Students’ preferences and teachers’ beliefs towards written corrective feedback

________________


INTRODUCTION
English is an international language, and therefore learning English is necessary for global communication. Besides, being able to use English for communication gives numerous advantages for individuals, especially for foreign language learners. Khunaivi and Hartono (2015) stated, "In Indonesia, English belongs to a foreign language in which it is used for academic purposes, job vacancies' requirement, and traveling overseas" (p. 15). Hence, it is very helpful to learn and be capable of using English.
Writing is one of the most essential skills for EFL students alongside reading, listening, and speaking. It is also the most difficult skill to learn among others. "when compared with other fundamental skills such as listening, speaking, and reading; writing is the most difficult skill because it requires writers to have a great deal of lexical and syntactical knowledge as well as principal of organization in L2 to produce a good writing." (Tangpermpoon, 2008, p. 1). It has been teachers' job to find appropriate teaching techniques to encourage students' success in learning such skill.
One of the techniques commonly employed by teachers to improve students' writing skill is through the provision of written corrective feedback (WCF). In this context, WCF is a written response made by a teacher that aims to correct linguistic errors found in students' written text. Bitchener and Storch (2016) added that "it seeks to either correct the inaccurate usage or provide information about where the error has occurred and/or about the cause of the error and how it may be corrected" (p.1).
Many aspects of writing can be given feedback by teachers, such as form (grammar, mechanics, and vocabulary) and non-form (organization and ideas). More than often WCF has been used as a technique for correcting grammatical errors as well as other errors found in students' written text. Even so, the effectiveness of WCF to improve students' writing skills is still debatable.
Based on prior investigations conducted by the researcher, it is found that some of the students claimed that they have difficulties in handling their teachers' written feedback given to their written errors. After being investigated in further, it turned out that some of them preferred certain kind of feedback rather than the ones given by their teachers. They also criticized their teachers' written feedback because they often receive not enough or too many feedback which made some of them discouraged. The dissimilarity between students' and teachers' perceptions may lead to misunderstandings and ineffective learning. This is supported by Horwitz (1990), Kern (1995), andSchulz (1996) as cited in Brown (2009) who stated, "mismatches between FL students' and teachers' expectations can negatively affect the students' satisfaction with the language class and can potentially lead to the discontinuation of study" (p. 46). Therefore, some studies are needed to look into both students' and teachers' perceptions regarding the WCF in order to give better decisions for the teachers in using certain types and amount of WCF.
There are also many pros and cons related to the effectiveness of WCF in writing classes. Truscott (1996) initiated the debate against the usefulness of WCF specifically on grammar correction. He stated in his review article: My thesis is that grammar correction has no place in writing courses and should be abandoned. The reasons are: (a) Research evidence shows that grammar correction is ineffective; (b) this lack of effectiveness is exactly what should be expected, given the nature of the correction process and the nature of language learning; (c) grammar correction has significant harmful effects; and (d) the various arguments offered for continuing it all lack merit (p. 328).
This argument received many cons by several researchers in the same research field. Responding to Truscott's views, Ferris (1999) argued as cited in Tseng (2018): Teachers should continue correcting grammatical errors because: 1) L2 students wanted it, 2) students needed to produce academic text with manageable errors to proceed to mainstream curriculum, and 3) students should become self-sufficient in editing (p. 160).
Teachers and students are the primary subjects involved in WCF. Hence, their perceptions and preferences towards WCF are considerable. This is supported by Lee (2008) who stated that, "without understanding how students feel about and respond to teacher feedback, teachers may run the risk of continually using strategies that are counter-productive" (p. 145). Accordingly, in order to achieve effective WCF practice, it is crucial to see whether students' preferences are in line with teachers' beliefs in practicing WCF or not.
There are many studies focusing on the effectiveness of WCF in specific (e.g., Ahmad, Saeed, & Salam, 2013;Bitchener, Young, & Cameron, 2005;Ellis et al., 2008;Baleghizadeh & Dadashi, 2011;Sheen, 2007;Bitchener & Knoch, 2010b). However, a few have explored the aspects which determine its usefulness: students' preferences and teachers' beliefs towards the usefulness of WCF itself. Moreover, published literature that investigates this topic in Indonesia is still scarce. Therefore, the researcher is interested to conduct a study in this area.

METHODS
The objective of the present study is to explain students' preferences and teachers' beliefs towards the usefulness of WCF. Thus, the descriptive design is adopted. The study will combine quantitative and qualitative research methods. The researcher uses quantitative research in order to gain statistical data regarding students' preferences and teachers' beliefs towards WCF. In addition, qualitative research is conducted to obtain more descriptive information regarding students' and teachers' reasons why they preferred certain types of WCF and particular error types that should be corrected.
The subject of this study included five EFL teachers and 35 EFL students majoring in Software Engineering in SMK Negeri 1 Bawang Banjarnegara. The student participants consist of 15-16 years old males and females. The majority of the students have been learning English since junior high school. The teacher participants consist of one male and four females. Most of the teachers have been teaching for at least nine years. In addition, one of the teachers has exact 29 years teaching experience to this date. Thus, the teachers are considered as experienced teachers.
In collecting the data, the researcher distributed questionnaire sheets to the students and interview questions delivered to the teachers. Both instruments had almost similar questions that focused on certain questions: (1) what amounts of WCF do you think are most useful?; (2) what types of WCF do you think are most useful?; (3) what types of errors do you think should be corrected?
The collected data were then imported into an excel spreadsheet for quantitative analysis. The data were categorized based on common themes to make the analysis easier and more readable. Afterwards, the data were imported into SPSS for statistical analysis. The analysis involved several tests which compared both students and teachers' responses whether there were significant differences between the two subjects. After doing the analysis, the researcher concluded students' and teachers' responses. Finally, the results can answer the present study's research questions.

Useful Amounts of Written Corrective Feedback
On questionnaire and interview question item number one, participants were asked regarding their preferences about how many WCF teachers should provide on students' written errors. On this item, participants were allowed to choose one or more options; thus the total percentages of participants' choices can be more than 100%. Table 1 demonstrated the statistics for options chosen by student and teacher participants. The researcher then asked participants' explanations for their preferences towards different amounts of WCF. 64.41% of the participants (students=69%; teachers=40%) pointed out that students should know all of the errors. In addition, 24.39% of the participants (students=23%; teachers=40%) believe that teachers should only mark major errors focusing on the current topic. Interestingly, 9.76% of the participants (students=6%; teachers=40%) think that too many markings can make students discouraged. Finally, 2.44% of the participants (students=3%) argued that teachers should prioritize the content by only marking errors interfering the communication ideas.
In order to obtain more data that support the findings, questionnaire and interview question number nine asks the participants whether teachers must correct repeated errors on students' written work. As shown in table 2, the results demonstrated that most teachers and learners think that it is helpful to mark a repeated error whenever a learner makes the same error. Fisher's exact test shows that significant difference was not found between students' and teachers' views in this regard (p=.427). The researcher also asked participants' reasons for Correction for Repeated Errors. 60% of the participants (students=64%; teachers=40%) believe that teachers should correct repeated errors so that students understand their errors better. In addition, 17.5% of the participants (student=20%) think the same because correction for repeated errors can remind students of their errors. Moreover, 12.5% of the participants (students=9%; teachers=40%) argued that correction for repeated errors will not make students repeat the same errors. In contrast, 5% of the participants (students=6%) believe that correction for repeated errors is not helpful because students should think about their errors and do it themselves. Moreover, 2.5% of the participants (teachers=20%) think similarly by noting that students can instead ask their peers for their reoccurring errors. Finally, another 2.5% of the participants (students=3%) stated that teachers should do it once or twice only.

Useful Types of Written Corrective Feedback
Questionnaire item number one and interview question item number one to seven asked the opinions of participants regarding the helpfulness of different WCF types. The participants rated each question from scale 1 to 5 (1 = least helpful, 5 = most helpful). Table 3 presents preferences of both participants regarding useful types of WCF. The researcher also asked the reasons of participants for choosing Clues or Directions on How to Fix Error. The test displayed that there is no significant difference between student participants and teacher participants (p=.636). According to the result, 52.5% of the participants (students=57%; teachers=20%) think that clues or directions are helpful because they let students practice selfcorrection and they will make students remember the errors better. Additionally, 5% of the participants (students=3%; teachers=20%) also believe that clues or directions are promising to follow since they will surely guide students to find the correct answer. On the other hand, 22.5% of the participants (students=26%) believe that clues or directions are not enough because students need clearer explanations so the students can find the correct answer effectively. Moreover, 5% of the participants (students=6%) demonstrated that they believe clues or directions are not straightforward and leave too much work for the students. Furthermore, 12.5% of the participants (students=6%; teachers=60%) believe that clues or directions are only suitable for clever students. In addition, 2.5% of the participants (students=3%) stated that some students may not have access to find out the clues or follow the directions because some of the students may not have the resources like books or any references that the clues or directions usually require.
The researcher also asked for participants' explanations for Error Identification. The test displayed that there is a significant difference between student participants and teacher participants (p=.013). The result shows that 47.5% of the participants (students=51%; teachers=20%) consider error identification useful because it helps students get noticed where the errors occur so then they can make follow-up corrections. 2.5% of the participants (students=3%) also think similarly that error identification makes students more observant of occurring errors. In addition, another 2.5% of the participants (teachers=20%) believe that error identification makes students more self-reliant because they have to do follow-up corrections by themselves. However, on the opposing side, 40% of the participants (students=40%; teachers=40%) believe that error identification is not useful because it does not provide enough information for the students and they need more information on how to handle the errors. Additionally, 5% of the participants (students=6%) believe the same because error identification gives students a lot of work to do and they have to put quite effort to correct the errors. Furthermore, 2.5% of the participants (teachers=20%) think that error identification is not always practical as it is only suitable when students study certain topics.
The researcher then asked for participants' explanations for Error Correction with Comment. The test displayed that there is no significant difference between student and teacher participants (p=.618). According to the findings, there are 37.5% of the participants (students=14%) who believe that error correction with comments is helpful for students because it helps them know why an error exists and also the correct form to the occurred error. In addition, 37.5% of the participants (students=43%; teachers=60%) pointed out that although it provides commentary for the error, the error correction alone is helpful for the students to solve the occurring error. Moreover, 12.5% of the participants (students=14%) think similarly that error correction with comments gives students detailed information about the error handling so the students can learn much from there. Furthermore, 12.5% of the participants (students=14%) pointed out only for the value of comment that it helps students understand why errors exist. Moreover, 5% of the participants (students=3%; teachers=20%) believe that error correction with comments is useful; however, it has a negative impact that it makes students effortless because the teacher gives spoon-feeding to the students. Additionally, 2.5% of the participants (students=3%) think that error correction with comments is helpful because students do not have to do much work for the correction since the information is quite detailed. Another 2.5% of the participants (students=3%) argued that error correction with comments is useful because it gives a new insight to students that learning English is not difficult. In contrast, on the opposing side, 2.5% of the participants (students=3%) demonstrated that error correction with a comment is not helpful because it does not promote self-correction for the students.
The researcher also asked participants' reasons for choosing Overt Correction by the Teacher. The test displayed that there is no significant difference between student participants and teacher participants (p=.369). According to the findings, 57% of the participants (students=63%; teachers=20%) think that it is helpful for students because it helps them to know the correct forms of occurring errors. In addition, 17.5% of participants (students=17%; teachers=20%) agreed that error correction is useful; however, they believe that it is not enough. They also consider the addition of comments to the feedback because they are necessary. Moreover, 7.5% of the participants (students=9%) think that error correction without comment promotes more self-correction for students; and therefore it is helpful. Furthermore, another 7.5% of the participants (students=3%; teachers=40%) stated that error correction without comment is more straightforward because students can directly see the correct form of their errors. Contrarily, 7.5% of the participants (students=6%; teachers=20%) believe that it is not helpful for students because they do not pay attention to the nature of why the error occurred. They only see the correct form without explanations. Additionally, 2.5% of the participants (students=3%) argued that error correction does not promote self-correction for students.
The researcher also asked participants for choosing Comment without Correction. The test displayed that there is a significant difference between student participants and teacher participants (p=.047). According to the result, 27.5% of the participants (students=29%; teachers=20%) believe that comment without correction is helpful because it makes students more active in looking for the correct form. In addition, 20% of the participants (students=20%; teachers=20%) also think that comment without correction is useful as long as it is explanatory. In contrast, 27.5% of the participants (students=31%) consider that comment without correction is not helpful because it is too confusing for students; and thus they do not understand it. 22.5% of the participants (students=20%; teachers=40%) believe the same as well regarding its usefulness. They think that comments are not enough and students need correction in addition. Besides, 2.5% of the participants (teachers=2.5%) argued that comment without correction is only suitable for clever and high-level students.
The researcher also asked participants' explanations for No Feedback on students' written errors. The test displayed that there is no significant difference between student participants and teacher participants (p=.373). The majority of the participants believe that giving no feedback to students' written error is not helpful. 92.5% of the participants (students=94%; teachers=80%) argued that without feedback students would assume that there are no errors in their written work. Moreover, 2.5% of the participants (teachers=20%) commented that it is the teachers' duty to provide feedback on students' written errors. Furthermore, another 2.5% of the participants (students=3%) argued that giving no feedback to students' written error indicates that teachers are ignorant to students. On the other hand, there are 2.5% of the participants (students=3%) who argued that giving no feedback to students' written error is not a problem. It is said that without receiving feedback from teachers, students will experience less stress regarding the occurring errors.
The researcher also asked for participants' explanations for a particular WCF type, specifically Comment on Content/Ideas. The test displayed that there is no significant difference between student participants and teacher participants (p=.567). According to the result, 37.5% of the participants (students=37%; teachers=40%) believe that giving comments about the content or ideas on students' written work is helpful because it makes students feel motivated and dedicated. In addition, 2.5% of the participants (students=3%) stated that comment on content is acceptable. Moreover, another 2.5% of the participants (students=3%) argued that comment on content challenges students to correct errors by themselves. At the same time, 52.5% of the participants (students=51%; teachers=60%) think that comment on content/ideas is not helpful because it is not enough. They believe that grammar errors should be responded too. Additionally, 2.5% of the participants (students=3%) believe the same that students prefer grammar correction to comment on content. Furthermore, another 2.5% of the participants (3%) argued that comment on content will not affect students in learning the material after all.

Types of Errors that should be corrected
Item number four of the questionnaire and number ten to sixteen of the interview questions ask the participants' opinions about what types of errors that should be corrected. Participants chose rating for each question that describe their preferences (1 = least helpful, 5 = most helpful). Table 4 presents students' and teachers' mean results for types of error that should be corrected. The test for WCF on organization errors displayed that there is no significant difference between student participants and teacher participants (p=.310). According to the data, 65% of the participants (students=74%) believe that teachers need to give WCF on organization errors because it helps students understand correct writing organization. In addition, 20% of the participants (students=20%; teachers=20%) stated that giving WCF on organization errors helps to make students' writing more understandable. Moreover, 10% of the participants (teachers=80%) argued that as long as giving WCF on organization errors affects students' writing quality, then the teacher should do it. Furthermore, 2.5% of the participants (students=3%) pointed out that giving WCF on organization errors motivates students to learn more about writing organization. Interestingly, 2.5% of the participants (students=3%) believe that grammar is more important than organization so teachers should focus on grammar instead.
The test for WCF on Grammar Errors displayed that there is no significant difference between student participants and teacher participants (p=.146). 82.5% of the participants (students=91%; teachers=20%) argued that it is important because it helps students to understand the correct grammar. Moreover, 10% of the participants (students=6%; teachers=40%) argued that grammar is important so students need to receive WCF on it. Additionally, 5% of the participants (students=3%; teachers=20%) think that correcting grammar is the most important; thus giving WCF on it is simply necessary. Finally, 2.5% of the participants (teachers=20%) believe that as long as giving WCF on grammar errors affects students' writing quality, then the teacher should do it.
The test for WCF on Content/Ideas displayed that there is no significant difference between student participants and teacher participants (p=.319). According to the findings, 57.5% of the participants (students=66%) believe that giving WCF on content or ideas is important for students because it makes students feel motivated. Moreover, 7.5% of the participants (students=3%; teachers=40%) argued that giving WCF on content or ideas helps content improvement. In addition, 5% of the participants (students=3%; teachers=20%) think that it makes students know whether their content is correct or incorrect. Additionally, 2.5% of the participants (students=3%) stated that students appreciate any comment on content. Furthermore, the other 2.5% of the participants (students=3%) believe that WCF on content or ideas is always necessary so teachers should do it. Moreover, the other 2.5% of the participants (students=3%) think that WCF on content or ideas is okay as long as it does not judge the idea wrong. Likewise, the other 2.5% of the participants (students=3%) pointed out that WCF on content or ideas is okay as long as it is positive for students. Finally, the other 2.5% of the participants (students=3%) stated that WCF on content or ideas makes students learn to receive criticism on their writing content. In contrast, 7.5% of the participants (students=6%; teachers=20%) believe that WCF on content or ideas is not a priority because other writing errors should be complementary to the feedback. Additionally, 5% of the participants (students=6%) argued that WCF on content or ideas will not affect students' writing. Finally, another 5% of the participants (students=3%; teachers=20%) pinpoint the importance of focusing on linguistics errors rather than focusing on content or ideas.
The test for WCF on Punctuation Errors displayed that there is a significant difference between student participants and teacher participants (p=.041). Based on the findings, 37.5% of the participants (students=37%; teachers=40%) believe that by giving WCF on punctuation errors, students can use correct punctuation. In addition, 27.5% of the participants (students=26%; teachers=40%) explained that punctuation is important for the quality of writing so teachers should give WCF on it. Moreover, 20% of the participants (students=23%) think that punctuation gives clarity to the writing so it is important for students to receive teachers' WCF on punctuation errors. Furthermore, 7.5% of the participants (students=3%) argued that students tend to forget the use of punctuation so teachers should remind them by giving WCF on punctuation errors. At the same time, 7.5% of the participants (students=6%; teachers=20%) stated that giving WCF on punctuation errors is not important. They believe that punctuation errors are not a big problem so teachers do not have to prioritize it.
The test for WCF on Spelling Errors displayed that there is no significant difference between student participants and teacher participants (p=.150). According to the results, 55% of the participants (students=54%; teachers=40%) think that giving WCF on spelling errors is important because spelling errors can lead to misunderstanding. In addition, 40% of the participants (students=40%; teachers=20%) agreed as well because with the WCF provision students can use correct spellings. Moreover, 2.5% of the participants (students=3%) think similarly because with WCF on spelling errors, students will not repeat the same error. In contrast, 2.5% of the participants (students=3%) think differently that giving WCF on spelling errors is not important. They argued that spelling errors are not a big problem as long as the reader understands the meaning.
The test for WCF on Vocabulary Errors displayed that there is no significant difference between student participants and teacher participants (p=.319). 72.5% of the participants (students=77%; teachers=40%) believe that giving WCF on vocabulary errors is important because it makes students know the correct and suitable vocabulary to use in their current written work. In addition, 17.5% of the participants (students=11%; teachers=60%) think the same because it helps students to know more vocabulary. Moreover, 5% of the participants (students=6%) argued that vocabulary is simply important so teachers should provide WCF on vocabulary errors. In contrast, 5% of the participants (students=6%) do not agree if WCF on vocabulary errors is important. They explained that vocabulary error is not a big problem so teachers should not focus too much on it.

Useful Amount of WCF Types
The majority of the students prefer that teachers should provide WCF on all errors that are found on students' written text. Therefore, for the majority of the students, they think that the greater the amount of WCF given by the teacher, then the more valuable it is. Similar to students, several teachers also believe that they should give WCF on all errors, but some of them also consider providing WCF only on most of the major errors. Overall, teachers' opinions are similar to the majority of students' perceptions and both of them demonstrated that the students and the teachers prefer WCF provision on large quantities of errors. The findings are consistent with that of Sayyar and Zamanian (2015), which found that most students and teachers prefer comprehensive correction on students' writings.
When both participants were asked whether teachers should mark students' reoccurring errors, the majority of them consider that errors should be corrected although they occur again. One of the students (S4) stated, "Yes, so students will get reminded with the reoccurring errors." Furthermore, there is a teacher (T5) that argued, "I will surely correct the same errors that students make. Students may not understand the first correction so the second or third is simply required." This preference is in line with the research conducted by Amrhein and Nassaji (2010), which demonstrates that both students and teachers value consistency.

Useful Types of Feedback
The second research question asked the types of WCF that students and teachers believe to be the most helpful. The students argued that they appreciate any kind of feedback even though they prefer receiving more linguistics feedback rather than personal comment on content or ideas. The finding is in line with that of Sayyar and Zamanian (2015), which found that students expected all types of linguistics errors to be corrected by teachers and they saw problems in the content are not too important. The students in the present study tend to choose to have overt correction with a comment because they claimed that they will receive better understandings. One of the students (S16) explained, "If the teacher told students the location of the errors as well as the explanation, students will understand better." Similar to students' preferences, teachers also agreed that error correction with explanatory is the best. One of the teachers (T3) explained, "I like this type of WCF. It gives corrections as well as the explanations, so students will completely understand their errors. However, it is too time-consuming." This finding is consistent with that of Amrhein and Nassaji (2010) that to minimize time-consumption, the teachers focus only on correcting the important errors.

Types of Errors that should be corrected
The third research question asked what types of errors students and teachers think should be corrected. Both students and teachers showed overall positive opinions about the usefulness of WCF on certain errors such as organization errors, grammar errors, punctuation errors, spelling errors, and vocabulary errors. However, the students saw WCF on content or ideas to be the least of their interest. These findings partly support the research done by Halimi (2008), which found that students tended to value teachers' WCF provision on surface-level errors (grammar, spelling, punctuation, vocabulary choice) than correction of other surface-level errors (content).
For organization errors, most students believed that it is important to receive WCF on this type of errors because it helps students to know a good and correct writing organization. One of the students (S21) argued, "It improves students' writing skill especially in paragraph writing." This finding is not in line with that of Diab (2006), which found that there were slightly fewer students that consider organization in writing. Similarly, one of the teachers (T1) stated, "As long as it gives good impact to students' writing quality, teachers should correct students' written errors. Thus, giving feedback on organization errors is simply necessary." For grammar errors, the majority of the students argued that correction on grammar errors is important because it helps students to know the correct grammar. The student (S21) stated, "It helps students in understanding grammar such as word order and sentence structure." This finding confirms that of Jodaie (2011), which showed that a majority of students expected their teachers to correct grammatical errors on every draft. Most teachers also believed that grammar is important so students need to receive WCF on it. One of the teachers (T2) argued, "Correction on grammar is very important. Let's say that I instructed the students to make a recount text. Then, they made errors on their writing regarding the tenses. It will affect the quality of the writing. It needs to be corrected." For WCF on content or ideas, most students believed that it simply makes students feel motivated. One of the students (S25) stated, "It is useful. Positive comments improve students' motivation." In addition, the teachers explained that students should not only concern improvement on form-focused errors but also the content or ideas as well. One of the teachers (T5) stated, "Emphasizing feedback on content or idea is important so students' writing idea quality will improve." Interestingly, this finding contradicts with that of Hartshorn et al., (2010), which suggested that ...utilizing WCF in many ESL writing contexts is overwhelming for both the teacher and the student. Providing quality feedback can be time-consuming for the teacher, and the tasks of processing and implementing large amounts of feedback can be unrealistic for the student. (p. 86).
It indicates that providing WCF on all kinds of errors requires much effort yet time-consuming. Thus, it is considered to be inefficient and unrealistic practice.
For punctuation errors, the majority of the students argued that WCF on punctuation errors is needed so students can use correct punctuation. One of the students (S32) stated, "Punctuation is important because it affects written text quality." This finding is consistent with that of Seker and Dincer (2014), which found that students preferred to receive feedback for both content and form, including the punctuation errors. Most of the teachers also believed that punctuation is important for the quality of writing. One of the teachers (T4) explained, "It is clearly important. Imagine that there are a lot of punctuation errors on students' written work. Of course, it will degrade the quality of the writing. Teachers should correct this kind of errors." For spelling errors, the majority of the students believed that spelling errors can lead to misunderstanding; thus, correction on spelling errors is important. One of the students (S17) stated, "Incorrect spelling can lead to misunderstanding for the readers. Thus, teacher's feedback on spelling errors will be useful." This finding is in line with that of Amrhein and Nassaji (2010), which displayed the same students' preferences on the importance of correcting form-focused written errors such as spelling.
Most of the teachers also believed that spelling errors can lead to misunderstanding. The teacher (T4) stated, "Spelling errors are common in students' writing. Small mistake on a letter can even confuse the reader. Hence, it is clear that spelling errors need to be considered." Lastly, for vocabulary errors, the majority of the students believed that correcting vocabulary errors is necessary so that students are able to know the correct and suitable vocabulary in their writing. One of the students (S10) saw the importance of correcting vocabulary errors because, "It helps us to know the words that we are supposed to use or the words that suit the context." In addition, most of the teachers agreed that vocabulary errors are important to be noted. One of the teachers (T3) stated, "Vocabulary is one of the most important aspects in writing. If the students incorrectly use certain word, then the meaning can be different with the one that the students want to express. Thus, vocabulary mastery is needed. Whenever I notice a vocabulary error, I always provide alternative words that suit the context. Therefore, the students can learn new words from there." This finding is in line with that of Iswandari (2016), which found that vocabulary errors are believed to be useful to correct.

CONCLUSION
Both students and teachers showed similar views about the useful amount of WCF and correction for repeated errors. Most of the students preferred to receive large quantities of feedback. The majority of the students also preferred to receive correction on repeated errors. The teachers had similar beliefs that they mostly chose comprehensive feedback. The majority of the teachers also chose to provide corrections on repeated errors.
Both students and teachers also share similar perspectives about the useful types of WCF. Most of the students expect to receive comprehensive feedback which includes correction and explanations. The teachers also agreed that comprehensive feedback is the most useful form of feedback; however, they consider it as time-consuming.
Ultimately, both students and teachers also have similar opinions about the error types that should receive correction. The majority of the participants stated that personal comment on content or ideas is not a big problem, so teachers should not put much effort into providing this kind of feedback. Instead, the findings demonstrated that teachers should focus on linguistics errors especially grammar, punctuation, spelling, vocabulary, and organization.