and AA found lesser problems in their writing. UA found 86 alerts of correctness, 132 related to clarity, and two related to delivery. Meanwhile, AA found several errors in her writing, that were 77 alerts related to the correctness, 108 related to clarity, and 21 connected to delivery. They did not revise their writing with application assistance as extra as IW. It showed how they used a metacognitive action strategy by monitoring the errors, even though they verified the accuracy of Grammarly's feedback without needing extra assistance and awareness after receiving the feedback regarding the errors in their writing.
Furthermore, IW realized he made a lot of errors in his writing. One of the errors was typos of 'definition' with misspelling /e/, and 'properly' with missing /r/. The screenshot of errors committed is in Figure 1.

The Influence of Grammarly towards Indonesian EFL Students'
First-Degree Thesis Writing Confidence It also illustrated how IW accepted the suggestions and found the other mistakes, as in sentence […] the teaching-learning process in the classroom will not go properly. In the sentence, a modal verb or auxiliary verb, will, is an alternative to 'goes', the verb with the suffix -es for singular third personal pronoun subject of simple present tense.
Moreover, Grammarly identified two mistakes in a sentence in terms of punctuation, that is, coma, and unnecessary prepositions, that is, /of/, made by UA. For detail, please see the red underline and the highlight in the following Figure 2.

Figure 2. An Example of UA's Punctuation and Preposition Error
Finally, AA also encountered errors in terms of linguistics, that is, passive voice for simple present tense. The verb component for the tense is To Be + Verb 3. Thus, the correct form, as shown by the application, is also found. Look at Figure 3 for the detail.  Figures 1, 2, and 3, the three participants tended to ignore some feedback that Grammarly gave because they thought it did not affect their writing. The evidence demonstrated how both UA and AA used context to determine the appropriate revision action. It indicated that their cognitive engagement with Grammarly is slight. Meaning to say, the way they monitored and understood the explanation and context of their writing carefully by reading the draft before deciding whether or not to accept feedback using Grammarly indicated that their metacognitive strategies were implemented without any extra assistance. Unlike the two participants, IW's cognitive engagement was quite extensive, manifested in the use of several meta-cognitive actions to regulate his thinking process. It made IW need more context to analyse the accuracy and decide whether to accept or reject the feedback.

Affective engagement
Affective engagement is related to students' immediate emotional reactions to Grammarly feedback upon receiving the reaction and their attitude towards Grammarly feedback. In terms of this reaction and attitude, IW showed his emotional response towards Grammarly feedback that was a bit distrustful. The data from the interview on Affective Engagement (AE), that is, Reaction and Attitude (RA), as excerpted from participant IW, is in the following quotation: The Influence of Grammarly towards Indonesian EFL Students' First-Degree Thesis Writing Confidence "I don't believe 100% Grammarly feedback because I have to analyse, I have to look for the same materials about the feedback from the Internet may be, and decide to change it or not. But anyhow, it helped although I still have to check particular vocabulary or terminology in dictionary for correct language use and context." Excerpt from data AE-RA-IW Despite his distrustfulness of Grammarly feedback, he was surprised about the amount of feedback received and how conscientious Grammarly was. However, he was a bit sceptical because some errors were due to his carelessness. Unlike IW's distrust, participants UA and AA showed emotional reactions upon receiving Grammarly feedback. They were surprised and grateful that their linguistic errors were shown and suggested revised. The interview made with the three participants implied, in general, that emotional reactions toward Grammarly feedback affected cognitive and behavioural engagement.
In terms of Grammarly Impression (GI), the interview excerption on Affective Engagement (AE), is as in the following citation from participant UA: "My first impression towards Grammarly is, I think, is like "What a good application! It is really helpful for students to make their writing better than before. When I started writing my thesis, I felt that I was certain to be able to complete my study in time."

Excerpt from data AE-GI-UA
When presented with students' engagement, it is shown that participant IW led significant levels of engagement in terms of behavioral and cognitive, followed respectively by UA and AA. His leading was due to how prominent IW was in the revision strategies using the assistance of the application and how he understood by questioning the accuracy of feedback. In contrast, UA and AA used their intuition to verify the accuracy of the feedback.
Substantial affective engagement was demonstrated by participants UA and AA. They showed an emotional reaction and attitude response. From the description, it is concluded that the level of engagement of participants towards Grammarly is strategic compliance. They have high attention on one side but low commitment on another side.

Feedback is given during Writing Activities
Giving feedback during writing activities was an essential factor in developing writing confidence. Thus, all the participants reported their writing quality, cognitive understanding, and confidence while receiving Grammarly feedback in writing. The following interview result citation on the Feedback (FB) that influenced Writing Confidence (WC) is excerpted from participant AA: "I wasn't confident about my writing, but after using Grammarly, I felt like my writing seemed to be better because previously there were many mistakes, and Grammarly helped me in fixing the writing mistakes. It helped me in improving my writing and confidence. For example, when I wrote, there were some mistakes that I didn't know or didn't realize. So, Grammarly influenced me to know or realize them. I felt helped and confident that my writing got better."

Excerpt from data WC-FB-AA
The above interview result represented by participant AA inferred that Grammarly improves English writing quality, expands cognitive understanding, and promotes confidence through feedback. Moreover, the feedback also signified that Grammarly develops an awareness of linguistic errors. The interview result indicated that the participant was confident to submit or resubmit their thesis draft to supervisors.

The Influence of Grammarly towards Indonesian EFL Students'
First-Degree Thesis Writing Confidence

Affective Filters
According to the excerpt from the interview, it was concluded that the participants had a considerable impact on confidence in EFL learning, especially in English thesis writing. In terms of linguistics and grammar knowledge, the Anxiety (AX) as an Affective Filter (AF) was expressed by IW as follows: "You know, even the one that is best at English proficiency, may still encounter linguistics and grammar anxiety. But at least, Grammarly has helped me to reduce that anxiety. Grammarly was an essential tool for me to overcome my linguistics or grammar difficulties in writing my thesis." Excerpt from data AF-AX-IW The above excerption explains how participant IW felt linguistics and grammar knowledge anxiety he encountered have been reduced because of Grammarly application assistance. Linguistics and grammar knowledge, especially the English language, deal with form and meaning, or what form signifies. Form, that is lexical and grammatical, associates with part of speech, tenses and plurality. The parts of speech is verb, noun, adjective and adverb. The tense covers active and passive verb in a sentence. The verb is first, second and third form that is regular and irregular. Meanwhile, the plurality the form signifies is countable and uncountable noun. Premium Grammarly features covers all those English linguistics and grammar knowledge, and overcomes their hindrances in writing, especially thesis writing.
Moreover, in terms of Motivation and Writing Workload (MWW) as subsection of Affective Filter (AF), UA stated how Grammarly has increased motivation and released workload. Here is the statement: "When we studied writing and grammar or structure subjects, I was not motivated because those subjects were too difficult to catch. They were too burdening, workloads. Magically, my motivation in writing and grammar increased especially in writing my thesis. Grammarly, with its features, helped reduce the workload."

Excerpt from data AF-MWW-UA
The excerpt taken from data as expressed in the interview by UA shows that motivation and workload are associated with each other. The association is the high motivation, the light workload; or vice versa, the low motivation, the heavy workload. The conclusion from the interview was that UA had high motivation to enlighten his thesis writing workload because of Grammarly help.
In terms of Self-esteem (SE) that is also breakdown of Affective Filter (AF), AA put forward that Grammarly has promoted self-esteem when writing the thesis. The AA's expression goes in the following: "I think writing a thesis in English is very much different compared to the one written in, for example, Indonesian. It is even different from general writings of English subjects. Grammarly as an application proved that my selfesteem was developed when I rewrote or revised my thesis corrected by supervisors. The supervision and submission were not so frequent."

Excerpt from data AF-SE-AA
The participant AA as read in the statement above indicated that his self-esteem or self-confidence, that is the breakdown of affective filters had been developed because of the Grammarly application that was helpful for him. For the self-esteem he had, his thesis writing revision was faster and the submission to supervisor was also infrequent.
From the interview excerption results, it was signified that affective filters have influenced participants' self-confidence. The use of Grammarly reduced anxiety about linguistics and grammar knowledge, and selfdoubt. Moreover, it was also inferred that the use of the application immensely increased motivation and lowered the writing workload. In addition, the application also increased participants' self-esteem considerably.

The Influence of Grammarly towards Indonesian EFL Students'
First-Degree Thesis Writing Confidence

The Indicators of Writing Confidence
The indicators of writing confidence, as excerpted from the interview results, comprises flexibility in behavior, cognitive effort, emotional stability, motivation, and self-esteem. In terms of the Indicator of Writing Confidence (IWC), that is Behavioral Flexibility (BC), the data from the interview, as excerpted from participant IW, is in the following citation: "I spent a long time making my revisions. I was confident to decide or to accept the feedback shown by Grammarly. The feedback shown by Grammarly made me flexible. I was not afraid of missing word choice or alternatives because Grammarly has them all." Excerpt from data IWC-BC-IW The above citation indicated that the participant showed reactions toward Grammarly when writing. The behavior was adapted according to the feedback provided when making revisions, showing actions and strategies worked and whether or not to accept the feedback.
Furthermore, based on the Indicator of Writing Confidence (IWC), that is, Cognitive Effort (EF), the interview from participant UA, as in the following excerpt, is: "I made some effort to monitor the bold mistakes shown by Grammarly in my writing and had to revise accordingly, not to accept the feedback immediately. In this case, as I observed this far, Grammarly often suggests the active voice meanwhile the sentence I wrote was passive voice. If I accept the suggestion for revision, the meaning will imply differently."

Excerpt from data IWC-EF-UA
The above interview excerpt implied that the participant used metacognitive strategies, that is, planning, monitoring, and evaluation when revising, and remaining confident with the choices offered by Grammarly. It infers that metacognitive strategies were used by the participant to plan, monitor and evaluate the revision. The strategies enabled her to accept or to reject the passive voice suggested by the Grammarly.
Moreover, based on the Indicator of Writing Confidence (IWC), that is, Emotional Stability (ES), the participant AA's interview summary, is as follows: "Grammarly helped me revise my writing, and the feedback overcame my difficulties. So, it was relieved in terms of emotional stability, and therefore, I think it promoted my confidence."

Excerpt from data IWC-ES-AA
The summary of the interview revealed that the emotion was stabilized during and after the revision. The indicator of emotional stability of writing confidence using Grammarly supported writing capability and reduced anxiety. Emotional stability was maintained for not being stressful because the difficulties shown through the feedback from Grammarly application was also be overcome by the Grammarly. To add more, it was concluded from the interview result on Motivation (M) as one of the Indicators of Writing Confidence (IWC), that the participants of the research were motivated in identifying undetected errors using Grammarly. The workload of writing errors was also solved and decreased.
Finally, the last Indicator of Writing Confidence (IWC) identified from the interview was Self-esteem (SE). Based on the interview, two participants felt that their self-esteem in writing increased. This SE as an indicator of IWC is related to the ability, including the effort to not give up, to succeed in the thesis writing and revision before submitting them to supervisors.

Discussion
The participants' engagement with Grammarly exposed as a case for this research was behavioral, cognitive, and affective engagement. The engagements influenced participants' confidence in writing and revising their thesis draft before the submission to their supervisors.

The Influence of Grammarly towards Indonesian EFL Students'
First-Degree Thesis Writing Confidence The behavioral engagement involved revision actions, strategies, and time spent. The general description of the participants on behavioral engagements were the prompt action to revise, the proper strategies used, the less time spent, and the thesis writing revision was quicker to submit to supervisors. The appropriate revision actions, strategies, and time spent was influenced by use of Grammarly as grammar checker that helped to put ideas in thesis writing by context. In line with this behavioral engagement, participants IW and UA, based on Table 2, made a moderate change to their draft, which focused on eliminating Grammarly-detected errors. The errors detection corrected were 66.6% for IW and 67.2% for UA of the total errors. Meanwhile, participant AA made a slight change to her draft, as she corrected 59% of her total errors. Furthermore, the participants, in general, spent over 30 minutes on each part of their thesis writing. It indicated the number of errors to revise and the speed to revise using Grammarly. It meant that the participants' behavioral strategy was at the surface level, for they barely used revision strategies to refine their draft. The total errors and the speed of revision of the three EFL thesis writing participants were in connection with Zhang et al. (2020), stating that L2 proficiency seemed to influence how participants conducted revision actions. The lower levels of L2 proficiency tended to adopt a form-focused approach and make surface-level changes in revision. It concluded that time spent was less because of the service provided from the premium Grammarly features.
The cognitive engagement included the depth or the level of understanding of feedback given. This cognitive engagement implemented metacognitive and cognitive actions. The level of understanding of the feedback given, in this case, refers to linguistics and grammar knowledge of the English language, and how to put ideas by the context of the thesis subject. Since the case of this study is in EFL context, hence, the cognitive engagement involving metacognitive and cognitive action is the participant's capability to make their thesis writing correct in English senses.
To refer to the cognitive engagement, the three participants noticed Grammarly's feedback and recognized its corrective intentions, implicitly and explicitly. The participants seemed to understand the causes of errors in their writing and how to correct them, therefore, the finding of this research was considered contradictory to Han & Hyland's (2015) insights. The insights were the students' engagement with written corrective feedback that was insufficient because of linguistic knowledge awareness. The participants of this research were confident to revise in terms of linguistics awareness because of Grammarly's assistance. Despite the recommendation suggesting that it was flawed to understand, the participants' thesis writing was successfully revised.
The affective engagement involved immediate emotional reactions and attitudinal responses to the feedback given. The three participants experienced different emotional responses. The differences were distrust of feedback given for sufficient linguistics knowledge, and trust of the feedback given for insufficient knowledge on the linguistic errors. It meant that the trust and distrust to feedback given by Grammarly were influenced by sufficient knowledge on linguistics and grammar of English language. The sufficient knowledge the participant had, the less dependent they were on the Grammarly. This less dependence represents distrust. Conversely, the less sufficient knowledge of the English language linguistics and grammar, the more dependent they were on the Grammarly assistance. This dependence signifies the trust. Nevertheless, this affective engagement, overall, corresponded to the tendency of Grammarly use in writing (Cavaleri & Dianati, 2016) because the percentage of errors detection and revision was averagely more than 50% as seen in Table 2.
Moreover, this research indicated that a positive attitude towards Grammarly supported and influenced writing confidence. The use of Grammarly engaged sufficient time to do good revision and a deep understanding of rules and forms of language as put forward by Lucas (2022). The feedback given during writing activities, as engaged by the three participants, was an influential factor in determining writing The Influence of Grammarly towards Indonesian EFL Students' First-Degree Thesis Writing Confidence confidence, especially in terms of grammatical concepts. To summarize the discussion, this research finding is generally supporting O'Neill & Russell's (2019) positive perception of Grammarly for corrective feedback.
Grammarly's assistance in thesis writing, based on the indication described in this research, was proven to influence confidence. The use of Grammarly boosted flexibility in behavior, cognitive efforts, emotional stability, motivation, and self-esteem in writing.
The participants of this research were first-degree students of English education study programs who worked on their thesis writing completion; therefore, the influence of Grammarly towards thesis writing confidence resulted and described in this research was not generalizable, and only applies as a case to expose to public readers or audiences to be taken precedence. The influence of Grammarly on writing thesis confidence does not apply to general students of any study programs. The precedence is important as postulate of how writing thesis completion was the obligation must be fulfilled to graduate from the study from English Education Study Program.

Conclusion
This current research found that participants' engagement in terms of behavioral, cognitive and affective was complex and diverse. The complex and diverse engagement required behavioral reactions, cognitive efforts, and affective responses to produce effective thesis writing revisions with Grammarly. Furthermore, it was found that the participants' Grammarly influenced towards thesis writing self-confidence in terms of feedback given, affective filters and indicators of writing confidence. In terms of Grammarly and thesis writing self-confidence, this research implied that premium Grammarly positively influenced or impacted students' writing confidence. The influences were on the feedback provided during writing activities through premium Grammarly features, monitor model for learning, and affective filters like anxiety reduction. Moreover, this research excerpted that participants' thesis writing confidence was influenced by indicators of behavioral flexibility, cognitive efforts, emotional stability, motivation, and self-esteem. Since this research was a case, the good precedence described could be imitated and escalated to promote the number of Grammarly users, especially the students of English education study programs across the country working on English writing subjects, particularly on thesis writing completion.