The Effect of Error Correction vs. Error Detection on Iranian Pre-Intermediate EFL Learners' Writing Achievement

This study tries to answer some ever-existent questions in writing fields regarding approaching the most effective ways to give feedback to students' errors in writing by comparing the effect of error correction and error detection on the improvement of students' writing ability. In order to achieve this goal, 60 pre-intermediate English learners were randomly divided into two groups: the first one was Direct Feedback Group, receiving feedback on their writing through error correction (DFG) and the other one was Indirect Feedback Group (IFG), receiving feedback in their writing through error detection along with the codes. The learners were taking English classes in a private English center and were supposed to self-correct and hand in their writings when received indirect error feedback. The results suggested that error detection along with the codes led to better improvement in the learners’ writing than the error correction treatment.


Introduction
Providing feedback to students' writing errors has always been one of the teachers' difficult tasks. Bearing this fact in mind, a lot has been written on the issue surrounding error feedback to students' writings most of which are controversial and even inconclusive. Trusscott (1999) argued that error correction is of little benefit and even harmful and therefore, it should be kept aside in EFL writing classrooms. On the other hand, Ferris (2002) claimed that error correction is widely seen by teachers and students as an important factor in improvement of the writing ability. Some other studies (Cohen, 1987;Ferris and Roberts, 2001) also pointed to the importance of accuracy and therefore error correction in students' writings and the students' willingness to receive feedback on their errors. Fathman and Wally (1990) in a study investigated the effect of error feedback on students' improvement in writing. They compared groups who received error feedback to those receiving little -if any-feedback. They found out that the groups obtaining error feedback did much better in grammatical writing than those receiving little-if anyfeedback on their writings. Ferris and Roberts (2001) state that controversy continues as to whether error feedback improves students' writing accuracy and their overall writing ability.
More literature in this area confirms that error correction debate continues. Tim Ashwell (2000) has recommended a focus on form with novice writers and claims that only concrete content feedback is useful for novices. Jean Chandler (2003) has found that students who received error correction and are asked to act on those corrections are less likely to repeat those same errors in subsequent assignments.
Since providing feedback to students' writings is difficult and time-consuming, it is worth investigating the most effective way to react to errors. There has also been a bulk of research concerning the effect of different feedback types on students' writing improvement (e.g., Ferris & Helt, 2000;Frantzen, 1995, Sheppard, 1992. Several issues have always been under investigation in such studies as (a) the idea of providing feedback or not providing it, (b) the question of giving direct or indirect feedback, and finally (c) the question of giving feedback to all or some specific kinds of errors.
According to Ferris & Roberts (2001), direct feedback is given when the teacher provides the correct form for the student writer. On the other hand, indirect feedback occurs when the teacher indicates in some way (underlying or providing codes) that an error exists but does not provide the correction, therefore letting the writer know that there is a problem but leaving it to the them to solve it. Ferris and Roberts (2001) found out that the groups who received feedback outperformed the no-feedback group on the self-editing task but there were no significant difference between codes and no-codes group. Ferris (2002) maintains that the danger of direct error feedback is that teachers may misinterpret students' meaning and put words into their mouths, though direct feedback may be appropriate for beginner students and when the errors are "untreatable" (i.e., errors that students are not able to self-correct, such as syntax and vocabulary errors) ( Ferris,2002( Ferris, ,2003. Chandler (2003) and Ferris (2003) have advocated the use of indirect coded feedback especially with novice writers. Leki (1991) also indicated that students prefer indirect feedback on college writing rather than corrections.
In a study, Atai (2000) compared the effect of self-versus peer-correction on students' writing ability; the results showed an advantage for the students receiving peer-correction in terms of their general writing ability, therefore, he concluded that peer-evaluation helped in achieving much better pieces of writing as a whole.
Alfi (2004) found that the more explicit the degree of feedback is, the more accurate the learners' written performance would be. The research findings, although inconclusive, have raised several significant issues for teachers to consider.
There is no doubt that writing is the most difficult skill for L2 learners to master. According to Richards and Renandya (2008) the difficulty lies not only in generating and organizing ideas, but also in translating these ideas into a readable text. They state that the skills involved in writing are highly complex; L2 writers have to pay attention to higher level skills of planning and organizing as well as lower level skills of spelling, punctuation, word choice, and so on .The difficulty becomes even more pronounced if their language proficiency is weak. Therefore, writing should play a more prominent role in classroom-based studies of second language acquisition.
There have been so many conflicting approaches surrounding writing instruction, two of which that mainly have affected writing strands are process vs. product writing.
Even as late as 1970, teachers were mostly concerned with the final product of writing: the essay, the report, the story and what that product should look like. Compositions were supposed to (a) meet certain standards of prescribed English rhetorical style, (b) reflect accurate grammar, and (c) be organized in conformity with what the audience would consider to be conventional (Brown, 2001). Later, teachers shifted their focus on content and message i.e., process approach to writing when they discovered the advantage of looking at learners as creators of language.
Seow (1995) holds that the process approach to writing comprises four basic stages-planning, drafting, revising and editing. For each stage, suggestions are provided as to the kinds of classroom activities that support the learning of specific writing skills. For example, at the planning stage, teachers can help students generate ideas through such activities as brainstorming, clustering and rapid free writing. At revising stage, feedback has been considered as a key element contributing to skill development. In revision stage the role of feedback by teacher would gain prominence.
Although there have been controversies regarding the efficiency of providing error feedback on students' writings (Ferris 1999a, Truscott, 1999, there are some pieces of evidence supporting the idea of providing feedback and its efficiency in overall quality of students' writings. A lot of studies have investigated and compared the effect of direct versus indirect strategies in helping learners in accuracy and overall writing quality (Ferris, 1995a;Ferris, 1995b;Lelande, 1982;Robb, Ross, & Shortreed, 1986) Direct or explicit feedback occurs when the teacher identifies an error and provides the correct form, while indirect strategies refer to situations when the teacher indicates that an error has been made but does not provide a correction, thereby leaving the student to diagnose and correct it.
Additionally, studies examining the effect of indirect feedback strategies have tended to make a further distinction between those that do or do not use a code. Coded feedback points to the exact location of an error, and the type of error involved is indicated with a code. (For example PS means an error in the use of the past simple tense). Not coded feedback refers to instances when the teacher underlines an error, circles an error or places an error tally in the margin, but, in each case, leaves students to diagnose and correct the error.
Since effective written communication is the main goal of many L2 courses and is a key to achievement in academic grounds, drawing students' attention to mismatches between the target language writing features and their own output would be of great importance.
The present study tries to identify one of the effective strategies in teaching writing at the revision stage and especially writing feedback and evaluation.

Research question
To what extent would the type of written feedback (direct or indirect) determine the performance improvement in new pieces of writing?

Participants
Unlike most error correction studies to date that have focused on more advanced learners in academic setting, this study comprised 30 female pre-intermediate students learning English as a foreign language in a language institute in Iran ages ranged from early teens to late twenties and they were divided into two groups. Direct Feedback Group (DFG) and Indirect Feedback Group (IFG). Majority of the learners were taking English classes for one consecutive year. Besides, due to their elementary and beginner levels of proficiency until then, they had received little direct or serious instruction in writing essays.
For one semester (the five week period of the treatment), they followed an eclectic method in their curriculum, the aim of which was to improve their communicative skills in the four macro-skills (reading, writing, peaking, listening). As part of their course, they had to achieve competency in writing essays which was similar to the task set for the research. The research task, therefore, provided practice with feedback for these assessments which will be explained below.
Pre and post test Learners were given The American Headway Placement Test to be homogenized in addition to a short essay writing task as their pretest. The same short essay writing task was given to the learners as post test and both pre and post tests were corrected by Analytic Scoring Rubric Checklist adopted from Roebuck (2001) Procedure In order to insure the homogeneity of the subjects, they were given two-pre-tests before the treatment; The American Headway Placement Test was given to all the participants to make sure that they were homogeneous and have similar general language proficiency. They also were asked to write short essays to make sure that there is no significant difference between their writing ability.
Then they were randomly divided into two groups: group one who received direct feedback on their writing through error correction and group (DFG) two who received indirect feedback in their writing, namely error detection along with codes (IFG).
Both groups received the same amount of instruction, and the same amount of time was spent teaching writing in each class Group one received direct written corrective feedback in their essays delivered to the teacher. In this group, the teacher underlined the students' writing errors and provided the correct form for them; and then, the corrected writings were given back to the students in order for them to be aware of their errors and the corrected form and to improve them in the following writings.
Group two received indirect corrective feedback (detection). In this group, the error codes were introduced and explained to the students in the beginning of the term by the teacher; and then, the errors of the students' writings were detected, i.e, they were underlined, but the corrected forms were not provided for them; the related codes were written next to each error to give a clue to the kind of errors which was made.
These error codes were a clue to the type of errors made by students; for example, wherever "S.P." was written under a word, it meant that an error of spelling had occurred; or wherever ""W.O" was written under a phrase or sentence, it mentioned that there was a problem regarding the word order (The complete list of these error codes are available in the appendix part). Then the coded corrected writings were given back to the students and they were asked to correct it and hand in the corrected writings along with the original ones in the following sessions. The students were asked to write short essays-about two or three paragraphs-whose topics revolved mainly around a central theme. All error types were corrected or detected in the two groups and no selective error correction or detection was involved.
The treatment lasted for five weeks and 3 sessions per week. In total, each student was required to hand in at least 8 writings which equals almost one every other session. After this five-week-period, a writing post-test was given to the participants to examine the short term effect of writing detection treatment versus writing correction treatment.
Both the pre-test writings and the post-test writings were checked and scored based on a modified version of the Analytic Scoring Rubric Checklist adopted from Roebuck (2001). (see appendix2). And for the data analysis a T-test was run to compare the mean scores of the groups.

Results
This section presents the results of investigating the extent to which different types of written feedback on students' writings helped the learners to improve their general ability in writing when wiring new texts.
The means and standard deviations of the immediate writing post-test scores for each treatment group are presented in table 1. The mean score for the error detection subjects exceeded that for the error correction subjects.
Table 1 (see appendix 1). shows that there is a significant difference in the scores obtained from the two groups, direct feedback group and indirect feedback group in the post test (t= -6.284, p=.000). Accordingly, our research question was answered in affirmative in favor of indirect or coded feedback group, hence using coded feedback was shown to exert a positive effect on the writing ability improvement of the learners compared to direct ones. In fact, the subjects performed better on writing test through exposure to the coded feedback, and not the direct counterpart

Discussion
According to Ferris (2002), error correction is widely seen by teachers as crucial for writing development and generally expected and welcomed by L2 students. In the present study, too, the students welcomed error correction on their writings and even demanded it when was not provided.
Truscott's review of studies by Kepner (1991), Semke (1984, and Sheppard (1992) claimed that error correction does not have a significant effect on improving L2 student writing. Given this limited range of studies, the present study sought to expand the base by investigating the effect of two types of feedback on the general improvement of students' ability in writing.
To address the gap, the results of the present study tentatively support the idea of the usefulness of error feedback on students' writings ( Fathman & Walley, 1990;Ferris, 1995;Ferris, 1997;Ferris & Helt, 2000;Lenande, 1987;Polio et al., 1998) provided that it is not direct. However, it can be mentioned that the results also tacitly support Truscott's claim that provision of direct corrective feedback on L2 writing is ineffective since the learners' of direct feedback group (DFG) have shown no significant improvement compared to their indirect feedback group rival. Truscott also questions the long term effect of direct feedback claiming that it won't last long, further studies are required to examine the long term effect of indirect feedback which is not in format of direct grammar teaching and correction.
The present study also found that the type of feedback provided, had a significant effect on the learners writing improvement. It was observed that those learners who received indirect feedback on their writing through error detection along with codes, showed greater improvement in producing new pieces of writing. And, they also welcomed this type of feedback more than the direct error correction form. Therefore, the findings of this study seem to confirm the idea of indirect error feedback over the direct error feedback (Ferris, 2003;Frantzen, 1995;Lanande, 1982) through increased engagement and attention to problems.
Besides, it can be argued that the reason of better progress on behalf of the learners in indirect feedback, could be the effort made to locate and provide codes for them and this would cause the more reflection on the papers which is similar to consciousness raising task (Ellis, 2003) and this led to more encouragement and more independency.

Most subjects stated:
When our errors are underlined and the correct forms are provided for us (error correction), and then, they are handed back to us, we just take a cursory look at it to learn about our mistakes ( supporting truscott's rejection of ineffectiveness of direct feedback); but, when the errors are just underlined and we had to work out the correct form ourselves hence, we would learn it better. Therefore, it can be concluded that error detection can be used as a complementary strategy to implicit teaching grammatical forms in order for learners to reach to the maximum level of improvement in writing. It also may alleviate the problem of direct error correction which affectively filters learners' development in language acquisition (Krashen, 2003).
Moreover, since the participants had enrolled in an English course which was supposed to provide teaching and feedback in all four language macro-skills (speaking, reading, listening, and writing), no control group was designated for the present study. Therefore it would be mentioned as one of the limitations of the study Additionally, the present study studied the short-term effect of providing direct versus indirect feedback in pre-intermediate Iranian students learning English as the foreign language. Future research could survey the long-term effect as well and explore the best possible way of providing indirect feedback on students' writings. Sheppard, K. (1992