Preferences and Problems in Feedback Techniques of Bangladeshi Tertiary Level Students

The study attempts to find out the preferences of students for different feedback strategies in writing, their views about the implementation and usefulness of the existing feedback practice, and their problems and expectations from it. Feedback in writing is a vital act for its development where the perception of the students is important. It is hypothesized that students have certain preferences, problems, and suggestions for the improvement of feedback. A group of 50 students of tertiary level from one institution who experiences the same feedback strategies were surveyed online for the research. The hypothesis is found to be strongly supported by the findings and discussion with a number of implications. The research concludes by indicating its benefits and giving directions for further study.


Introduction
Feedback is the general name for various procedures that are currently used in many L2 courses to react to students' writing and it is defined as a way to inform learners of whether an instructional performance is right or wrong (Lalande, 1982). According to Harmer (2007), both positive and negative feedback are necessary. Often the comfort and outlook of the students are not paid heed to, as a result of which the practice lags behind as a teaching technique. In exploration of the nature and effectiveness of feedback on students' writing, a set of questions are of primary importance: how and from whom do the students want to receive feedback; are there any specific styles preferred over others; what are the students' problems and expectations from feedback? These are crucial points to identify for the researchers and practitioners since development of skills is likely to be hindered if the learners are not satisfied with the feedback.

Purpose and Hypothesis
This study examines the students' expectations from feedback and how they want it to be shaped by the feedback providers. It is hypothesized that the students of tertiary level in Bangladesh have distinct likes and dislikes regarding feedback along with a number of problems and expectations from the practice. Preference of students for the different strategies of feedback (e.g., source, style, aspects, and type) and the steps to fulfill those, their perceptions regarding its different implementations and usefulness, and their problems and suggestions for improving the procedure are the concerns of the paper.

Literature Review
Language researchers consider feedback as an important part to be considered for development of the second language of learners. Feedback is responsible for improving the performance of the students by giving information to the learner about his or her performance on a learning task (Ur, 1996). According to Kulhavy and Wager (1993), in the context of writing, three broad meanings of feedback can be explored; first, feedback is a motivator when provided as a praise; second, it can reinforce learning by rewarding and punishing on prior behavior; and third, it gives information to change behavior on a particular direction. According to Hyland and Hyland (2001), the role of written feedback has mostly been seen as informational, a means of guiding reactions and recommendation to facilitate improvements. Ravand and Rasekh (2011) discussed all feedback variables where the mode, source, style, focus, complexity, and specificity can be considered as the basis of feedback strategies. The mode refers to whether the feedback is written or oral, provided in physical presence, or through the internet. The sources of feedback are mainly feedback from teacher and feedback from peer. "Feedback style refers to the linguistic and pragmatic characteristics of the feedback, including the illocutionary force of feedback (e.g., whether teacher uses questions, suggestions, or commands) and the explicitness of the feedback (e.g. whether codes or explicit corrections are provided)," say Ravand andRasekh (2011, p. 1140). Here, directness or indirectness is also considered as a feedback providing style. The two focuses in providing feedback are form and content based. The authors also talk about feedback in relation to complexity and specificity. Their study mentions that the least complex feedback (i.e., correct answer) revealed greater learner benefits in terms of efficiency and result than complex feedback. About specificity in providing feedback, they pointed out that several researchers have stated that feedback is considerably more effective when it delivers details of how to improve the writing rather than just specifying whether the student's work is correct or not.

Strategies of Feedback.
Feedback Process for Writing. Often feedback strategies for writing include responding, correcting, training and involving students, and finishing the feedback process (Harmer, 2007). Responding was observed during the writing process (i.e., the teacher says how the work appears to him or her and how successful it is before suggesting how it could be improved, by giving helpful comments). Correcting the written work was observed using correction codes (e.g., 'ww'-wrong word) or through summarizing comments saying what needed to be corrected. Harmer (2007) emphasized on training the students and finishing the feedback process which is only really finished when the students make the directed changes in their writing, tending to be autonomous. Some of the strategies are known to foster learner autonomy. Self-assessment, self-repair, and portfolio assessment are all evaluation processes that are associated with developing writing (Ekbatani, 2000). Self-assessment can act as an enhancement to teacher assessment and provide one of the most effective means of developing critical self-awareness to achieve the Munzerin Mohiuddin and Asifa Sultana required skills. Self-assessment heading for self-repair lets learners judge their performance in short sections and improve on their own creativity when they find deviances from the standard forms.

Corrective Feedback (CF).
Corrective Feedback is a suggestion to a learner in second language teaching, asserting that his or her use of the target language is incorrect (Ferreira et al., 2007). Lyster and Panova (2002) illustrated the significance of corrective feedback (CF) as it provides feedback on error which the students are unable to discover on their own. Ellis's (2008) typology can be considered useful as it outlined all the styles of CF. His typology of corrective feedback stretched a basis for scrutinizing the options to the teachers and for thoroughly experimenting with them in their own teaching. The typology is briefly presented below: • Direct CF: The teacher provides the correct form.
• Indirect CF: The teacher indicates the error but does not provide the correct form by giving the indication only or giving the indication by locating it.
• Metalinguistic CF: Some metalinguistic clue is provided as to the nature of the error by using error codes (e.g., art-article) or by giving a brief grammatical description.
• Focused (intensive) or unfocused (extensive) CF: This concerns whether the teacher chooses to correct all (or most) of the errors or only one or two specific types of errors.
• Electronic feedback: Teacher provides a hyperlink which contains correction of the error.
• Reformulation: A native speaker's reworking to make the students' writing seem like native language, keeping the original intact.

Studies on Students' Perception on Feedback.
Students' perspectives in research on teaching and learning are not emphasized enough in studies examining feedback in writing. The following sections discuss the available research on different aspects of perceptions.
Perceptions on Source of Feedback. Hyland and Hyland (2006) pointed out that surveys of students' feedback preferences commonly indicate that ESL students value teacher written feedback more highly than other forms like peer and oral feedback (e.g., Radecki & Swales 1988;Leki 1991;Enginarlar 1993;Saito 1994;Ferris 1995). Students often disfavor nonteacher feedback like that of peer evaluation and self-assessment teachers grading (Saito, 1994). However, there were also studies that revealed students' preferences towards peer review or combined participation of both teachers and peers in the feedback process (Best, et al., 2015;Chaudron, 1984;Mendonca & Johnson, 1994). An analysis of Chinese and Spanishspeaking students pointed out that both groups preferred constructive criticism and teachers' comments over those of other students (Nelson & Garson, 1998). In addition, when students were asked about their preference towards e-feedback, they favored peer feedback over e-feedback due to computer-anxiety (Lai, 2010). Lee (2008) in her research identified that 72.2% high-proficiency students and 40.9% low-proficiency students prefer "comments+marks/grades+error feedback." The students expressed that it is important to know the desired level of performance indicated through grades as well as why their writing is good or bad which can be expressed through comments. The need for positive feedback was identified in studies by several researchers where students expressed that approval of their performance often motivated them (e.g., Best, et al., 2015;Mahfoodh, 2011). They also pointed out that teachers' comments are helpful when they are "specific,""understandable," and "manageable" (Best, et al., 2015). Students' preferences towards reformulation of the language items were identified from studies on L2 graduates where they expressed their desire to know how the same ideas would be expressed by native speakers (Leki, 2006, as cited in Hyland & Hyland, 2006).

Perceptions on Focus/Aspects for Providing Feedback.
In a review of the current practices of corrective feedback on writing, Bitchener and Ferris (2012) noted that the students felt that the teacher gave more feedback on grammar in comparison with any other aspects of writing. However, the students were reported to assign equal importance to both grammar and ideas. This view was reiterated in other studies where the significance of having feedback on both meaning and structure was emphasized by all students, as they believed that one is a failure without the other (e.g., Mustafa, 2012). However, findings from L2 studies were not completely uniform where some students were reported to prefer feedback on both content and grammar, and some preferred grammar over content (Hyland & Hyland, 2006).

Perceptions on Types of Feedback (Indirect/Direct/Metalinguistic Cues).
Investigations of students' preferences towards direct and indirect feedback also produced diverse findings. Hyland and Hyland (2006) identified that students favored indirect clues rather than direct corrections. On the other hand, Bitchener and Ferris (2012) explored that students preferred direct feedback on their errors as it involved less effort for them in correcting. However, they believed that indirect feedback where errors are located and labeled with some kind of clarification or at least with error codes would help them to improve their writing in future.
Perceptions on Other Aspects of Feedback. Lee (2008) found that many researchers claim that L2 students desired a high quality on accuracy in writing, they were keen to have all their errors located by the teacher, and also wanted overt correction of errors (i.e., direct error feedback) from teachers (e.g. Komura, 1999;Lee, 2005;Leki, 1991;Rennie, 2000;Radecki &Swales, 1988, andLee, 2005). Similar perceptions were found in Grami's (2005) study on students in Saudi Arabia which claimed that the students "want,""expect," and "appreciate" teachers' written feedback about their surface-level errors. The study of Norouzian and Farahani (2012) discovered a huge difference between the perception of teachers and students preferring focused and unfocused feedback where the students wanted to have more focused feedback. Studies conducted on the students' responses towards feedback revealed that the highproficiency students were more positive than their low-proficiency peers in terms of their understanding, their view of the usefulness of teacher feedback, and their aptitude to correct their errors (Lee, 2008). Moreover, specific types of feedback were reported to obstruct the efficacy of the feedback (Mustafa, 2012).

The Present Study
The existing studies on feedback on students' writing exhibit a varied range of findings on the current practices with regard to feedback, and students' preferences towards these practices. Therefore, the present study explores the most preferred feedback strategies in writing by Bangladeshi L2 learners at the tertiary level who are currently doing or have recently done English language foundation courses, and identifies their problems and expectations from the existing feedback processes. The specific research questions are:  What are the feedback strategies for writing?
 What are the most preferred feedback techniques by the students?
 Why do the students prefer one technique/strategy over the other?
 What are their problems in the provided feedback?
 What are their expectations from feedback providers?

Methodology
The methodology section describes the research design, the procedure for data collection and means of data analysis including instruments for research and participants.
Research Design.This research has employed both the quantitative and qualitative methods. The quantitative method has been exemplified through the use of a questionnaire to investigate which feedback strategies are preferred most. The problems and expectations of the students regarding the existing feedback processes are identified through a qualitative approach.

Data Collection.
Since the study is based on the perceptions of the tertiary level students, the researcher selected the students who were doing and recently did the English language courses (ENG 091, ENG101, ENG102, ENG 201) of a university, namely BRAC University. Only one University was taken into account to have a stable data from the students of different perceptions experiencing the same or similar feedback processes. The process of the data collection was held by questionnaire using an online survey.
Instrument of the Research. The questionnaire consisted of 14 items in total. The question categories included: two small text-type items, five multiple choice questions, five five-point Likert-scale questions, and two open-ended questions where students had to answer on their own.

Participants.
A total of 50 participants took part in the survey for the research. All the respondents were tertiary level students of different departments, doing or recently completed the English language courses from the same university. The study included both male and female (1:2) and the age range was 20-24 years, irrespective of cultural background.

Data Analysis
The multiple choice questions are presented in piecharts to show the percentage of each preference level. To analyze the Likert scale (Likert, 1932) five-point items, the following range of points and interpretation key for finding results are used: Strongly agree -5, agree -4, neutral -3, disagree -2, strongly disagree -1. The interpretation key for finding the results as seen in several other quantitative researches arestrongly disagree: 1.00 -2.25, disagree: 2.26 -3.00, agree: 3.01 -3.75, strongly agree: 3.76 -5.0.
The final results of each question were given by the mean score which refers to the sum of all scores of the respondents in a group divided by the number of respondents, [X = ΣX/n]. The mean score provided the average score of each item that helped to get information by shortening large amounts of data (Seliger & Shohamy, 1989).
Qualitative data from openended questions are summarized, and later, analysis with proper insights are done with references to earlier studies from the literature review.

Findings
For each question, 50 responses are calculated to collect the data. The sample consists of different respondents with around 10-20 years of learning experience of English as a second language, which provides a chance to survey students from different proficiency levels.

Findings Related to Feedback Strategies (Source, Style, Aspects, Types, and Steps of Feedback Process).
Question no. 1-5 in the questionnaire is to investigate the preferences for the feedback strategies.

Perception on source of feedback.
The demonstration of data about the students' preference of the source of feedback is presented below (Fig. 1). Evidently, it showed that 40% (20 out of 50) of the respondents preferred feedback from the teacher; this represented the highest number of participants. Perception on styles of feedback. The following distribution showed the percentages of preferences for the different styles of feedback (Fig. 2). The findings show that 30% (15 out of 50) subjects liked to get comments, 28% (14 out of 50) preferred grades, 30% (15 out of 50) preferred to have an implementation of all the styles.

Fig. 2. Percentages of preferences for different styles of feedback
Percentages showing preferences for different types of feedback. Fig. 3 displays the response regarding the preference for the types of feedback. A remarkable percentage of students preferred explicit corrective feedback or direct feedback -54% (27 out of 50) of participants.

Fig. 3. Percentages showing preferences for different types of feedback
Perception on steps of feedback process. The steps of the feedback process in Fig. 4 are adapted from Harmer's (2007) steps of feedback process for writing tasks. The highest number of respondents, which is about 30% (15 out of 50) of the total, recognizes that all the steps are equally necessary for the process of feedback.

Findings from perceptions of feedback practice (implementations and usefulness).
Questions 6-10 in the questionnaire are designed to investigate the perception related to implementation (necessity for feedback, praise/criticism, focused/unfocused, form/meaning, usefulness, and understanding) and usefulness of the existing feedback practices. 50 responses are surveyed for each question in Table 1. The Likert scale was used to measure these question. The results of the Table 1 show how the students accepted or rejected some facts regarding the usefulness and implementation of feedback practice. It showed that most agreed on the importance of feedback and mostly the positive ones. They agreed on both focused and unfocused feedback but disagreed that form is more important than meaning. It is also found that they feel the existing feedback practices to be helpful.

Findings Related to Problems and Expectations from Feedback
The following sections discuss the problems and expectations in relation to various feedback practices.

Problems in existing feedback practices.
The students' satisfaction in case of both progress and comprehending the feedback given are examined by the question "Are you happy with the feedback you get? Are there any difficulties in understanding the feedback you get on your writing?" The participants are more or less satisfied with the feedback they get as they feel the feedback varies with the varying performance.
However, they complained that feedback with detailed reasoning is missing in their scripts. One of the participants demanded constructive criticism and claimed, "Mere pointing out of errors without adequate reasoning and explanation is not useful." Some teachers mention a problem and provide no correction; on the other hand, some only correct without mentioning what is exactly wrong or why it is an error. Another mentioned, "I want more intensive feedback like formation of a sentence in a different way than I do, different vocabulary, more structural directions, and providing ideas can be very helpful." The causes of dissatisfaction found out is regarding the communication gap between students and feedback providers which in many cases leads in providing feedback without understanding what the studentshas tried to express in the writings. A number of students pointed out the problem of communication gap as a difficulty.
Another problem pointed out by the student is that the teachers only provide negative feedback which is a reason for getting demotivated in the absence of any positive feedback. "What does it take just to write 'good!' at the end of my answer? It is able to create great enthusiasm," remarked one of the respondents.
Expectations from feedback process. A number of expectations and suggestions from students are categorized and are presented accordingly in the points below: • Additional slots: The respondents proposed extra classes to explain the provided feedback besides the consultation hours for more communication. In the classes they can learn more from others' mistakes where the teachers will explain the errors. • Extra care for weaker ones: "Be polite while giving feedback and need to encourage students," a respondent suggested, claiming that they should be encouraged to ask for help. The consultation hours can have separate time allocation for the weaker ones, low proficient ones, and introvert ones while time is allocated properly for all students having different learning styles. • Pre-writing and pre-feedback tasks: Teachers can give a sample write-up having feedback on it, from which the students can learn common errors, feedback styles of the teacher, etc. before they write and submit their own writing. "Peer feedback should be mandatory before teacher feedback," a student demanded. This can be a practice of self-correction and also save time and hard work of the teachers. • Post-writing and post-feedback tasks: Students demanded scope for self-correction in class before submitting a work which they miss out due to short time span. One of them proposed, "After giving feedback, if teachers check whether students are making corrections and improving or not, and then take necessary steps, that can help a lot." • New supplements in the process: The respondents claim that it is not necessary that writing must always have written feedback; oral feedback procedures can bring a variation sometimes. A teacher can provide an example in the feedback to show a correction without directly correcting the students' writing. A friendly conversation developing communication between a teacher and a studentis considered useful more than a formal appointed consultation. A student claiming to be highly proficient expected that sometimes a native English teacher can provide feedback which is of better help.

Data Analysis
This section presents the analysis of the data described in the previous section, with references to the earlier studies presented in the literature review. The discussion will scrutinize the perceptions from the findings of the current study and that of the other researches and distinguish to what extent they match or differ.
Discussion on perception of source of feedback. The study shows that 40% of participants prefer teacher feedback, 8% prefer peer feedback, and 6% prefer computer assisted e-feedback. The data provides similar views from the earlier studies where the students prefer teacher feedback from more than any other sources and chose peer feedback over e-feedback which might be due to computer anxiety as drafted in those studies (Hyland & Hyland, 2006;Nelson & Garson, 1998;Lai, 2008). The study shows that only 6% students preferred self-assessment which is equal to that of the peer feedback. It can be stated that both the sources are equally preferable by the respondents. Saito (1994) also found out teacher feedback is preferred more than peer and self-correction. Hence, the feedback from the teacher is accepted the most. However, it cannot be denied that 38% of respondents wanted a process that uses all the sources equally. The strategies can be undertaken by prioritizing teacher-feedback and balancing it with a mixture of all the other sources of feedback. The popularity of teacher feedback in the entire context indicates that the teachers are the most reliable source of feedback to the students.

Discussion on perception on styles of feedback.
Most research shows that the students prefer comments over grades and they prefer grades only along with comments (e.g., Lee, 2008;Best, et al., 2015;Mahfoodh, 2011). The present study shows 6% of the participants preferred reformulation while it is noted in Leki's (2006) research that the style is desired by graduates mostly, although the study shows the preference percentage of comments and grades are not very different. The study showing 30% of the participants preferring all the styles equally disagrees with research like that of Lee's (2008), which finds 72.2% of HP students and 40.9% of LP students want "comments+marks/grades+error feedback." The possible reason for the dissimilarity can be the difference of comfort of having all the styles together. The 6% of students preferring reformulation is apparently the high proficient ones who want the advanced evaluation more. The comparisons show that comments and grades are preferred more than corrections or error feedback in earlier studies and the present research too.

Discussion on perception on aspects of feedback.
Writing skill demands that all aspects of writing (content, organization, vocabulary, and mechanics) be developed. The research shows that 24% of students prefer feedback mostly on content, 14% on organization, 12% on vocabulary, 28% on mechanics or grammar, and 22% prefer all to be considered with equal importance. The studies also show similar views about students preferring more feedback on mechanics rather than content and other aspects (e.g., Bitchener & Ferris, 2012;Hyland & Hyland, 2006). 22% of students preferring all the aspects which is not remarkably less than the 28% preference for mechanics, justifies that a sustainable number of students feel the importance of all the aspects to be considered equally in their writing. It can be predicted that grammar or syntax is comparatively a more complex aspect of writing than the other ones; hence, it requires more care to bring proficiency in it.
Discussion on perception on types of feedback. The high popularity of direct feedback is also seen in the study of Bitchener and Ferris (2012). As a whole, the preference for mostly the direct feedback (54% of students) can be interpreted as a demand to know the errors clearly by the students and work out those easily. The studies show that students choose feedback mostly with metalinguistic cues as they want to reorganize the writing (e.g., Hyland & Hyland, 2006). 8% selected indirect feedback without locating errors and 4% preferred locating errors that differentiatedtheir work with the earlier studies. The students criticize the practice of only indicating the error (Mustafa, 2012). It might have appeared because more subjects selected the indirect feedback without locating errors to try harder to improve. However, the large section of subjects preferring direct feedback and that which preferred all the types simultaneously, are the ones to be given more attention.
Discussion on perception on steps of feedback process. Specific research on Harmer's (2007) steps of feedback process was not found. So, the present study explores the approach by investigating students' preference for each of the steps. The study claims that 24% thinks students should be trained about reading and comprehending the feedback styles and types which denotes that many students are having problems in understanding the feedback and 30% favored all the steps to be followed with equal importance. The largest percentage at the end show that a greater number of students want all the proposed steps of the feedback processes to be observed for their writing. The other steps are voted by a small percentage but the distribution can help the teachers to understand the desire and need of the students for a systematic feedback process.
Discussion on perceptions on positive/negative, focused/unfocused, precise/detailed feedback. The development of writing depends on the implementation of feedback on students which can be positive and negative feedback, focused and unfocused feedback, and precise and detailed feedback. The participants "strongly agreed" that positive feedback is more useful than negative feedback. In the earlier studies, it is found that both are equally preferred by the students (e.g., Mahfoodh, 2011). Positive feedback inspires students and cheers them up, while Lee's (2008) research reveals that students want more comments and that includes both positive and negative feedback. The present study finds the participants more inclined to positive feedback; receiving negative feedback appears demotivating and monotonous to them.
It is "agreed" by the respondents that the teachers should focus on all types of errors rather than one or two specific ones. This view differs with some studies and agrees with others.
Research reveals that students want overt correction in their writings including the surface level errors (e.g., Lee, 2008;Grami, 2005). However, Norouzian and Farahani (2012) found that only the teacher feels they should correct all the errors, not the students. Often the focused feedback helps the student to concentrate on specific errors which is not easy in case of unfocused ones. The fact is a bit controversial among the students for which the present study found them "agreed" but not "strongly agreed" on it.
The open-ended answers discovered that the students desire detailed feedback; they prefer "precise comments" but detailed feedback. The study of Best, et al. (2015) also says students are against grading as it is not a detailed evaluation and, in case of comments, they want specific and manageable comments. It suggests that students want explanation of their faults and clear-cut comments which are easy to decipher.
Discussion on problems and expectations from feedback. The three main problems spotted are: absence of detailed feedback and reasoning errors, communication gap between student and teacher causing misinterpretation in writing, and excess negative feedback which is demotivating. Often the students do not understand why their mistakes are considered as "errors," and it is the teacher's duty to give exact reasoning behind each of those which will help them understand both the nature and importance of those for development. The distance or communication gap between the teacher and student does not let the teachers understand what the students tried to express in their writing. For this, a face to face conversation or conference is necessary sometimes, and is recommended by the students as well.The teachers mainly focus on negative feedback due to their own necessity and time constraints. The students thoughtfully pointed out this practice to be a problem in feedback. Providing only negative feedback discourages students and thereby the purpose of positive feedback remains unfulfilled.
The expectations of the students are to resolve the issues by adding extra slots to mitigate the communication gap through training and conferencing. The need for extra slots has been suggested since it is possible to feel that the regular class schedule makes it difficult for both students and teachers to work with feedback. They wished for proper allocation and distribution of time for each type of students since not all are equally proficient. They also mentioned pre-feedback and post-feedback tasks for detailed feedback and reasoning. Tasks like peer checking before teacher feedback, self-correction, and so on would make the process easier and efficient which the students fairly mentioned as their needs. To add flavor and bring variety to the feedback strategies, the respondents asked for some supplements in the feedback process for adapting it for all types of students including weaker ones, introverts, and advanced learners. Providing the students with examples of feedback may help them to comprehend the teachers' adopted styles, focus, and types of feedback. The suggestion of involving a native feedback provider once in a while for the advanced students might add an additional level by enriching the feedback process for their further development in writing.

Conclusion
To conclude it can be said that students have their own perspectives including preferences, problems, and expectations regarding the feedback process which does not necessarily agree with all the aspects of the practice. Their views and attitudes should concern the feedback providers, researchers, and experts planning ESL and EFL classroom needs, not being solely dependent on the existing tradition and language theories. Due to time constraints, the selection of the students for the research is very random which would have been richer if the students are categorized into high and low proficient ones. Further studies can be done in a larger scale by comparing the students' views with their teachers, two or more contexts or institutions, and cultural backgrounds. The same research might bring new results by conducting it in a large scale. In brief, this research can be helpful in adding a new dimension to feedback in writing or can be adapted by researchers in future for further studies.