The Effectiveness of Indirect Written Corrective Feedback as Perceived By Teachers and Students of a Public University in Vietnam

The current study investigates the effectiveness of indirect written corrective feedback (WCF) on five different aspects of writing (grammar, language use, mechanic use, content and organization) as perceived by teachers and English-major sophomores in a Vietnamese public university. Specifically, it provides an insight into the effectiveness of this feedback pattern as perceived by the teachers and their students. To fulfill the stated aims, this research utilizes two main instruments, namely questionnaire and in-depth interview. The data from the questionnaire was analysed using statistical procedures. Meanwhile, the data from the interviews was processed using qualitative analysis. With regards to the findings, teachers and students agree that given feedback suits students’ understandability, but somewhat exceeds their self-correction ability. Teachers and students’ perceptions match on the efficacy of indirect WCF for the treatment of grammatical errors and its inefficacy for the betterment of content. Both parties also remain neutral about the correction efficiency of this feedback pattern regarding lexical and mechanical errors. Finally, organization is the aspect on which the perceptions between two sides mismatch the most. Teachers and students also share several reasons in accounting for their perceptions. The findings implicate that changes should be made to feedback-giving practices of teachers and feedback-handling practices of students to enhance the effectiveness of indirect WCF.


INTRODUCTION
Writing has been long known to play the central role in second language acquisition process.It has been widely approved that learning writing not only serves to shape L2 leaners language ability (Ismail, 2011), but also enriches their content knowledge through the process of searching and organizing information to write.However, writing is probably the most challenging skill to master in L2 learning.Learning how to write in L2 is a long process requiring continuous practice and meta-cognitive strategies (Myles, 2002).For those aforementioned reasons, writing instruction has always been the focal point of attention in Second Language Acquisition.In light of that, written corrective feedback (WCF) emerges as a pedagogical approach widely recognized by L2 teachers and SLA experts for its role in fostering students' writing ability (Nakaruma, 2016).
Realizing the necessity of WCF, L2 teachers has put great endeavor into finding how to employ feedback to enhance its efficiency.The two most common approaches applied by many L2 teachers are direct written correction and indirect written correction.While researchers, such as Hosseiny (2014), are in support of the former, others including Lalande (1982) are in favor of the latter.Supporters of im-plicit correction argue that using indirect WCF will be beneficial to students' long-term writing development.Since students are engaged in self-correction, they will remember the mistake and avoid it in the new writing tasks.On the other hand, indirect WCF is criticized because it can present learners with challenges in figuring out the correction themselves.
Apart fromt that heated debate, another burning question to ponder is whether perceptions of teachers and students of each type of WCF should match.Because students are the users of feedback, it is also necessary for their perceptions of the feedback to be considered.Many L2 experts like Ferris (1999) strongly believe that learners are more likely to use a certain type of feedback, if they feel it is effective for them and vice versa.Therefore, it is necessary for teachers and students to reach an agreement of what works best for both sides.This implies the need for studies investigating teachers and students' opinion of the effectiveness of certain types of WCF.
In the context of the studied university, indirect WCF is a strategy found to be widely employed by many teachers for students with high-level proficiency like second-year English majors in Department of English.However, the available body of research placing focus on this type of feedback is very limited.Plus, hardly has any research been conducted to investigate teachers and students' perceptions

The Effectiveness of Indirect Written Corrective Feedback as Perceived by Teachers and Students of a Public University in Vietnam 153
of effectiveness of WCF.All these gaps provide a chance for this study to take place.Specifically, it provides an insight into the teachers' practices of using this kind of feedback in marking students' writing.Afterwards, teachers and students' perceptions of its effectiveness on five aspects, namely grammar, language use, mechanic use, content and organization are looked into, analyzed and compared with each other.With aforementioned aims and objectives, this study will focus on addressing two research questions: 1.What is the effectiveness of indirect WCF as perceived by teachers of second year mainstream program? 2. What is the effectiveness of indirect WCF as perceived by students of second year mainstream program?Once completed, this research will inform teachers of how students perceive the effectiveness of indirect WCF and vice versa.This enables teachers and students to come to an agreement of what should be done to enhance the effectiveness of indirect WCF.On a larger scale, the study is expected to come out with practical solutions and implications to improve feedback-giving practices of teachers and feedback-handling practices of students, with a view to contributing to the ultimate goal of fostering students' academic writing performance.

Writing skills in L2 learning
Writing has always been playing an important role in second language acquisition.This role is actually derived from "the fact that it reinforces grammatical structures, vocabulary and idioms" or the language that teachers have been teaching their students (Ismail, 2011).Some studies emphasize that students' learning progress are boosted because of their having to make use of "meta-cognitive and self-regulation activities" in order to convey their meaning through writing (Hurley & Wilkinson, 2004).Moreover, other research highlights the importance of writing in that it broadens students' knowledge and learning opportunities since they need to look for and organize information.This allows students to "sharpen their understanding and flourish their knowledge" (Nguyen, 2016).
In second language acquisition, writing is seen as the most difficult skill to master.As Nakaruma (2016) cites from many studies, writing is "a complex task which requires specialized skills" and huge amounts of "cognitive and linguistic" resources.The ability to write well is not naturally acquired, which has to be practiced and allowed adequate time for (Myles, 2002).Although writing has been "accepted as the common goals of learning English", it is often the skill that "falls through" (Zen, 2005).It is not rare to see L2 learners struggling when they have to communicate in the target language using written mode (Zen, 2005).In the light of this, teaching L2 writing has received much attention in recent years.One of the "pedagogical techniques", which has been widely recognized by both L2 teachers and SLA researchers, is written corrective feedback (WCF) (Nakaruma, 2016).Jacob (1981) asserts that there are five main aspects naming content of writing, organization, language use, mechanic use, and vocabulary.

Aspects of writing
First, content aspect refers to the presentation of knowledge, "substantive", relevance of the ideas to the topic, and evidence in support of the ideas (Jacob, 1981).
Second, organization refers to the presentation of ideas in a clear, logical, and way (Jacob, 1981).
Mechanic use lays emphasis on spelling and appropriate punctuation, citation and referencing in the text (Jacob, 1981).
Lastly, language use refers to the use of language, choice of vocabulary, expressions and appropriateness of words to convey the wanted messages (Brown, 2007).
The study will base the model of Jacob (1981).These aspects mentioned above will provide the researcher the basis to form the criteria for assessing the effectiveness of indirect WCF and answer the research questions.

Definition of written corrective feedback
It is widely believed that corrective feedback is one of the central aspects of L2 writing development.Kepner (1991) even addresses that WCF "is of perennial concerns to L2 teachers" (as cited in Sun, 2013).On the whole, most dominant researchers in the field agree on the definition of WCF as a reminder to learners that their "use of target language is incorrect" and "reinforcement of error correction" (Ellis et al, 2008;Beuningen, 2010;Ellis, 2009).According to Sun (2013), written corrective feedback can come from any source such as "a random reader of the composition", the writer's peers or the teacher.However, in L2 classrooms, it is the instructor or teacher that has key responsibilities in providing WCF for L2 learners (Sun, 2013).Moreover, numerous research has shown that learners also "want, expect and value" WCF and would like to receive it more than other kinds of written feedback such as oral feedback and peer feedback (Corpuz, 2011).
Since WCF became a widely explored topic in the research world, different researchers have proposed various approaches to categorize different types of written corrective feedback.However, the most common approach used by many researchers is through the implicitness and explicitness of feedback, evidenced by the model of Sheen (2011): • Direct non-metalinguistic written correction Students are provided with the correct form, by e.g., crossing out the error and replacing it with the correct word or adding something that is missing.

• Direct metalinguistic written correction
Students are provided with the correct form and giving a written explanation of some sort, for instance by numbering the errors and giving the answer with an accompanying explanation at the end of the page.
IJELS 6(4):152-162 • Indirect written correction (non-located error) Students are provided with an indication that an error has occurred but not locating or correcting it, these indicators appear only in the margin.

• Indirect written correction (located error)
This type is rather more obvious because the location of the mistakes is identified.The teacher indicates where students commit the errors by underlining, circling the errors or putting a cross to show omission.• Indirect written correction using error codes This type of feedback includes the underlining or circling the errors, plus coded feedback to show the "nature" of the error (e.g., "sp" for spelling or "w.c" for wrong word choice).

• Indirect metalinguistic written correction
This type of feedback also is presented as numbering the errors and giving a metalinguistic explanation of the error, but in the form of a question to withhold the correct form.E.g: In what form should the verb be in a passive sentence?

Significance of teacher written corrective feedback in developing L2 learners' writing
On the basis of findings from vast numbers of past studies the role of written feedback as a "useful editing tool" has been confirmed.To name but a few, Chandler (2003) and Ferris (2006) come out with evidence that those who actually received feedback showed greater improvement in accuracy in the subsequent writings than those who did not.However, it is still controversial whether written corrective feedback can result in long-term development of L2 writing and acquisition.Very few studies have gone beyond the role of feedback as a "facilitator of learning" by studying new writing tasks.One typical example of this is by Ferris (2006), which reports a reduction in the numbers of both grammatical and lexical errors L2 learners made over a semester.Ellis et al. (2008) and Sheen (2007) also have also indicated the improvement in terms of accuracy of L2 learners in the new writing entries.Although it is too early to come to conclusion that WCF leads to long-term writing development, Ferris (2004) did point out evidence that predict the positive effects of WCF in fostering L2 acquisition.

Requirements of teacher written corrective feedback
Most research to date has all agreed that teacher should take students' preferences and attitude into consideration when giving written comment.Just like Hyland & Hyland (2006) assert, it is necessary for teachers to "weigh their choice of comments to accomplish informational, pedagogic, and interpersonal goals simultaneously while taking account of likely student reactions".Regarding the students' positive view of effectiveness of feedback, some studies have found that teacher's WCF has also played an important role in not only developing L2 writing, but also in helping students to "to identify problems and giving them information about academic and disciplinary expectations" (Leki, 1991; as cited in Hyland & Hyland, 2006).In some other studies, it is found that students are even "positive" about the use of indirect written corrective feedback in which they are provided with clues for corrections, as it can turn them into "active" users of feedback (Hyland and Hyland, 2006).
On the other hand, evidence from research also suggests students' lack of understanding of the meaning of much feedback given on their writing papers and have no idea of "what they are expected to do with WCF" (Amrhein &Nassaji, 2010).For example, Ferris (1999) and Hyland (1998) prove in their studies that students had difficulty in understanding the feedback and the way they handle the feedback was out of sync with the teachers' intentions.This is actually originated from the conflict between teachers and students regarding perceptions of effectiveness of various kinds of feedback.Amrhein and Nassaji (2010) suggest this conflict is rather "pedagogically problematic".As pointed out by many studies, if students perceive a kind of feedback does not work for them, they are not likely to pay attention to it.Therefore, it is necessary for teachers and students to come to an agreement of what works the best for both sides (Amrhein & Nassaji, 2010).
As the main source of feedback, teachers must guarantee the understandability of the feedback and students are "susceptible of doing anything with it".It is also required that teacher must be "congruous with the feedback" and "accommodate to students' proficiency and competency to self-repair" (Salteh & Sadeghi, 2012).

Definition of indirect written corrective feedback
Broadly, implicit/indirect written feedback can be seen as the type of feedback that point outs to learners that they have made a mistake, without providing the correct form.In other words, Sun (2013) points out that all types of indirect WCF share a common feature which is "withholding the correct form in the hope of eliciting the correct form" or self-correction from the students.
Based on Sheen (2011)'s typology, indirect written corrective feedback consists of 4 types: indirect WCF (located error), indirect written corrective feedback (non-located error), indirect WCF using error codes, indirect meta-linguistic written correction (the description of each type is presented in the model).Bitchener (2008) adds that indirect WCF is also presented as "confirmation checks" or "request for clarification" From the researcher's viewpoint, all the definitions given by aforementioned researchers are considered as different patterns of "indirect WCF" as they all share one common feature of withholding the correction to trigger students' self-correction.
However, from the study of some of students' writing, the researcher also finds that the indirect WCF which names the nature of the errors is also considered as indirect WCF.This pattern is only different from the one using error code in that it names directly the nature of the errors, because there is no code or the teacher prefers to name directly the type of error.It also corresponds to the common feature of indirect WCF, which is withholding the correct use of the target language to provide opportunities for students' self-correction.

The Effectiveness of Indirect Written Corrective Feedback as Perceived by Teachers and Students of a Public University in Vietnam 155
For example: Parallelism => T implicitly tells students that writing has problems with parallelism at this place.
Given all the definitions in consideration, it can be concluded that indirect written corrective feedback can take any form ranging from "underlining or circling the error", "using cursors to show omissions", "indication at the margin", "written correction using error codes" "metalinguistic written correction", "confirmation checks", "request for clarification", comments showing the nature of the errors.This definition also provides the theoretical basis for the research when conducting the actual study.

The effectiveness of indirect WCF in debate
In fact, the effectiveness of indirect WCF has always been at the center of heated debates.While many researchers are in support of its usefulness, indirect WCF is also subjected to criticism for its correction-withdrawing nature.For the proposition side, Lalande (1982) argues that indirect WCF is more beneficial in fostering learners' long-term development of writing.The significance of indirect WCF is argued based on its ability to "involve learners in guided problem-solving" and encourage them to be more responsible for their progress (Wang & Jiang, 2015).Besides, indirect WCF proves superior to direct WCF when it engages learners "in a more profound form of language processing by promoting self-reflection, attention, and noticing" (Wang & Jiang, 2015).Simultaneously, the use of indirect WCF is criticized because it requires much effort from learners.Moreover, it may fail to help learners "resolve complex errors" and learners do not know "if their own hypothesized corrections are indeed accurate" (Hosseiny, 2014).
Another issue which should not be ignored is whether indirect should be used to correct different kinds of errors.Ferris (1999) categorizes errors into two "treatable" and "untreatable" errors.She argues that indirect WCF is more suitable for the treatment of "treatable" errors including "verbs, subject-verb agreement, run-ons, fragments, noun endings, articles, pronouns, and possibly spelling", which "occur in a patterned, rule-governed way" (as cited in Hyland & Hyland, 2006).On the contrary, it may not be effective for the correction of errors related to word choice or language use as there is no "set of rules that students can consult to avoid or fix these kinds of errors" (Ferris, 1999).

The Writing Assignment
According to the syllabus of the course chosen, there are two writing assignments focusing on the two main types of essays.Once the topic is released, students have to look for the reading materials related to the topic.After completing the first draft, students will have to submit the first draft to their teachers to get their comment.The teachers will give comments directly into the draft and give it back to students.Students will base on the feedback to make revision and write the second draft.
Unlike normal exam essay which lays most emphasis on language use and accuracy, the essay assignment of the course also lays emphasis on content and organization.For content, much emphasis is laid on the quality of ideas and research-based evidence to support for the ideas.Meanwhile, organization of ideas and coherent link between the ideas are the central aspects of organization.As students have to use sources from the internet as supporting materials, they have to include in-text citation and referencing.Therefore, the focus of mechanic use will be on citation and referencing.Mistakes like punctuation and spelling are not focused on since they can be checked by computer.

METHOD Participants
The participants of this study include three teachers, namely A, B, C, who were in charge of teaching the chosen course to three different classes.All three of them are master degree holders with profound expertise in teaching written proficiency courses.In the pre-research investigation of students' writing pieces, those teachers are found to make dominant use of indirect written corrective feedback in marking their students' writing.While teacher A and teacher C use a combination of both explicit and implicit WCF, teacher B only employs indirect written correction, which is much of a surprise to the researcher.Through the pre-research study of students writing and informal interviews with the teachers, there was no discrepancy regarding their "believed" and "actual" practices of giving WCF, which was observed in the past studies including that of Corpuz (2011).
There are also 90 second-year students from 6 classes in Department of English.Their expected level of proficiency is around B2/B2+ (according to CEFR).They are from the classes whose the chosen course was taught by the aforementioned three teachers.In the first semester of their second-year program, they had to do writing assignment which was selected to be the subject of this study.That semester also marked the first time they were exposed to research essay, citation, reference.Therefore, it was likely that students encounter loads of difficulty acquainting themselves with the new type of essay.

Questionnaire
Questionnaire is widely recognized as a method which allows researchers to collect a huge amount of data in a short period of time.Given the researcher's constraints in terms of time, financial resources as well as huge number of participants, questionnaire proves to be the most appropriate and economical instrument.The questionnaires are distributed to students to enquire about their perceptions of the effectiveness of indirect written corrective feedback, and the difficulties they encounter in handling this kind of feedback.
Following the standard procedure of designing a questionnaire with high level of validity and reliability, the researcher thoroughly examines the literature review.This helps to identify key themes based on which the items are constructed.After the items are written and scrutinized for relevance, they will be sequenced into an appropriate order.The questionnaire consists of 2 parts: IJELS 6(4):152-162 • Part (1): Before students actually work on the survey, they have to read a sheet containing the definition of indirect WCF and the focus of different aspects of writing.Several of their actual writing excerpts with comments by their teacher were printed in the sheet, so that students can have better understanding of indirect WCF and the subject of the study • Part (2) presents questions to ask for students' perceptions of the effectiveness of indirect WCF.There will be both close-ended questions (including Likert scale and multiple ticking questions) to save students' time and open-ended questions so that students' response will not be limited.The data will be collected for analysis to make generalizations and interpretations.Besides, the researcher also inquires students' personal information including name and phone number with a view to reaching them later for the interview.
To ensure the content validity of the questionnaire, the researcher seeks for judgment and feedback from two experts who have considerable experience in teaching writing and questionnaire design.The painstaking expert investigation culminates in several items being omitted due to irrelevance and some other being revised for word redundancy.The wording of the questionnaire is also modified and simplified in avoidance of any misunderstanding which may adversely affect the findings when the participants attempt to answer the questions.
The piloting of the revised the questionnaire is conducted with 15 of the informants to determine the reliability of the questionnaire and the appropriateness of the items.The results come out with acceptable reliability and no further feedback regarding the items.All the participants in the piloting study state that there is no problem about the comprehensibility of the questions, since the language is well below their level of proficiency.This also eases the need for translation.

Semi-structure interview
According to Le (2012), interview enables the researcher to "investigate phenomena that are not directly observable such a perceptions and attitudes".The current research aims at investigating perceptions of teachers and students of indirect WCF, so the chosen instrument proves a suitable one.
As the interviews are semi-structured, the researcher prepares a list of questions to ask in advance.However, there may also be questions asked on the spot to fit in with the situation.The literature review and questionnaire serves as theoretical validation for the interview questions, which are subsequently sent to experts for review.In response to expert judgment, wording of the questions is modified for the sake of comprehensibility, but no further reversion and redundancy is needed as they are judged to be relevant and suitable.The relevance of the questions is also confirmed through the piloted interviews put on before the actual ones.All the participants are more comfortable with the interviews being conducted in Vietnamese, so the questions are translated for their sake.The translated questions are approved by a professional translator before putting to use.
The semi-structured interviews are conducted to gain insights into how teachers perceive the effectiveness of indi-rect WCF in helping students improve writing skills.Before the interview, the researcher presents the teachers with the definition of indirect WCF with its patterns and examples to demonstrate these patterns.The examples are actually taken from the indirect WCF they gave in their students' writing.This allows the teachers to have deep understanding of the definition of indirect WCF and give the most judicious perceptions.Next, the researcher explains to the participants about the procedures, the purposes and answers any confusing questions from them so that the interview will run smoothly.
There are also interviews with students in order to gain deeper insights into their response of the questionnaire.These interviews will be conducted after the researcher has collected the questionnaire and analyzed the data.The researcher selects the interviewees whose responses match the concern of the research.Before the interview, the students will have a peek of the explanatory sheets and questionnaires completed by them.Afterwards, the researcher asks students some questions to gain in-depth insight into students' perceptions of the effectiveness of indirect written corrective feedback and the difficulties they encounter while handling this type of feedback.

Analysis of data from questionnaire
Once collected from the students, the data will be converted to statistics which the researcher will base on to draw tables and charts.Afterwards, the thorough analysis of the figures will be conducted to describe, associate, and make implication and interpretation about the effectiveness of indirect WCF as perceived by students.

Analysis of the data from interview
Since the interviews are conducted in Vietnamese, they have to be translated after being transcribed.The translated transcripts are then examined and finalized with the help of a translation expert.This text-based data is then stored, controlled and brought out in a more organized way with the aid of NVIVO software.However, all of these processes are put under the monitor of the researcher in avoidance of any unexpected error.The data from the interview with the students will be used to explain and reinforce the findings from questionnaire instrument.Meanwhile, the data from the interview with teachers will be synthesized and collated to make comparisons between the views of the teachers.The responses of both parties will also be compared to see to what extent their perceptions meet and differ from each other.

Answers to the First Question: What is the Effectiveness of Indirect Wcf as Perceived by Teachers of Second Year Mainstream Program?
Regarding the understandability of the feedback, all three teachers agreed that students could understand the feedback,

The Effectiveness of Indirect Written Corrective Feedback as Perceived by Teachers and Students of a Public University in Vietnam 157
but do not know how to correct the mistakes based on the feedback.Teacher B believed that it is the learner autonomy that decided the effectiveness of the indirect WCF."The indirect WCF I gave was not quite effective, but that was students' fault not to invest time and effort into the correction".Meanwhile, teacher A and teacher C seemed to have more sympathetic views for students' difficulties when making the correction of the mistakes.This somehow affects their perceptions of indirect written corrective feedback.

Effectiveness of indirect WCF on grammatical mistakes as perceived by teachers
In terms of grammar, teacher A believed that the effectiveness of indirect WCF was the most significant on grammar aspect.Adding to her point, teacher A explained that most of the grammatical mistakes found on her students' essays are mostly "slips caused by their carelessness".As a result, she just needed to "underline or circle" them, students would understand and correct it without any difficulty.Teacher B and C also agreed that indirect WCF enabled students to make improvement in terms of grammar.However, "the improvement is not significant as expected".To explain for this unexpected belief, teacher B added that "students seemed lazy and did not pay much attention to the feedback.They only use the check by computer, or just revise easy or obvious mistakes".Meanwhile, they did not spend time correcting more errors related to "subject-verb agreement", because these cannot be checked by computer or "recognized at first sight".Giving a different explanation to her response, teacher C supposed "Students were not flexible in their correction.For example, they may give wrong verb form and I pointed this out for them.However, if they changed the sentence structure or I asked them to, they may forget to put it in the right form again".

Effectiveness of indirect WCF on lexical mistakes as perceived by teachers
However, when it comes to the effectiveness of indirect WCF on lexical aspect, three teachers seemed confused.All teachers perceived that the errors on this aspect, which were mostly related to word choice and expression, were not revised effectively.To support for the claim, teacher A and teacher C explained that: "For errors related to word choice, most students find the new word, but the new word is still wrong".Meanwhile, teacher B disappointedly shared that students did not even bother replace the old word, "even if I mark it as W.C (word choice).It seemed that they saw these as superficial errors and do not pay much attention to it".

Effectiveness of indirect WCF on mistakes related to mechanic use as perceived by teachers
Regarding mechanic use, all three teachers agree that indirect WCF was not effective enough in helping students correct these errors.Teacher B perceives that students still cannot correct these errors, even if "I had given them detailed instruction in class and even pointed out for them".Therefore, it is still "the students to take the blame".On the contrary, teacher A and teacher C seemed to have a more sympathetic view.Teacher A associated this with many elements and requirements of APA citation and referencing.Hence, "students had never made perfect revision.For example, I underlined the mistake and wrote APA style, they managed to correct the author's name, but then forgot to give italic form for the title of the research in the same reference".Having the same opinion, teacher C added that there were examples in the textbook but examples are not representative enough for the numerous types and requirements of citation and referencing.Also, because of time limitation, teacher could just give implicit feedback like "underlining" or simply "APA style".This somehow explains why students have trouble correcting these errors.

Effectiveness of indirect WCF on the betterment of content as perceived by teachers
As for content, all teachers agreed that this was the aspect where students make the least improvement.From the essays, it is easily found that the main problems were mainly associated with students' lack of evidence from research and explanation for their arguments.All the teachers revealed they often gave indirect WCF by asking their students to give evidence or research grounds, but most students did not manage to.Each teacher had different explanations regarding this."Some students managed to give more grounds, but not satisfactorily", added by teacher A. "Many of them gave personal example or explanation, even if I pointed out that it needed to be from research".In the same position, teacher B supposed it would take time for students themselves to improve on this aspect, the revision process was too short for them to make noticeable improvement.On the other hand, the feedback teacher could give on this aspect can only be open-ended and implicit "due to the limit of paper".It is the students to gradually figure out their problems and improve by themselves.Meanwhile, teacher C attributed this to the availability of evidence and lack of suggestion on how to find the evidence.Specificially, this was because research articles on the internet often required payment or students did not have any suggestion on how to find the evidence.

Effectiveness of indirect WCF on the revision of organization
Surprisingly, teacher B and teacher C saw organization as the aspect on which students made the most improvement based on indirect WCF.The two teachers associated the arrangement of ideas and clarity of thesis statement as the main problems with organization in students' writing.Teacher B explained "I always ask students to follow the fixed organization for each essay in the textbook with topic sentence at the beginning of each paragraph.Therefore, when students failed to do so, I just needed to circle the sentence and wrote "topic sentence".They would understand my point and manage to put the sentence to the front".Likewise, teacher C said she always wanted students to be specific in thesis statement."When students give a general thesis like: IJELS 6(4):152-162 "this essay will outline the reasons…I would circle and put a question mark, students would understand and specify what the reasons are in the next draft".In contrast, teacher A perceived that students do not make any noticeable improve on this aspect in the next drafts.This difference in perceptions among the teachers is quite explainable.While teacher A associated the main problem related to organization as coherence between ideas, teacher B and C attribute it to arrangement of ideas and thesis statement.
In conclusion, it seems that the teachers were able to guarantee the understandability of feedback, but unable to adapt it to the level of self-repair of students.However, through the interview, many factors are found to affect the effectiveness of indirect WCF.For content and mechanic use, it is nearly impossible for teachers give direct correction due to the limit of paper, the requirement of the evidence and the numerous types and elements of referencing and requirement.Therefore, it is not the teachers to put the blame.The only conclusion to reach is that indirect WCF alone was not effective enough for the treatment of these complex errors.

Effectiveness of indirect WCF during as perceived by students of second year mainstream program
A look at the Table 1 reveals that indirect WCF enables students to make the greatest improvement in grammar with 59.3 % and 23.4% taking "agree" and "strongly agree" respectively.
The effectiveness of indirect WCF in helping students to revise errors is quite equal for the two aspects: mechanic use and language use.Regarding language use, 45.7% of students agreed on its effectiveness, while 42 % remained neutral.Roughly the same numbers of participants (48.2% & 42%) choose "neutral" and "agree" options respectively when it comes to the effectiveness of indirect WCF on mechanic use.
Specially, content and organization seem to be the two aspects on which students benefited the least from indirect WCF.The majority of students (56.8% and 53.1%) were confused about the efficiency of indirect WCF on content and organization respectively.Regarding content, a remarkable of 25.9% disagreed that indirect WCF is effective.While 19.7% of students disagreed on the effectiveness of indirect WCF on organization, 16.1% were in agreement with it.
From Table 1, it can also be seen that the majority of students (over 60%) were in agreement that indirect WCF allows more freedom for correction and makes them more independent and responsible.This reinforces the findings of previous study that student have "positive attitude" towards to the use of indirect WCF by teachers.
To account for students' response towards effectiveness of indirect WCF in different aspects of writing, the researcher asked students to reflect on the difficulty they had when dealing with indirect WCF: It can be seen that the findings of the Table 1 quite correlate to those of Table 2 revealing students' evaluation of effectiveness of indirect WCF.Therefore, this implicates that students' perceptions of effectiveness of indirect WCF are influenced by the extent to which they have difficulties in finding out how to correct the errors.
For grammar aspect, it can be seen that most students (nearly 80%) take "rarely" and "never" option, implicating that most of them had no trouble finding out how to correct these mistakes.For language use and mechanic use, it can be seen that almost half of students took "sometimes" and the other half took "rarely" option.This explains why they were also divided between remaining "neutral" and agreeing with the effectiveness of indirect WCF on these two aspects.
Regarding the effectiveness of indirect WCF on language use, Quinn -the ninth interviewee said that: "Most mistakes I make related to language use is word choice.Luckily, the teacher pointed that out for me and I was able to select a more suitable word".Meanwhile, Bach-the fourth interviewee rather confused: "After the teacher gave feedback, I read again and find that there was something wrong with the word, but finding another word is not easy for me.Sometimes, I managed to find the new word; the new word may not be suitable again".Bach's concern can be associated with Ferris (1999)'s view mentioned in the literature review that errors related to word choice may not be corrected if marked by indirect WCF.This is because there is no "set of rules that students can consult to avoid or fix these kinds of errors" For errors related to grammar or mechanic use, there are rules to follow to decide accuracy and inaccuracy.However, it is still striking to find half of respondents "sometimes" have difficulty.Through the in-depth interviews with the students, the researcher found out that it was the numerous types and elements required in citation and referencing that caused the students to get confused.Hong -the second interviewee elaborated: "There were just few examples in the textbook, not enough to demonstrate for all types that we had to do in the essays.When giving correction, the teacher just simply underlined the mistake or wrote "APA style".Of course, I knew my referencing was wrong, but did not know how to correct since I didn't have the example of that type to follow".
Finally, errors related to content and organization prove the most complex ones, with the majority (over 80%) of students taking "sometimes" and "often" option.This accounts the reasons why the majority of students were confused or even disagreed with the effectiveness of indirect WCF on these two aspects.Through the interviews with students, all students admitted understanding these errors, but having trouble how to correct them.
For content, it seems that their main problem was lacking evidence or ground for their claims.However, they had different reasoning's for this issue.Data -the sixth interviewee explained: "Thanks to teacher's feedback, I recognized I lacked evidence but I couldn't find the evidence available from research to support my claim".Meanwhile, Giang -the fifth interviewee said: "I know I have to give evidence, so I expect the teachers to give me suggestion on what kind of evidence I should look for, but the teacher only asks for more evidence".

The Effectiveness of Indirect Written Corrective Feedback as Perceived by Teachers and Students of a Public University in Vietnam 159
For organization, most students associated their biggest problem with coherence, namely the link between ideas, especially between claim and evidence.In support of this idea, Phuong -the first interviewee elaborated: "Because of teacher feedback, I realized the evidence I gave was not connected to the claim I made; however, finding a more suitable one was not an easy task.Those that supported the claims I made was from unreliable source, which was not approved by the teacher".Meanwhile, Hein -the third interviewee revealed: "My teacher just left an implicit comment as "weak link".I could figure out the mistake after a while, but I did not know how to correct it.Therefore, I would like to be given suggestion".

Students' other difficulties in dealing with indirect WCF
From Figure 1, it can be seen that there was almost no student who did not experience difficulty.The most striking number of nearly 80% admitted their concern that their correction would not come up to teacher's expectation.However, this may be due to their lack of autonomy to ask for help from teacher.Regarding this issue, teacher C supposes: "Actually there were some students who came and ask me whether their correction is good or not.Therefore, it is up to whether students actively ask for help or not".
A remarkable percentage of over 60% admitted having to spend more time and effort to deal with indirect WCF.This is also one demerit of indirect WCF, but the teachers saw it as a strong point.Teacher B believed "because they have to find the correction themselves, they will remember the mistake and avoid it in the next essay".
Meanwhile, only one fourth of students expressed their fear that their correction may not be correct.This could be attributed to students' concern for the many elements and required in referencing and citation, which caused them to trouble, especially when teachers just give superficial feedback as APA style or underlining.
As shown in Figure 2 most students still paid attention to indirect WCF given by their teacher.Just over 5% percent of students decided to "ignore the feedback" or "delete the sentences containing the feedback".Self-correction attempt was the method preferred by most students (over 70%).Meanwhile, only one fourth of students chose to make correction in another way.
To the researcher's surprise, "asking for clarification/suggestion", which was expected to top the list, is only taken by over 35 % percent of students.This implicates students' lack of autonomy to ask for help from teachers.Through the interview, the teachers also showed concern for this: "It seems that students were passive, always waiting for teachers to come and help them", said teacher B. In agreement with teacher B, teacher C added: "Even if I gave them time to ask question, they would be too lazy to ask".

Similarities and Mismatches Between Teachers and Students' Perceptions
Through the answers of two research questions, the researcher can generalize the similarities and mismatches between perceptions of students and those of teachers regarding the effectiveness of indirect WCF.
Regarding the general perceptions, teachers and students meet each other in that indirect WCF will serve to enhance learner autonomy and give them more freedom in terms of correction.
Apparently, perceptions and teachers and students match on the efficacy of implicit correction in treating grammatical mistakes.Nearly 80% of students "agree" and even "strongly agree" that indirect WCF is effective for the correction of grammatical mistakes.Three teachers were also in agreement with this, even though teacher B and teacher C supposed the effectiveness was not "noticeable as expected".This is quite understandable since mistakes related to grammar are seen as "treatable", with clear border between accuracy and inaccuracy.Therefore, most students can easily recognize and correct those errors.
In terms of mechanic use and language use, while students were divided in half-half between "neutral" and "agree" options, three teachers were confused and even "negative" about the effectiveness of indirect WCF for treatment of errors related to those two aspects.For mechanic use, teacher C and students also met each other in the numerous elements, types and requirements of referencing and citation as the main reason that downgraded the effectiveness of indirect WCF.Specifically, the teachers just gave implicit comments like "APA style" or "simply underlining".Students could easily recognize, but did not know how to correct since they did have any example to follow.In regards to language use, both sides associated the common errors to word choice.Teacher A and teacher C also shared the same concern with students that they could find a new word to replace the old word, but the new word may be wrong again.
As for organization, while teacher B and teacher C believes this was the aspect in which students made the most improvement thanks to indirect WCF.However, teacher A and students perceived that indirect WCF was generally not effective for the revision of organization.This mismatch is actually derived from their perceptions of the dominant type of errors related to this aspect.While teacher B and teacher C attributed this to arrangement of ideas, thesis statement and topic sentence, students perceived errors related to coherence (linking in terms of meaning between ideas, between claims and evidence) as their main problem.
Unlike grammar, content is perceived by both teachers and students as the aspect in which indirect WCF was the least effective.Both teachers and students associated the main problems with content to students' lack of evidence and ground from research for their claim.Meanwhile, teachers' comments on this aspect were implicit like "any evidence?""Any evidence to support this claim?"Teacher C also shared the sympathetic view with students that the availability of evidence and students' lack of suggestion to find the evidence would be a barrier to students' betterment of content.That affects the effectiveness of indirect WCF.Nonetheless, because the requirement of the evidence is that it must be from research, it is impossible for teachers to provide direct correction on aspects like content and organization.This implicates the need for solution to enhance the effectiveness of indirect WCF on these two aspects.
Regarding other obstacles, the students had while dealing with indirect WCF, all three teachers attributed these to students' hesitation in asking for help.Furthermore, they all saw students having to spend more time dealing with indirect WCF as a plus, since students themselves would remember the mistakes and avoid them in the coming writing tasks.
The findings of this study did reinforce those of previous studies investigating the effectiveness of indirect written corrective feedback.Notably, students are positive about the employ of this feedback pattern on the grounds that it boosts their learner autonomy as well as self-independence, and allows them to have more freedom for error correction (Hyland & Hyland, 2006;Wang & Jiang;2015).It also responded to the growing concern of many L2 experts in the field as to whether students can correct their errors based on indirect WCF (Bitchener, 2012).The positive side was evidenced by that the majority of students and teachers agreed that indirect WCF was helpful in the treatment of grammatical errors or even in the acquisition of grammatical knowledge, as pointed out by Ferris et al. (2013).On the negative side, participants in the study confused or even disagreed on the efficacy of implicit feedback on the more complex errors related to language use, reference, content and organization.Even if correction was possible, students were uncertain whether their correction was accurate or met the expectation of their teachers.These findings really corresponded to the concern of Hosseiny (2014) as mentioned in the literature review.In light of available theoretical basis and studies in the same areas, this study also calls for teachers and students to take into consideration the learning context to decide on the best feedback-giving and feedback-handling practices (Hyland, 2003).All of these serve to bring about the positive impacts for students' academic writing performance.

CONCLUSION
Despite restriction in terms of time and resources, this study has achieved remarkable results.First and foremost, it has provided an insight into practices and purposes of teachers in employing indirect WCF to mark students' writing.Secondly, it has informed teachers of students' perceptions of effectiveness of indirect WCF and vice versa.Moreover, the mismatches and similarities in perceptions of each side have also been figured out.Hopefully, these mismatches will initiate teachers and students to come to agreement of what works best for them.On a larger scale, this study will contribute to the limited number of studies investigating both teachers and students' perceptions of written corrective feedback, especially in the context of the university.
As for the betterment of content and organization, students seemed to have difficulty with finding of evidence and needed suggestion from the teachers.However, it is nearly impossible for teachers to give direct correction on these aspects due to research-derived requirement of the evidence.This finding implicates that there is a need for face-to-face review session after students finish second draft, so that teachers can give extra oral suggestion on the treatment of these complex errors.This is also the suggestion of most students make when being asked for solutions to enhance the effectiveness of indirect WCF.
As for mechanic use, it is the numerous types and requirements, and elements of reference as well as citation that caused students difficulty in correcting these errors.This is also the feature that hinders teachers to give direct correction on paper.Therefore, it is reasonable that that there should be more in-class practices to familiarize students with citation and referencing (the course is the first time students encounter with these two concepts), so that the risk of making mistakes is decreased.
From the study, it is found that students were not active enough in asking teachers for explanation and suggestion of the correction of complex errors.While a face-to-face review section should be implemented, it is necessary for students to be more active not only in self-correction, but also in asking teachers for help in the correction of complex errors related to content and organization.It is also vital for teachers to be more understanding of struggles students have to go through when dealing with such a demanding writing project.

Figure 1 .Figure 2 .
Figure 1.Students' other difficulties in dealing with indirect WCF

Table 1 .
The effectiveness of indirect WCF given on different aspects as perceived by students

Table 2 .
Students' difficulties in correcting different kinds of errors marked by indirect WCF