Abstract: This study examines the effects of focused and unfocused indirect feedback on the use of simple past tense in L2 writing. About 60 Azari learners of English were assigned to a focused group, an unfocused group and a control group (20 in each). A packet including a short response test, an error correction test, a fill in the blank test and a narrative writing test was given to the participating groups. The groups turned out to be homogeneous in terms of the form in focus. Any past tense used incorrectly in the production of the participants of the focused group was underlined. All erroneous structures in the production of the unfocused group were underlined while the control group received no corrective feedback. The tests which carried feedback for the 2 experimental groups and were untouched for the control group were returned to the participants. The experimental groups were asked to reflect upon the feedback provided. The same packet of tests was given to the groups after 1 week and 3 weeks later. Split-plot ANOVAs run on the data showed that there was no statistically significant difference among the 3 groups of the study, the interaction of time and treatment was not statistically significant while the accuracy of all the 3 groups appeared to improve from time 1-2 but not from time 2-3.

INTRODUCTION

Since L2 learners find writing in an L2 a big challenge (Zacharias, 2007), there have been many attempts to help learners to improve their writing quality and increase their motivation for writing tasks. Providing feedback has been one of these attempts which has already been intimately investigated in different contexts. Feedback is widely seen as a key for encouraging and consolidating learning. In language teaching and learning it is used to facilitate the learning process (Hyland, 1990; Richards and Lockhart, 1994). As Keh (1990) asserts, feedback is a fundamental element of a process approach to writing. She defines feedback in the context of writing as an input from a reader with the effect of providing information to the writer for revision.

Feedback might be operationalized directly or indirectly (Ferris and Hedgocok, 2005; Ferris and Roberts, 2001; Lalande, 1982; Robb et al., 1986; Terry, 1989; Zamel, 1985). Direct corrective feedback can be defined as the provision of the correct linguistic form or structure by the teacher to the learner on the linguistic error. On the other hand, indirect corrective feedback indicates in some way that an error has occurred without drawing attention explicitly (Ferris, 2003).

Another dichotomy is made between unfocused and focused corrective feedback. Unfocused CF corresponds to what might be considered normal practice in writing instruction; teachers correct all (or at least a range of) the errors in learners written work.

In contrast, focused CF selects specific errors to be corrected and ignores other errors (Ellis et al., 2008). There has always been a dispute among researchers and practitioners regarding the role of feedback in helping students to learn how to write.

Although many studies have been conducted to examine this issue, a lot of confusion remains regarding the kind of corrective feedback which can improve their writing process. So, many ESL/EFL writing teachers do not have some clear guidelines at their disposal to help their students.

Although, some researchers (Fazio, 2001; Truscott, 1996) hold negative views regarding the effectiveness of feedback many studies (Chandler, 2003; Ferris, 2003; Lalande, 1982) have reported that feedback on errors can help students improve grammatical accuracy in writing. Debates over the value of providing corrective feedback and the efficacy of certain feedback options on writing have been prominent in recent years, so it calls for further research. The purpose of this study is to examine the effects of focused and unfocused indirect feedback on the use of simple past tense in L2 writing.

Research question: Do focused and unfocused indirect corrective feedback affect accuracy in EFL writing differently?

Research hypothesis: It is predicted that focused and unfocused indirect corrective feedback will affect accuracy in EFL writing differently. Learners are more likely to attend to corrections directed at a single error type and more likely to develop a clearer understanding of the nature of the error and the correction needed (Bitchener and Knoch, 2009). If attention and understanding are important for acquisition, as cognitive theoreticians of L2 acquisition (Ellis, 2005; Schmidt, 1994) have claimed focused CF is more likely to produce positive results.

MATERIALS AND METHODS

Participants: This study was conducted at Zeynabieh High School in Ardebil, Iran. The participants in this study were 60 female EFL learners in 3 classes (their L1 was Azari-Turkish). They were in pre-university grade and aged 17. They were majoring in Natural Sciences and Humanities. The learners received 3 h of English language instruction 2 days a week. They were divided into 3 groups randomly (an experimental group, a contrast group and a control group). Each group included 20 learners. The control group did not receive any feedback. In order to get assurance as to the homogeneity of the learners in terms of the form in focus, we administered a pre-test and put the scores into a one-way ANOVA.

The result indicated that there were no significant differences across the three groups of the participants in their mastery over the target structure (F = 0.003, p = 0.997).

Target structure: Simple past tense (regular and irregular) was chosen as the target structure of the study for 2 solid reasons. Firstly, the learners in the pre-university grade are already familiar with and have explicit knowledge of this structure. The reaserchers did not aim to examine whether corrective feedback assists learning a completely new structure but whether it enables learners to gain greater control over a structure they have already been exposed to. The second reason was that the simple past tense is known to be problematic for learners. So, it was hypothesized that although, learners in this grade would have explicit knowledge of this structure they would make errors in its use. The pre-test demonstrated that this was indeed the case. All the 3 groups completed a pre-test, an immediate post-test and a delayed post-test; tests involved a short response-test, an error correction test, a fill in the blank test and a narrative writing. The short response test included 20 questions about the learners' real life. About 16 out of 20 questions were in simple past tense and 4 questions were in simple present tense. The learners were asked to write a complete answer to each question.

The 2nd test that was an error correction test included 20 sentences. Each sentence contained an error that was underlined. The participants were required to give the correct form of the underlined verb. About 16 out of 20 sentences were in simple past tense. The 3rd test (fill in the blank test) also included 20 sentences. The verb of each sentence was given in a bracket. The learners were asked to give the correct form of the verbs in the bracket. The last test which was a narrative writing, included 6 sequential pictures. The participants were asked to write a short story according to the pictures.

Procedure: The design of this study was pre-test, immediate post-test, delayed post-test and the entire study took 6 weeks. One week prior to the pretest, the teacher and the learners were provided with the essential and related information about the study. The teacher of the participants was instructed not to provide any explicit instruction on the targeted error categories during the period of the study. The learners were not told when they would take the tests. The reason for this was to eliminate the possibility of any student studying the feedback that they had already been given. In the 1st week, the pre-test was administered and the learners in the 3 groups took 4 tests: short response test; error correction test; fill in the blank test and narrative writing test. The teacher collected the tests and provided the treatment.

The experimental group received focused indirect feedback (errors involving the use of past simple tense were underlined). The contrast group received unfocused indirect feedback (all linguistic errors, including past simple tense were underlined. The control group did not receive any feedback. On the next session, the teacher returned to the class and gave back the tests to the learners and asked them (in the experimental and contrast groups) to examine the feedback at home. One week later, the immediate post-test was administered.

The immediate post-test was the same as the pre-test including a short response test, an error correction test, a cloze test and a narrative writing test. The teacher collected the tests and provided the treatment. On the next day, again she handed them back to the learners and asked them to look at the feedback at home. The delayed post-test was administered after 3 weeks to find out the effects of the treatment over time. It also included the same 4 tests used as the pre-test and immediate post-test.

RESULTS

A split-plot repeated measures ANOVA test was run for the short response test to examine the effects of time, feedback type and the interaction of time and feedback type on L2 accuracy (Table 1).

It revealed that the effect of time was statistically significant at 0.05 level (F = 3.67, p = 0.032) while the effect of feedback type (group) was not statistically significant (F = 1.36, p = 0.266). The effect of the interaction of time and feedback type was also significant (F = 3.94, p = 0.005). To report the exact location of the differences across the 3 times (the pre-test, post-test and delayed post-test) a post-hoc LSD was run. It revealed that there was a statistically significant difference between time 1 and 2 (p = 0.009) and between 2 and time 3 (p = 0.043). But there was not any statistically significant difference between time 1 and 3 (p = 0.264).

The result of the split-plot repeated measures ANOVA test run for the Fill in the blank test (Table 2) revealed that the effect of time was statistically significant at 0.05 level (F = 5.33, p<0.008). There was not any significant difference between the groups in the filll in the blank test (F = 0.19, p = 0.827). The effect of the interaction of time and feedback type was not significant (F = 1.61, p = 0.176).

The post-hoc LSD run revealed that there was a statistically significant difference between time 1 and 2 (p = 0.005). The difference between time 2 and 3 was also significant (p = 0.010). But there was not any statistically significant difference between 1 and time 3 (p = 0.612). Based on the results obtained from the split-plot repeated measures ANOVA run for the Error correction test (Table 3), it became clear that the effect of time was statistically significant at 0.05 level (F = 14.86, p<0.001).

The test of between-subjects effect revealed a non-significant difference between the groups (F = 0.28, p = 0.827). The effect of the interaction of time and group was not statistically significant (F = 1.71, p = 0.176).

The result of the post-hoc LSD confirmed that there was a statistically significant difference between time 1 and 2 (p = 0.004), 2 and time 3 (p = 0.003) and time 2 and 3 (p = 0.002). The results of the split-plot repeated measures ANOVA test run on the data obtained from the Narrative writing test Table 4 indicated that the effect of time was statistically significant at .05 level (F = 7.51, p<0.001), the between-groups differences in the Narrative writing test, however was not significant (F = 0.49, p = 0.613).

Table 1:	Split-plot ANOVA for short response test

Table 2:	Split-plot ANOVA for Fill-in-the-blank test

Table 3:	Split-plot ANOVA for Error correction test

Table 4:	Split-plot ANOVA for the Narrative written test

The effect of the interaction of time and group was not statistically significant (F = 1.71, p = 0.153). The post-hoc comparisons revealed a significant difference between time 1 and 2 (p = 0.004) and time 2 and 3 (p<0.0001). The difference between 1 and time 3 was not statistically significant (p = 0.282).

DISCUSSION

Considering the results and interpretations of the tests, the research question can be answered now: focused feedback and unfocused indirect corrective feedback do not affect accuracy (in using simple past tense) in EFL writing differently.

As the hypothesis of the study, it was predicted that the students who would receive focused indirect corrective feedback would outperform those who would receive unfocused indirect feedback but as the findings indicated this hypothesis is rejected. In other words, indirect focused feedback and indirect unfocused feedback have exercised almost an equal degree of effect on the accuracy in using simple past tense in EFL learners' written production.

The results of this study are in line with those of some studies which reported that the corrective feedback on error cannot help the learners improve accuracy in their writing (Bitchener and Knoch, 2009; Fazio, 2001; Truscott and Hsu, 2008). The results, however are different from those of some other studies which reported the efficacy of written corrective feedback (Sheen, 2007; Zacharias, 2007).

Based on the findings of this study, one could easily jump to the conclusion that Truscott (1996) was right when he claimed that provision of corrective feedback is ineffective on FL writing. But it must be noted that many variables such as context, time, age, level of proficiency and the level of explicitness of feedback affect the effectiveness of feedback on writing. Considering all the variables, the unexpected results of the study can be justified through some points.

One of the points in this study might be associated with allowing the participants to take their studys with them and reflect on them. Even the participants in the control group might have become curious about the structures aimed at and have tried to learn them. Moreover, the fact that the focused group received more total corrections than the unfocused group should be taken into consideration. Many learners in the unfocused group made only one or two errors and thus received only minimal correction of any misuse of simple past tense they committed.

However, the fact that the number of corrections of simple past tense in the focused group outweighed corrections of any other single linguistic feature may explain why the post-test differences did not reach significance. So, it is possible that the learners in both groups became equally aware that the focus of the study was on the simple past tense. Even the control group might have had access to the other groups’ corrected study because they were from the same school. It is also possible that the control group have avoided using the structure in focus in the narrative writing test. Truscott (2004) made an important point that correction may cause learners to avoid constructions on which they expect to be corrected, thereby reducing the number of errors they make in these constructions.

Although, no advantage was evident for any of the feedback types in the present study, it can be suggested that as long as some form of the written corrective feedback is provided, it may facilitate acquisition and accuracy improvement with low proficiency learners (Bitchener and Knoch, 2008). It is argued that the learners who are at the lower level of proficiency might take more benefit from the corrective feedback than those who are at the higher level of proficiency.

In other words, it can be said that the most important problem of this study was the ceiling effect. In the view if the learners were at the lower level of proficiency, the result would be different that is indirect focused and unfocused feedback would affect accuracy in using simple past tense in EFL writing differently. Learners are more likely to attend to corrections directed at a single error type and more likely to develop a clearer understanding of the nature of the error and the correction needed (Bitchener and Knoch, 2008). If attention and understanding are important for acquisition, as cognitive theories of L2 acquisition have claimed (Ellis, 2005; Schmidt, 1994), then focused CF is clearly better equipped to produce positive results.

Another important point is that the bulk of the written CF studies has examined unfocused correction. The studies that have shown written CF to be effective have all been much more narrowly focused, addressing the effects of CF directed at specific grammatical features. It remains a possibility that unfocused written CF is ineffective or even damaging as Truscott (2007) suggests while narrowly focused CF is effective. Truscott (1999) himself has acknowledged that it might be possible for highly focused written correction to affect acquisition. We need to know how focused CF needs to be to enable learners to attend to and understand the corrections. Corrections directed at a diverse set of linguistic phenomena are unlikely to foster noticing and cognition that may be needed for CF to lead to acquisition (Ellis et al., 2008).

CONCLUSION

In order to contribute to the need for further research on the value of providing corrective feedback to the learners, the present study was carried out to investigate whether targeted corrective feedbacks on EFL students' writing results in improved accuracy in new pieces of writing over time and whether there is a differential effect on accuracy for different corrective feedback options. Considering the result of the study, it was found that the targeted corrective feedbacks (focused and unfocused indirect feedback) on EFL learners' writing did not result in improved accuracy (in using simple past tense) in new pieces of writing over time. Based on the findings of the study, it became clear that there was not a differential effect on accuracy (in using simple past tense) for different corrective feedback options.