November 2016 – Volume 20, Number 3
Matthew Coomber
Ritsumeikan University, Japan
<mcoomberhotmail.com>
Abstract
Second language writers need to develop the ability to revise their writing independently of third party advice; thus, it is important that teachers devise methods by which to promote habits of self-directed revision. This quasi-experimental study investigates three classroom activities designed to encourage students to independently revise essays prior to receiving teacher feedback. One class of second-year Japanese university students revised the first draft of an essay after completing the following activities: 1) an oral presentation of essays; 2) a grammar workshop; and 3) a 20-point checklist. A second class was designated as a control group and revised their essays with no additional input. Teacher feedback was provided on the second drafts in the same way to each group. It was found that the treatment group made almost three times as many revisions to their first drafts. This group’s revisions were also more successful, with 80% having improved on the first drafts, compared to 66% of those made by the control group. However, on the teacher-reviewed second drafts, the majority of revisions derived from teacher feedback; less than 15% of revisions made by both groups were self-directed.
Keywords: second language writing; revision; self-directed editing; feedback
Introduction
In their studies of the role of feedback in second language writing, both Connor and Asenavage (1994) and Paulus (1999) found that over half the revisions made by their students were unconnected either to the teacher or peer feedback they had received. Whether these revisions were self-directed or derived from some external source could not be established; however, either possibility indicates that these particular learners demonstrated a high level of autonomy with regard to taking responsibility for improving their writing. Both studies, however, were carried out in English as a Second Language (ESL) classes, in which it seems likely that extrinsic, and perhaps also intrinsic, motivation may be relatively high. The Japanese university students I teach, on the other hand, take English classes as a compulsory component of their degrees; these English as a Foreign Language (EFL) learners, in my experience, often seem to lack motivation to independently revise their written work and depend too heavily on teacher feedback. However, if writing is to be viewed as a long-term life skill, rather than just a short-term academic goal, learners need to develop the habit of self-directed revision, and motivating students to take responsibility for revising their own written work should thus be a central goal of second language (L2) writing classes. Relatively little research to date, however, has looked at ways in which teachers can do this. The current study seeks to address this deficiency.
Literature Review
Feedback and Revision
Despite the inevitable debate about how best to do so, it would be hard to dispute that the central goal of a writing teacher is to help students improve their writing, a task which Casanave (2004) describes as ‘the most consuming of all dilemmas for L2 writing teachers’ (p. 64). Solving the dilemma is problematic, and even defining it is far from straightforward: while teachers and learners may have an immediate focus on the local goal of improving a particular piece of writing through feedback and rewriting, both parties must also remain aware of the long-term goal of improved language proficiency.
Ideally then, teachers would be able to help their students improve both the quality of their drafts in the short-term and overall English ability in the long-term. Before considering how teachers can help learners to revise their papers, however, it may be instructive to make some further observations about the nature of revision.
First, as Bereiter and Scarmalia (1987) point out, revising written work is not a simple task: the very presence of an extant text may constrain the options available to a writer, especially a relatively unskilled one. Not only must writers have the ability to evaluate the text, but they also need the skills and knowledge to make the necessary changes; those who lack this may even resort to simply cutting sections they have identified as problematic (Conrad & Goldstein, 1999; Kobayashi & Rinnert, 2001). Thus, identifying and remedying a problem are two distinct steps in the process. As Hayes (2004, cited in Myhill & Jones, 2007) rightly clarifies, to characterise revision as simply fixing problems offers only a partial view of its role: to revise is also to make a text more coherent, more descriptive, less repetitive, or clearer. Learners, however, do not necessarily share this view, with many conceiving of revision as mainly an editing task (Christiansen & Bloch, 2016; Cresswell, 2000; Leki, 1991; Yagelski, 1995). Finally, revision is not a one-size-fits-all activity. Writers who plan more carefully and in greater detail may have less need to revise than those who adopt a stream-of-consciousness approach to text generation (Casanave, 2004). Certainly when considering only between-draft revision, amount is not the crucial factor: some papers will require less work than others; moreover, it is far from true that all revisions lead to better writing.
While the behaviours and outcomes involved in revising written work seem to be complex, there appears to be little dispute in terms of its value. Sengupta (2000) is a rare dissenting voice, claiming that “L2 revision does not seem to lead to improved texts” (p. 99), yet basing this claim in part on Yagelski (1995), an L1 study. For most, however, the question that remains is how teachers can best help students effectively revise their writing. Although a large and growing body of research addresses the issue, “there is little agreement among teachers and researchers about how teachers should respond to student writing” (Fathman & Whalley, 1990, p. 178). To this issue it could be added that there is likely to be no greater level of agreement among the students themselves: opinions will depend on a variety of factors (Cumming, 2001), including proficiency level, cultural background, goals, and motivation.
In a study of first language (L1) writers, Yagelski (1995) found student revisions to closely follow what they viewed as teacher preferences; it seems probable that EFL students, less sure of their own linguistic competence, would be even more likely to interpret teacher feedback as directives, rather than advice. Yet if revision becomes largely teacher-directed, then its full potential is no longer being realised: by critically re-reading and reappraising their own work, learners can develop both their writing skill and language ability. Although confident and self-motivated students may be able to see past the teacher as the sole source of wisdom, there is a danger that for others teacher feedback may encourage abdication of responsibility for their own writing (Hyland, 2000).
Providing appropriate and useful feedback is a difficult task for teachers, but it is the learner who is faced with the task of converting this feedback into effective revisions, although relatively little research seems to have focused on how they do this (Murphy, 2000).
First, as Lee and Schallert (2008) suggest, there is the potential for any teacher comment on student writing to be misunderstood, particularly by lower level learners who may unfortunately also be less likely to ask for clarification. Goldstein (2004, 2006) offers a more detailed overview of research into this issue, noting that learners may find feedback confusing, believe they understand it when they do not, try to use it without understanding it, or understand the feedback but lack the knowledge or strategies to revise their work accordingly. If for any of these reasons learners are unable or unwilling to make use of feedback, then the effort of the teacher providing it becomes a futile one (Lee, 2013). Moreover, even when students do engage with feedback, interpersonal and cultural features may also impact in many ways upon how feedback is interpreted and applied (Hyland & Hyland, 2006). Thus, as Ferris and Hedgcock (1998) note, the degree to which individual learners are able to revise their writing as a result of teacher feedback varies enormously.
Nevertheless, for all the discussion of how teachers should respond to their students’ writing, most researchers seem to agree on one issue: formative feedback on preliminary drafts of writing assignments, whether from teachers or peers, is an essential component of writing instruction. However, it is not a panacea. Providing written feedback is just one way in which learners can be encouraged to revise their compositions. Furthermore, it must be remembered that feedback is a means, not an end. The real goal is improved writing and language ability, and as useful as feedback may be, we must not assume that second language writers are unable to make improvements to their own work without the direct intervention of a third party.
Self-Directed Revision
Much of the literature has characterised between-draft revision as something that is done primarily in response to feedback. Self-directed revision, when discussed at all (Connor & Asenavage, 1994; Conrad & Goldstein, 1999; Paulus, 1999), has generally been mentioned as an aside; few researchers have explicitly set out to study this. Nonetheless, revision occurs at all stages of the writing process, and both mental revision pre-writing and the revision which takes place during writing are undoubtedly self-directed. As learners clearly have the ability to revise their own writing, could it be that an excessive focus on external input between drafts, be it teacher or peer-directed, robs them of the motivation and confidence to make revisions of their own? As Muncie (2000) points out, when interpreting teacher or peer feedback, “the learner does not have to decide what to do, only (at best) how to do it” (p. 47). Justifying his rather radical remedy of eliminating between-draft teacher feedback, Muncie suggests that the improvements seen on final drafts are both inevitable and fairly meaningless, being simply “the product of the organisational skills of an expert” (p. 49).
Although Muncie’s conclusion would appear to ignore the insights of sociocultural theory into the learning process and seem to gloss over the potential influence of the specific type of feedback offered, the contention that direct teacher intervention in the writing process may not be the best way to promote learner autonomy and improve writing ability in the long-term is worth considering. Studies have shown that learners may use similar comments in different ways (Conrad & Goldstein, 1999): whether or not they utilise feedback critically will depend on a variety of personal, interpersonal, cultural, and practical factors. For those who tend to adopt feedback uncritically, perhaps constrained by the dual role of teacher as both advisor and evaluator, the benefits may only be felt in the short-term. A greater emphasis on self-directed revision may therefore be one way teachers can re-focus writing classes towards longer-term goals.
The few studies which have specifically investigated self-directed revisions have, on the whole, produced positive results. For instance, Gass (1983) found that a diverse group of ESL learners could, overall, correctly identify which of their own sentences were ungrammatical 68% of the time, although it must be noted there was considerable variation among learners and their success rate in correcting these sentences was not as high. Makino (1993), working with university students in Japan, found that all his participants were able to correct their own sentence-level errors to some extent; however, he also noted that they had greater success in revising when given teacher cues for error location. Both studies, moreover, looked only at isolated sentences, which in the case of Makino had been written in Japanese and translated by the learners into English. Thus, while offering some indication as to learners’ ability to revise their own work, it does not necessarily follow that these students would have equal success with essay-level revision. In a study that did focus on essay writing, Polio, Fleck, and Leder (1998) asked learners to write 30-minute essays, then two days later, having provided no feedback on the drafts, gave them 60 minutes to revise these compositions. Although they evaluated only the linguistic accuracy of these essays, they found that on this measure there was significant improvement in the revised essays. It is also of interest that, beyond sentence-level editing, learners in this study exhibited a range of revision behaviours (e.g., using their 60 minutes to write new essays from scratch), all of which resulted in linguistically more accurate essays, providing further evidence that learners do not necessarily need third party feedback to successfully revise their own work.
In a study conducted in Japan, Suzuki (2008) compared the self-directed and peer-directed revisions of two groups of university students, finding that although more word-level changes were made during peer revision, self-editing generated more changes at the sentence and discourse- levels and almost twice as many overall. Using think-aloud protocols, recordings of peer review discussions, and stimulated recall interviews, Suzuki also found a high frequency of negotiation episodes during each type of review, providing an indication of the potential benefits of self-review of written work in both the long and short-term. Finally, Diab (2016) found that “self-feedback resulted in significantly reduced lexical errors compared to teacher feedback” (p. 64), offering further evidence of the need for teachers to incorporate a focus on self-directed revision into writing classes.
Teachers may readily accept the value of self-editing; learners, however, may not necessarily share this view. How then can we can encourage, persuade, or motivate our students to engage in it? After years of mounting frustration with first drafts of written work which gave the impression of having been rushed off at the last minute with the sole intention of meeting the deadline, I adopted the policy of collecting first drafts at the start of class, skimming through them quickly during the lesson, and at the end explaining that I had decided to give students an extra week to work on revising these essays. The results of this change were not encouraging. Some students made minor revisions, but most seemed to resubmit exactly the same draft the following week, knowing that at that point they would get my feedback anyway.
While learners with strong intrinsic motivation probably need little external stimulus, those studying English as a compulsory subject may be more reluctant to revise independently and less likely to see the value of doing so. In this situation, simply asking or encouraging students to revise may not be enough. What is needed is to create a greater incentive for students to return to their original drafts. The remainder of this paper examines how I attempted to do this.
The Research Questions
The first step to revising any piece of writing is to re-read it (Ferris, 2008; Ferris et al., 2013; Kamimura, 2000; Myhill & Jones, 2007). In the courses I teach, in addition to the composition element, students are required to give two oral presentations. Despite my efforts to encourage a more natural presentation style, almost all students approach this assignment as an exercise in memorisation: they write out a script, learn it, and attempt to repeat it verbatim. By specifying that the first presentation must be on the same topic as the students’ essays, I thus found an excellent opportunity to motivate students to re-read their first drafts, and not just once, but many times. Beyond getting students to read their drafts again though, I also wanted to give them some indication of what to look for when they did so. To this end, two further changes were introduced. First, before submission of the second drafts, one class was devoted to a grammar workshop targeting common problems. Second, I created a 20-point checklist for students to complete and submit along with their revised essay (Appendix A).
In a pilot study conducted with a class of 18 students, I found that 11 students had revised their essays substantially, and an additional five students had made some minor changes; only two had re-submitted an unchanged draft. Encouraged by this preliminary finding, I decided to undertake a more systematic study to investigate the following questions:
1) Did the use of these three treatments affect a) the number, b) the type, and c) the success of self-directed revisions that students made on the first draft of their essays?
2) How many self-directed revisions did students make on the teacher-reviewed second drafts?
Method
Participants and Course
Participants in this study included two classes of Japanese second-year university students. The study was carried out in the last of four writing-focused courses that all students were required to take. Class A comprised 24 students, and Class B comprised 23 students. One person from each class failed to complete the course, and one from Class B submitted a plagiarised first draft, leaving 23 students from Class A and 21 students from Class B whose compositions were analysed.
The course consists of two main elements: essay writing and formal oral presentation. The most important component, and the one on which the present research focuses, is a 500-600-word persuasive essay on a topic chosen from a list of 10 titles generated by small group brainstorming sessions.
Research Design
As Dörnyei (2007, p. 117) notes, “In educational contexts, true experimental designs with random group assignments are very rarely feasible.” While the students in Classes A and B had been randomly assigned to these classes, this placement occurred eight months prior to the commencement of this study, and in the intervening period, the two classes were taught the previous writing course by different instructors. This study cannot therefore be classified as experimental research, as the possibility of this differential prior instruction having been responsible for any differences observed between the two groups cannot be ruled out. Due to these issues, this research is best characterized as quasi-experimental, defined by Dörnyei (2007) and Nunan (1992) as being similar to an experimental approach in that it includes a control group but lacks a random assignment of participants. Although concerns have been raised by some regarding the ethics of using control groups in educational research, the method is nevertheless a widely accepted one in the field. In this study, Class A was designated as the control group and Class B the treatment group, receiving the three treatments described in more detail below. Every care was taken to ensure that the different procedures followed were not to the overall detriment of either group.
Schedule
Students wrote four drafts of their essays. After submitting their first drafts, Class A was asked to revise and resubmit with no additional input; Class B received the three treatments and was then asked to revise and resubmit. During these classes, the control group received an additional lesson on presentation skills and gave group presentations on non-essay topics the following week. Written teacher feedback on the second drafts was then provided to both classes in the same manner: comments on both surface and meaning-related issues were given on the same draft, and a simple system of codes was used to identify grammar problems. After returning these drafts, brief individual conferences were conducted during class time focusing on a) checking students could read and understand all the feedback comments, and b) answering any questions they had. Learners then had one week to revise their essays and brought the third drafts to the next class for peer review.
Treatment 1: Presentations
After the first drafts were collected, the remainder of the lesson was spent reviewing presentation skills that had been studied in previous courses. Students then had one week to prepare a five-minute poster presentation on their essay topics that was to be delivered without notes the following week. Learners were divided into groups of five or six and gave their presentations to their group members, who then gave feedback focusing on the presentation skills covered in the previous lesson. Beyond prompting them to re-read their first drafts repeatedly, it was also hoped that the prospect of preparing and giving an oral presentation to their peers would encourage learners to carefully review the content and organisation of their essays.
Treatment 2: Grammar Workshop
Both Ferris (2008) and Hall (1990) claim that instruction on error patterns must be connected to the learner’s own writing if it is to help them make successful revisions. Bearing this in mind, the lesson following the poster presentations involved a grammar workshop comprising 12 sentences in need of correction, each taken from a different student’s first draft and each including one or more common errors in these drafts. Students worked in groups of three or four to try to find and correct as many of these errors as they could, after which the sentences were discussed as a class.
Treatment 3: Checklist
The final treatment was a checklist of 20 statements addressing both general points about essay structure and common problems found in students’ written work. Students were asked to read the statements, act on them if necessary, and check a box to confirm they had done so. This checklist was distributed after students completed the grammar workshop. Rather than using the checklist provided in the course textbook, an original checklist was created for two reasons. The first reason was to cover the specific requirements of this essay and the points we had focused on during lessons; the second was to ask students to check the boxes, sign the sheet, and return it the following week with their second drafts. It was hoped that the checklist would also act as the learner’s statement to the teacher that he or she had indeed done all these things. With regard to this function, it may have been more effective to provide a bilingual version of the checklist; unfortunately, time constraints precluded this.
Analysis of Revisions
At each stage of the drafting process, students were asked to submit their essays both by e-mail and as a hard copy in class. In order to count and categorise revisions, Microsoft Word’s Compare function was used to compare the first and second drafts. The advantage of this function is that it automatically identifies every difference between the two versions, whereas there is always the possibility that a researcher, however diligent, may miss a small change when comparing drafts by the naked eye. However, although this function is invaluable in identifying the location of revisions, it is not useful in either counting or categorising them. Thus, after locating the site of revisions using the Compare document, I returned to the original drafts to analyse them.
The taxonomy of revisions used (Appendix B) is based on those of Cho and MacArthur (2010), Faigley and Witte (1981), and Stevenson, Schoonen and de Glopper (2006). Because all three studies investigated different points and were carried out in very different contexts, none of these taxonomies could be used in their original form. Instead, elements of each were adapted to create a taxonomy suitable for the purposes of this study. While Dimension A of this taxonomy, which categorises revisions as surface or meaning changes, is based on these studies, none of them set out to investigate whether the revisions made by their subjects were successful. Thus, Dimension B, looking at this point, is of my own creation. Conrad and Goldstein (1999) made a broad categorisation of revisions as either successful or unsuccessful; however, it may be of interest to break these categories down further and consider a finer level of detail. Also, it seemed unlikely that all revisions would be amenable to this binary division, and a further category was therefore added to account for this possibility.
All students’ first and second drafts were compared using this taxonomy. Revisions were first categorised according to Dimension A then categorised again according to Dimension B. This two-step process was advantageous because it offered the possibility of reviewing initial judgments when returning to a pair of drafts. After completing the Dimension B categorisation, one final check was made in order to ensure that all decisions were satisfactory. When analysing the revisions made between the second and third drafts, a simpler classification was used, dividing all changes into those which could be attributed to teacher feedback and those which could not.
Results and Discussion
Research Question One
Did the use of these three treatments affect a) the number, b) the type, and c) the success of self-directed revisions that students made on the first draft of their essays?
Table 1. All revisions by type
Revision Type | ||||||
Surface | Meaning | Total | ||||
Class | A | B | A | B | A | B |
Total | 117 | 290 | 54 | 133 | 171 | 423 |
% of total | 68.42 | 68.55 | 31.57 | 31.44 | n/a | n/a |
Maximum | 24 | 26 | 20 | 24 | 26 | 45 |
Mean | 5.04 | 13.86 | 2.35 | 6.33 | 7.39 | 20.14 |
Table 1 shows the number of revisions the two classes made to their first drafts. It can be seen that the students in Class B, the treatment group, made a total of 423 revisions at an average of just over 20 revisions per paper. On the other hand, Class A, the control group, made only 171 revisions at an average of fewer than 7.5 per paper. Per person, then, Class B students made almost three times as many changes to their first drafts as those in Class A did. As noted previously, although students were allocated to these classes randomly, this placement occurred eight months prior to the start of the course in which the research took place, during which time the two classes were taught by different instructors. Moreover, given the wide range of abilities represented in the two classes, the possibility that the overall English level of Class B students was higher than that of Class A cannot be ruled out. Thus, it cannot be claimed with certainty that the three treatments were wholly responsible for the difference in the amount of revision undertaken by the two groups. Nevertheless, in light of the extremely large difference in the number of revisions made, it seems likely that the treatments were effective in stimulating students to revise their papers.
Table 1 also shows that the proportion of revisions made at surface and meaning-levels were almost identical in each group, with both classes making slightly more than twice as many changes to surface aspects of their essays than changes affecting meaning. From this finding, it can be inferred that the treatments did not influence students to focus more heavily on one type of revision than they otherwise would have. Many studies have used such counts of surface versus meaning revisions to suggest that learners have a tendency to focus more on the former (Cresswell, 2000; Fitzgerald, 1987; Sze, 2002). While the statistics in Table 1 might initially appear to support these claims, a word of caution is necessary. To treat one surface-level revision, which may be something as minor as adding a comma, as being equal to one meaning-level revision, which could perhaps constitute the addition of an entire new point, is perfectly sensible when comparing the revisions made by two groups of learners. However, using such a numerical comparison to draw conclusions about the degree to which learners focus on either surface or meaning-level revision makes little sense; it is quite possible that what is counted as a single revision to meaning may be more time-consuming and cognitively challenging than twenty revisions to surface areas. Thus, in the present study, while it can be said that there was no difference between the control and treatment groups in terms of the proportions of surface and meaning revisions, the fact that over two-thirds of revisions were at surface-level must not be taken as an implication that students prioritised local issues in their writing at the expense of global issues.
Table 2. Surface revisions by type
Revision Type | ||||||
Grammar | Vocabulary | Mechanics | ||||
Class | A | B | A | B | A | B |
Total | 56 | 186 | 22 | 38 | 39 | 66 |
% of total | 47.86 | 64.14 | 18.80 | 13.10 | 33.33 | 22.76 |
Maximum | 9 | 22 | 7 | 8 | 18 | 8 |
Mean | 2.43 | 8.86 | 0.96 | 1.81 | 1.70 | 3.14 |
At a finer level of detail, Tables 2 and 3 show that there are several differences in the amount and proportions of different types of revision made. Looking first at surface changes, Table 2 shows that Class B made almost twice as many revisions to vocabulary and mechanics as did Class A. Although statement 1 of the checklist dealt with formatting, nothing else in the treatments specifically targeted mechanics or vocabulary; thus, it may be the case that simply motivating students to re-read their drafts led them to notice problems with these two areas. Grammar, of course, was the sole target of Treatment 2, and it is also mentioned in statements 17 and 18 of the checklist, perhaps explaining the fact that the treatment group made over three times as many revisions to the grammar of their essays as did the control group. Moreover, of the 186 grammar revisions made by Class B, 159 were points covered in the grammar workshop, suggesting fairly conclusively that this treatment was largely responsible for the substantial difference in the number of grammar revisions the two groups made.
Table 3. Meaning revisions by type
Revision Type | ||||||||||||
Organization | Complex repair | Extension | Minor addition | Major addition | Deletion | |||||||
Class | A | B | A | B | A | B | A | B | A | B | A | B |
Total | 10 | 9 | 23 | 47 | 11 | 31 | 1 | 21 | 2 | 9 | 7 | 16 |
% of total | 18.51 | 6.77 | 42.59 | 35.34 | 20.37 | 23.31 | 1.85 | 15.79 | 3.70 | 6.77 | 12.96 | 12.03 |
Maximum | 7 | 2 | 5 | 12 | 3 | 8 | 1 | 4 | 1 | 3 | 6 | 5 |
Mean | 0.43 | 0.43 | 1.00 | 2.24 | 0.48 | 1.48 | 0.04 | 1.00 | 0.09 | 0.43 | 0.30 | 0.76 |
Similarly, differences are apparent when looking in greater detail at meaning-focused revisions (Table 3). In all categories except Organization, Class B made over twice as many revisions, and it is also worth noting that of the ten revisions to organization made by Class A, seven were made by a single person. In Class A, over 60% of revisions dealt with changing existing content (Organization and Complex Repair), and only 25% dealt with adding further content either through the extension of existing points or the addition of new ones. In contrast, extension and addition of content accounted for 45% of the meaning revisions made by Class B. When considering only additions, whether they were minor or major, Table 3 shows that Class B made precisely ten times as many revisions as Class A did. Furthermore, this was not the work of only a few students: 10 of the 21 class members added further content to their second drafts, with another five having extended existing points. Although this cannot be said with certainty, it may be that the prospect of making oral presentations for their classmates stimulated Class B students to work on the content of their essays. Whereas some learners may find teacher feedback motivating, there are certainly others who seem to regard it with indifference; few people, however, feel a similar immunity to the judgment of their peers.
Moreover, while it was stressed to both classes from the outset that the purpose of a persuasive essay is to convince your reader of your point of view, it may be that the concept of the reader is too abstract to create genuine audience awareness among student writers. It seems possible, however, that the concrete reality of having five or six classmates listen to your argument would provide greater incentive to ensure that each point in the essay is both clear and well-supported.
As discussed in the literature review, it seems likely that the careful attention to language form required in revising writing is conducive to language acquisition. If so, then it may be the case that even revisions which do not improve the current text contribute to long-term learning. Nevertheless, the immediate goal of revising writing is, of course, to improve the composition at hand. From the point of view of learners, revisions which do not achieve this goal are likely to be seen as a waste of time and effort. It is thus important to consider not only the number of revisions made, but also whether or not they were successful, which is shown in Table 4.
Table 4. Revisions by effectiveness
1. Improvement | 2. Deterioration | 3. No change | ||||||||||
1a | 1b | 2a | 2b | 2c | 3 | |||||||
Class | A | B | A | B | A | B | A | B | A | B | A | B |
Total | 51 | 211 | 61 | 125 | 2 | 0 | 7 | 22 | 13 | 10 | 37 | 55 |
% of total | 29.82 | 49.88 | 36.67 | 29.55 | 1.12 | 0 | 4.09 | 5.20 | 7.02 | 2.36 | 21.64 | 13.00 |
Maximum | 21 | 23 | 9 | 20 | 1 | 0 | 2 | 5 | 8 | 5 | 8 | 12 |
Mean | 2.22 | 10.04 | 2.65 | 5.95 | 0.87 | 0 | 0.30 | 1.05 | 0.57 | 0.48 | 1.61 | 2.62 |
Note: see Appendix B for explanation of codes
The first point of importance to note is that a large majority of the revisions made by both classes led to improvements on the original drafts. Polio, Fleck and Leder (1998) suggest that “whether or not learners can correct their own errors in the short-term, without feedback, is still an open question” (p. 46). This is therefore an encouraging finding, indicating that when given only indirect input (Class B) and even when given no feedback at all (Class A), learners, if given time to do so, are capable of independently revising their own writing with a good degree of success, not only by correcting errors but also by improving content. Nevertheless, there was a clear difference between the two groups, with almost 80% of Class B’s revisions categorized as improvements, compared to just under two-thirds of those made by Class A. Findings in Table 1 showed that in absolute terms, Class B made 2.5 times as many revisions in total as Class A; when looking at only those revisions which improved the drafts, it can be seen that Class B made four times as many revisions which corrected clear errors (1a) and twice as many that enhanced the style, clarity, or level of detail (1b). Overall, Class B made exactly three times the number of successful revisions: 336 to class A’s 112. This is a considerable difference between two groups of learners which, while not strictly speaking statistically comparable, were certainly extremely similar in most respects. While it remains possible that other factors could have influenced this outcome, it seems likely that the three treatments both encouraged students to make more revisions and provided sufficient guidance on which areas to revise that the changes made by the treatment group were also more likely to succeed.
Research Question Two
How many self-directed revisions did students make on the teacher-reviewed second drafts?
One goal of introducing the three treatments to my writing classes had been to help students discover that they were capable of making improvements to their writing without relying on a teacher’s advice; in addition to improving the current essays, I had hoped to encourage the habit of self-directed revision in the long-term. As mentioned above, feedback was provided on both classes’ second drafts in the same way. Having made a large number of self-directed revisions to their first drafts, it was of interest whether the learners in Class B would continue to do so, or whether they would focus primarily on the teacher feedback. Table 5 shows the number of self and teacher-directed revisions both classes made to their second drafts. It should be noted that even though all revisions that could not be linked to teacher written feedback have been categorized as self-directed, it is possible that some may in fact have derived from other sources. For example, although the mini-conferences focused mainly on the written feedback, I may have given additional advice in response to student questions; moreover, learners may have consulted with others outside class.
Table 5. Self and teacher-directed revisions
First to second draft, self-directed revisions |
Second to third draft, teacher-directed revisions |
Second to third draft, self-directed revisions |
||||
Class | A | B | A | B | A | B |
Total | 171 | 423 | 640 | 682 | 108 | 88 |
% of total | n/a | n/a | 85.56 | 88.57 | 14.44 | 11.43 |
Maximum | 26 | 45 | 56 | 56 | 26 | 23 |
Mean | 7.39 | 20.14 | 27.87 | 32.48 | 4.70 | 4.19 |
As seen in Table 5, revisions to the second drafts were dominated by those which were based on teacher feedback, accounting for over 85% of revisions in both classes. This finding contrasts strongly with results of studies done in ESL settings, such as those of Connor and Asenavage (1994) and Paulus (1999), in which teacher-directed revisions comprised only 36% and 34% respectively of the total. Sze (2002), however, in a detailed study of a learner she describes as a “reluctant ESL student writer,” concluded that “written feedback resulted in more revising than the participant revising by himself” (p. 28), perhaps suggesting that motivation is a key factor influencing self-directed revision. While the two classes in this study certainly contained some highly motivated individuals, it seems reasonable to suppose that these EFL students, who were taking compulsory English classes and had no immediate need for L2 writing skills, are likely overall to be less motivated than those learners who have chosen to attend US universities.
More surprising is the fact that Class A actually made a greater number of self-directed revisions to their second drafts than Class B did. While the difference in itself is fairly small, it may indicate that the treatment group, despite having revised their first drafts extensively, did not gain either skills or confidence in their ability to self-revise through this experience. While this would be a disappointing outcome, it should be noted that another explanation is also possible: it may be that, having made an average of over 20 self-directed revisions on their first drafts, students in Class B had simply already made most of the changes they were capable of. To gain a clearer understanding of whether the treatments served to promote self-editing as a habit, it would perhaps have been preferable to investigate revisions made to a second composition on a new topic; due to the constraints of the course schedule, however, this was not feasible.
One further point of significance emerged from the data presented in Table 5. It was noted previously that factors other than the treatments could be responsible for the large difference in the number of revisions made to the two groups’ first drafts, possibilities including a difference in the overall English level of the two classes and the impact of their having had different writing teachers in the previous semester. It is thus interesting to note that there was in fact very little difference in the average number of revisions each class made to their second drafts, with Class A averaging 32.5 per student (both teacher and self-directed) and Class B averaging 36.7. What is more, two members of Class A made no revisions at all to their second drafts: if these outliers are excluded, the average of Class A rises to 35.5. The strong similarities in the revision behaviour of Classes A and B between the second and third drafts certainly seems to suggest that there were no pre-existing differences in either the ability or inclination of the two groups of students to self-revise. While not conclusive, this finding would appear to provide compelling evidence that the differences observed between the two classes with respect to research question one can likely be attributed to the effect of the three treatments.
Conclusion
The key finding of this research is that the treatment group made almost three times as many self-directed revisions per person to the first draft of their essays as did the control group. While it cannot be conclusively shown that this difference was a direct result of the treatments, the fact that the two groups made a very similar number of revisions on the teacher-reviewed second drafts provides fairly strong evidence in support of this explanation. It was also found that the treatment group made proportionately more additions of new content and revisions to grammar than the control group, and it is tentatively suggested that this was due to the influence of Treatments One and Two, respectively. While two-thirds of the self-directed revisions made by the control group were judged to have improved their essays, this figure rose to 80% for the treatment group, indicating that these learners have the ability to successfully revise their writing without teacher feedback, and perhaps suggesting that the treatments were effective in helping them do so.
Analysis of revisions made to the teacher-reviewed second drafts showed that for both groups, over 85% could be connected to teacher written feedback. While there was only minimal difference between the groups, it is interesting to note that although the treatment group made more teacher-directed revisions, the control group actually made slightly more self-directed revisions on these drafts. Thus, it seems that even if the treatments had encouraged students to independently revise their first drafts, this motivation may not have been carried forward.
Various weaknesses in both the design and implementation of the research have been noted throughout the paper. Nonetheless, several further points are worthy of mention. First, the decision to include three separate treatments meant that it was impossible to isolate the effect of each treatment. As it seems likely that one or more was effective in stimulating self-directed revision, further research into the specific role of each treatment could be worthwhile. Second, no insights have been gained into the process by which learners revise their own writing. While this was not a goal of this study, it would nonetheless be valuable, and further research that perhaps utilises think-aloud protocols or stimulated-recall interviews would be welcome. It should also be noted that counting and categorising revisions proved extremely difficult, particularly in cases where learners had made substantial changes to the content and organisation of their essays, and that whatever taxonomy is used, some decisions will come down to the researcher’s judgment. Most importantly, this study was unable to investigate either the long-term effect of the treatments on students’ ability to self-revise or any language gains made. As Dörnyei (2007) points out, logistical constraints mean that longitudinal research is relatively uncommon in the field of second language acquisition; nonetheless, investigation of the long-term role self-editing of writing can play in language learning would be invaluable.
The findings of this study indicate that when prompted in some way to do so, students are able to revise their own written work with considerable success. This, it is suggested, may have benefits for both learners and teachers, potentially contributing to language acquisition and fostering autonomy for learners and reducing the burden of feedback provision for teachers. Teacher feedback, of course, remains an essential component of writing instruction. It is only in the presence of a reader that writing functions as communication, and, as Goldstein (2004) notes, a trained and experienced teacher is likely able to provide more effective feedback than other readers. However, if teachers can encourage learners to carefully review and revise their written work before submission, the feedback process could become more efficient and productive for both parties.
About the Author
Matthew Coomber has been teaching in Japan since 2001 and is currently Associate Professor at Ritsumeikan University. He holds an MLitt in International Security Studies from the University of St Andrews and an MA in Applied Linguistics and TESOL from the University of Leicester. His research interests include intercultural communication and feedback in academic writing.
References
Bereiter, C., & Scarmalia, M. (1987). The psychology of written composition. Mahwah, NJ: Lawrence Erlbaum.
Casanave, C. (2004). Controversies in second language writing. Ann Arbor, MI: The University of Michigan Press.
Cho, K., & MacArthur, C. (2010). Student revision with peer and expert reviewing. Learning and Instruction, 20, 328-338.
Christiansen, M.S., & Bloch, J. (2016). “Papers are never finished, just abandoned”: The role of written teacher comments in the revision process. Journal of Response to Writing, 2(1), 6-42.
Connor, U., & Asenavage, K. (1994). Peer response groups in ESL writing classes: How much impact on revision? Journal of Second Language Writing, 3(3), 257-276.
Conrad, S., & Goldstein, L. (1999). ESL student revision after teacher-written comments: Text, contexts, and individuals. Journal of Second Language Writing, 8(2), 147-179.
Cresswell, A. (2000). Self-monitoring in student writing: Developing learner responsibility. ELT Journal, 54(3), 235-244.
Cumming, A. (2001). Learning to write in a second language: two decades of research. International Journal of English Studies, 1(2), 1-23.
Diab, N. (2016). A comparison of peer, teacher and self-feedback on the reduction of language errors in student essays. System, 57, 55-65.
Dörnyei, Z. (2007). Research methods in applied linguistics. Oxford: Oxford University Press.
Faigley, L., & Witte, S. (1981). Analyzing revision. College Composition and Communication, 32(4), 400-414.
Fathman, A., & Whalley, E. (1990). Teacher response to student writing: Focus on form versus content. In B. Kroll (Ed.), Second language writing (pp. 178-190). Cambridge: Cambridge University Press.
Ferris, D., & Hedgcock, J. (1998). Teaching ESL composition: Purpose, process, and practice. Mahwah, NJ: Lawrence Erlbaum.
Ferris, D. (2008). Myth 5. Students must learn to correct all their writing errors. In J. Reid (Ed.), Writing myths (pp. 90-114). Ann Arbor, MI: University of Michigan Press.
Ferris, D., Liu, H., Sinha, A., & Senna, M. (2013). Written corrective feedback for individual L2 writers. Journal of Second Language Writing, 22, 307-329.
Fitzgerald, J. (1987). Research on revision in writing. Review of Educational Research, 57(4), 481-506.
Gass, S. (1983). The development of L2 intuitions. TESOL Quarterly, 17(2) 273-291.
Goldstein, L. (2004). Questions and answers about teacher written commentary and student revision: Teachers and students working together. Journal of Second Language Writing, 13, 63-80.
Goldstein, L. (2006). Feedback and revision in second language writing: contextual, teacher, and student variables. In K. Hyland & F. Hyland (Eds.) Feedback in second language writing (pp. 185-205). Cambridge: Cambridge University Press.
Hall, C. (1990). Managing the complexity of revising across languages. TESOL Quarterly, 24(1), 43-60.
Hyland, F. (2000). ESL writers and feedback: Giving more autonomy to students. Language Teaching Research, 4(1), 33-54.
Hyland, K., & Hyland, F. (2006). Feedback in second language writing. Cambridge: Cambridge University Press.
Kamimura, T. (2000). Integration of process and product orientations in EFL writing instruction. RELC Journal, 31(2), 1-27.
Kobayashi, H., & Rinnert, C. (2001). Factors relating to EFL writers discourse level revision skills. International Journal of English Studies, 1(2), 71-101.
Lee, G., & Schallert, D. (2008). Meeting in the margins: Effects of the teacher-student relationship on revision processes of EFL college students taking a composition course. Journal of Second Language Writing, 17, 165-182.
Lee, I. (2013). Research into practice: Written corrective feedback. Language Teaching, 46, 108-119.
Leki, I. (1991). The preferences of ESL students for error correction in college-level writing classes. Foreign Language Annals, 24(3), 203-218.
Makino, T. (1993). Learner self-correction in EFL written compositions. ELT Journal, 47(4), 337–341.
Muncie, J. (2000). Using written teacher feedback in EFL composition classes. ELT Journal, 54(1), 47-53.
Murphy, S. (2000). A sociocultural perspective on teacher response: Is there a student in the room? Assessing Writing, 7, 79-90.
Myhill, D., & Jones, S. (2007). More than just error correction: students’ perspectives on their revision processes during writing. Written Communication, 24, 323-343.
Nunan, D. (1992). Research methods in language learning. Cambridge: Cambridge University Press.
Paulus, T. (1999). The effect of peer and teacher feedback on student writing. Journal of Second Language Writing, 8(3), 265-289.
Polio, C., Fleck, C., & Leder, N. (1998). If I only had more time: ESL learners’ changes in linguistic accuracy on essay revisions. Journal of Second Language Writing, 7(1), 43-68.
Sengupta, S. (2000). An investigation into the effects of revision strategy instruction on L2 secondary school learners. System, 28, 97-113.
Stevenson, M., Schoonen, R., & de Glopper, K. (2006). Revising in two languages: A multi-dimensional comparison of online writing revisions in L1 and FL. Journal of Second Language Writing, 15, 201-233.
Suzuki, M. (2008). Japanese learners’ self revisions and peer revisions of their written compositions in English. TESOL Quarterly 42(2), 209-228.
Sze, C. (2002). A case study of the revision process of a reluctant ESL student writer. TESL Canada Journal, 19(2), 21-36.
Yagelski, R. (1995). The role of classroom context in the revision strategies of student writers. Research in the Teaching of English, 29(2), 216-238.
Appendix A
Please complete this checklist and submit it with your essay next week.
Essay Second Draft Checklist |
Check ✓ |
1. I have read the layout guidelines and example on pages 14-15 of my supplementary materials booklet. My essay follows these guidelines. |
|
2. My essay has four or five paragraphs. | |
3. My essay is at least 500 words. | |
4. My introduction begins with an interesting hook. | |
5. My introduction gives background information about the topic. | |
6. The last sentence of my introduction is my thesis statement. | |
7. My thesis statement answers the question directly. | |
8. My thesis statement includes the topic of each body paragraph. | |
9. My essay has 2 or 3 body paragraphs. | |
10. Each body paragraph focuses on one topic. | |
11. Each body paragraph has a clear topic sentence giving the main point of the paragraph and mentioning a counter-argument. |
|
12. Each body paragraph has at least two different types of support. | |
13. Each body paragraph ends with a concluding sentence. | |
14. The conclusion includes a summary of the main points of the essay. | |
15. The conclusion includes a recommendation. | |
16. The conclusion finishes with powerful final comment. | |
17. I have read every sentence carefully at least twice to check for grammar mistakes. |
|
18. I have checked all sentences starting with So, But or And. | |
19. I have not used computer translation for any part of my essay. | |
20. I have not copied any of this essay from the internet or anywhere else. |
Name: _________________________________
Appendix B
Taxonomy of Revisions
Dimension A (type of revision)
1. Surface changes (changes involving simple repair which do not have a substantial effect on the meaning)
- a) grammar (including changes in tense, agreement, word form, word order, etc), divided into:
- a.i) a point covered in the grammar workshop or checklist
- a.ii) a point not covered
- b) vocabulary
- c) mechanics (spelling, capitalization, format and punctuation)
2. Meaning changes (changes relating to subject matter and ideas)
- a) organization (e.g. moving a clause, sentence or paragraph)
- b) complex repair (clarifying existing points at sentence or clause level)
- c) extension of existing content (e.g. elaborating on or adding an example of an existing point)
- d) addition of new content, divided into:
- d.i) minor (e.g. adding a new supporting point)
- d.ii) major (e.g. adding a new main point)
- e) deletion of content
Dimension B (effectiveness of revision)
1. Revision is an improvement on the original
- a) corrects a clear error
- b) improves the style, level of detail or clarity
2. Revision is worse than original
- a) makes an error worse
- b) introduces an error where none previously existed
- c) has a negative effect on style, level of detail or clarity
3. Revision cannot be judged either better or worse than the original
© Copyright rests with authors. Please cite TESL-EJ appropriately. Editor’s Note: The HTML version contains no page numbers. Please use the PDF version of this article for citations. |