The Effects of Direct and Indirect Corrective Feedback Techniques on EFL Students’ Writing

Abstract

This article presents the results of the comparison between two different Written Corrective feedback (WCF) techniques to 60 low-intermediate EFL students in Karaj, Iran. Assigned to 2 groups, the participants on one group received the direct red pen technique, whereas the participants on the other group received an indirect technique. The participants produced three pieces of writing (pre-test, immediate post-test, and delayed post-test). Simple past tense errors were targeted in the feedback. The study found that the indirect feedback group outperformed the direct feedback group on both immediate post-test and delayed post-test.

Full text

Procedia - Social and Behavioral Sciences 98 ( 2014 ) 445 – 452
Available online at www.sciencedirect.com
1877-0428 © 2014 The Authors. Published by Elsevier Ltd.
Selection and peer-review under responsibility of Urmia University, Iran.
doi: 10.1016/j.sbspro.2014.03.438
ScienceDirect
International Conference on Current Trends in ELT
The Effects of Direct and Indirect Corrective Feedback
Techniques on EFL Students' Writing
Elham Eslami
*
Tehran Institute of Technology, Karaj, 3153785476, Iran
Abstract
This article presents the results of the comparison between two different Written Corrective feedback (WCF) techniques to 60
low-intermediate EFL students in Karaj, Iran. Assigned to 2 groups, the participants on one group received the direct red pen
technique, whereas the participants on the other group received an indirect technique. The participants produced three pieces of
writing (pre-test, immediate post-test, and delayed post-test). Simple past tense errors were targeted in the feedback. The study
found that the indirect feedback group outperformed the direct feedback group on both immediate post-test and delayed post-test.
© 2014 Eslami. Published by Elsevier Ltd.
Selection and peer-review under responsibility of Urmia University, Iran.
Keywords: Written Corrective Feedback; Direct Feedback; Indirect Feedback; Writing Accuracy
1. Introduction
A debate was sparked ever since Truscott published his famous article in 1996, claiming written corrective
feedback (WCF) is ineffective or harmful. The debate lies in whether it is indeed harmful or helpful. If there was one
thing Truscott could profoundly affect, it was to alert the experts in the field to the lack of satisfactory research to
support the efficacy or inefficacy of CF techniques in improving students' writing accuracy. No firm conclusion
could be reached upon the studies with low-quality designs which had also used poor-quality feedback techniques,
and teaching methods and strategies.
Regardless of existing uncertainty about the efficacy of WCF, using them still maintains popularity in the field of
* Corresponding author. Tel.: +98-936-799-1638.
E-mail address: elham_eslami222@yahoo.com
© 2014 The Authors. Published by Elsevier Ltd.
Selection and peer-review under responsibility of Urmia University, Iran.
Open access under CC BY-NC-ND license.
Open access under CC BY-NC-ND license.

446 Elham Eslami / Procedia - Social and Behavioral Sciences 98 ( 2014 ) 445 – 452
teaching English language. Therefore, it finds great importance to discriminate between the effects of all of the
available techniques on different grammatical constructions and to select those that are at an acceptance level of
quality, usefulness, and effectiveness for the context and people with whom they are being used.
Most of the studies on WCF (Bates, Lane, & Lange, 1993; Ferris, 1995; Ferris & Hedgcock, 1998; Hendrickson,
1978, 1980; Lalande, 1982; Walz, 1982) make a distinction direct CF and indirect CF which are two common types
of written corrective feedback. In the case of direct CF the teacher gives the correct form to the students, and it is
desirable for low-level-of-proficiency students who are unable to self-correct and do not know what the correct form
might be. However, it requires minimal processing on the part of the learners and thus, it may not contribute to long-
term learning (Ellis, 2009). A recent study by Sheen (2007) suggests that direct CF can be effective in promoting
acquisition of only specific grammatical features.
Indirect feedback occurs when the teacher indicates in some way that an error exists but does not provide the
correction, thus leaving it to the student to find it. language acquisition theorists and ESL writing specialists alike
argue that indirect feedback is preferable for most student writers, because it engages them in ‘‘guided learning and
problem solving’’ (Lalande, 1982), leading to reflection about linguistic forms that may foster long-term acquisition
(Ferris and Roberts, 2001; James, 1998; Reid, 1998). However, the results of studies that have investigated the
difference between direct and indirect CF are very mixed. Some studies (Ferris & Helt, 2000; Lalande op. cit.)
suggest that indirect feedback is indeed more effective in enabling students to correct their errors, some suggest the
opposite (Chandler, 2003), and others (Robb et al., 1986;Frantzen, 1995) found no difference between direct and
indirect CF.
A further distinction that needs to be examined is the one between ‘unfocused’ and ‘focused’ CF. The former
corresponds to what might be considered normal practice in writing instruction (although not necessarily what L2
writing researchers advocate); teachers correct all (or at least a range of) the errors in learners’ written work. This
type of CF can be viewed as ‘extensive’ because it treats multiple errors. In contrast, focused CF selects specific
errors to be corrected and ignores other errors. Investigations into the most effective ways to provide ESL learners
with WCF have often been overly comprehensive in the range of error categories examined. As a result, clear
conclusions about the efficacy of such feedback have not been possible.
Although it is possible that different CF strategies have different effects on students' writing accuracy, in Iran the
most frequent CF technique, among all the other options (see Ellis, 2009), is the red pen. The researcher asked 85
teachers and 82% stated that they use the red pen technique. Teachers customarily use this technique to give
feedback to the students’ writings by means of writing the correct form of the error using a red pen; in some cases
they also use a meta-linguistic explanation for the most frequent errors. However, there exist many research findings
which indicate that indirect CF techniques are more efficient than direct techniques (Ferris & Helt, 2000; Frantzen,
1995; Lalande, 1982; Lee, 1997; Robb et al., 1986).
The aim of this study is to evaluate the effect of direct and indirect CF techniques on intermediate EFL learners'
writing accuracy. It is of great importance to know whether despite all the efforts teachers put into using direct
techniques and the fact that the red pen technique is extensively used in Iran, it is effective or not. More specifically,
the study aims to answer the following research questions:
1. Is there any significant difference between the impact of indirect and direct CF techniques on low-intermediate
EFL learners' writing accuracy comparing their immediate post-tests? If yes which technique has greater impact?
2. Is there any significant difference between the impact of indirect and direct CF techniques on low-intermediate
EFL learners' writing accuracy comparing their delayed post-test? If yes which technique has greater impact?
3. Is there any significant difference in the immediate post-test and delayed post-test performance of red pen
group?
4. Is there any significant difference in the immediate post-test and delayed post-test performance of indirect
group?

447
Elham Eslami / Procedia - Social and Behavioral Sciences 98 ( 2014 ) 445 – 452
In addition, the following null hypotheses are formulated:
H1: There is no significant difference between the impact of indirect and direct CF techniques on low-
intermediate EFL learners' writing accuracy comparing their immediate post-test.
H2: There is no significant difference between the impact of indirect and direct CF techniques on low-
intermediate EFL learners' writing accuracy comparing their delayed post-test.
H3: There is no significant difference between the immediate post-test and delayed post-test performance of red
pen group.
H4: There is no significant difference between the immediate post-test and delayed post-test performance of
indirect group.
2. Method
2.1. Participants
60 EFL learners participated in the study. They were randomly selected from among 93 students taking part in
low-intermediate courses at an institute in Karaj, Iran. The ratio of male to female participants was equal to avoid
bias caused by sex difference. The sample consisted of a homogeneous group in terms of age, first language, and the
English language background. The average age of the participants was 24.8 . Their first language was Persian, and
their English language proficiency was nearly equal.
2.2. Instruments
Two tests were employed in the present study. The first test was the Cambridge’s Preliminary English Test (PET)
used to ascertain the homogeneity of the participants with regard to their English proficiency. The second was a
writing test package which included a pre-test, an immediate post-test, and a delayed post-test in order to measure
the participants' achievement.
2.3. Reliability of the instruments
The participants' pieces of writing were evaluated and scored by two raters for assessing inter-rater reliability. To
estimate the inter-rater reliability of the test, we calculated the correlation coefficient between the two raters. Table 1
depicts the resulting inter-rater reliability indices.
Table 1. Inter-rater reliability indices.
Note. R1= first rater; R2= second rater. ** p < 0.1.
The test is shown to have very high reliability, 0.937, which is statistically significant at p < 0.1 level of
significance. The content validity of the instruments was also assured by a panel of experts.
3. Procedure
Class sessions were held three times a week for 12 weeks in winter 2012, with each session taking 105 minutes.
The course was incorporated into a competency-based syllabus, the objective of which was to promote the learner's
communicative skills. 93 students form 11 intact classes of “Low-intermediate-Level” were given a PET
(Preliminary English Test) in order to select a homogenized sample in terms of language proficiency. PET consisted
of 69 English language proficiency questions on the four skills of reading, writing, listening, and speaking. Out of
the participants who took the test, the eligible ones (those whose scores ranged from one standard deviation above
CM test R2
CM test R1
.937**

448 Elham Eslami / Procedia - Social and Behavioral Sciences 98 ( 2014 ) 445 – 452
and below the mean on the test) were selected to take part in the study. They were then randomly separated into two
30-member homogeneous groups. The conditions for both groups were exactly the same, except for the method used
for the provision of written feedback.
The experimental group named A received direct WCF in red pen, and those who received indirect CF technique
were considered as the experimental group B. A pre-test, an immediate post-test, and a delayed post-test were run as
part of the writing assessment staged at weeks 1, 12 and 20. The participants were given equal amount of time (15
minutes) for these writing activities.
4. Data Analysis
In order to examine the performance of the two groups in the immediate post-test of writing accuracy and also to
compare the performance of the two groups in the delayed post-test, the researcher ran two independent t-tests.
Furthermore, to examine the performance of the two groups on three different tests (pre-test, immediate post-test,
and delayed post-test), the researcher made repeated comparisons using the matched t-test for each group and then
corrected the significance level through the Bonferroni test.
5. Results
5.1. Performance of the two groups in the immediate post-test
In order to examine the performance of the two groups in the immediate post-test of writing accuracy an
independent t-test was conducted. The results are presented in the table 2.
Table 2. Immediate post-test independent sample t-test.
Levene's Test
for Equality of Variances
t-test for Equality of Means
F
Sig.
t
df
Sig. (2-tailed)
DV Scores
Equal variances assumed
4.084
.048
-7.982
58
.000
Equal variances not assumed
-7.982
49.709
.000
As the table shows, There was a significant difference in the scores for direct (M=12.8, SD=2.8) and indirect
(M=17.8, SD=1.8) groups; t (49.7)= -7.98, p = .000. These results suggest that using an indirect technique really
does have an effect on writing accuracy. Therefore, the participants in the two groups differed in their performance
in the immediate post-test.
5.2. Performance of the two groups in the delayed post-test
A second independent t-test was run to compare the performance of the two groups in the delayed post-test. The
results are presented in table 3.
Table 3. Immediate post-test independent sample t-test.
Levene's Test
for Equality of Variances
t-test for Equality of Means
F
Sig.
t
df
Sig. (2-tailed)
DV Scores
Equal variances assumed
1.181
.282
-10.706
58
.000

449
Elham Eslami / Procedia - Social and Behavioral Sciences 98 ( 2014 ) 445 – 452
Equal variances not assumed
-10.706
55.744
.000
As the table shows, there was a significant difference in the mean scores of direct (M=10.9, SD=2.07) and
indirect (M=17.3, SD=2.5) groups; t (58)= -10.71, p = 0.000. These results suggest that using an indirect technique
really does have an effect on writing accuracy over time. Therefore, the participants in the two groups differed in
their performance in the delayed post-test.
5.3. Red pen corrective feedback and writing accuracy
In order to examine the performance of the two groups on three different tests (pre-test, immediate post-test, and
delayed post-test) given to them during the study, repeated comparisons were made using the matched t-test for each
group and then corrected the significance level through the Bonferroni test. Based on the Bonferroni test, the level of
significance is first decided at 0.05 and then, since there was going to be three comparisons made in this part, the
level of significance was set at 0.05 / 3 = 0.017 (Pallant, 2005). The matched t-tests that were run were interpreted
according to the new level of significance (p < 0.017). The descriptive statistics information for the performance of
group A is provided in table 4.
Table 4. Paired samples statistics of group A.
M
N
Std. Deviation
Std. Error Mean
Pair 1
pre
7.67
30
2.919
.532
immediate
12.82
30
2.866
.523
Pair 2
pre
7.67
30
2.919
.532
delayed
10.89
30
2.073
.378
Pair 3
immediate
12.82
30
2.866
.523
delayed
10.89
30
2.073
.378
As the table shows, this group has the highest performance in the immediate post-test (M=12.82) followed by
delayed post-test (M=10.89) and the lowest performance in the pre-test (M=7.67).
A matched t-test was conducted to see whether or not there was any significant difference in the performance of
the first group in the pre-test and the two post-tests of writing accuracy. The results are presented in the table 5.
Table 5. Paired samples test of group A.
Paired Differences
t
df
Sig. (2-
tailed)
95% Confidence Interval
of the Difference
Mean
Std. Deviation
Std. Error Mean
Lower
Upper
Pair 1
Pre-immediate
-5.151
3.706
.676
-6.535
-3.766
-7.611
29
.000
Pair 2
Pre-delayed
-3.227
2.829
.516
-4.283
-2.170
-6.246
29
.000
Pair 3
Immediate- delayed
1.924
1.952
.356
1.194
2.653
5.397
29
.000
As the table shows, the observed mean differences are significant. This means that participants in the first group
had significantly different performances on the pre-test and the two post-tests. Comparing the mean of the pre-test

450 Elham Eslami / Procedia - Social and Behavioral Sciences 98 ( 2014 ) 445 – 452
and post-tests makes it clear that the participants performed significantly better in the immediate post-test than the
delayed post-test and obviously the pre-test.
5.4. Indirect corrective feedback and writing accuracy
The results of the descriptive statistics of the participants in the second group revealed that the mean scores of the
participants in the second group on the pre-test and two post-tests of writing accuracy were 8.06 and 15.96 and
15.60 respectively. These are presented below in table 6.
Table 6, Paired samples statistics of group B.
M
N
SD
Std. Error Mean
Pair 1
Pre
8.0667
30
.69149
.12625
Immediate
15.9667
30
.80872
.14765
Pair 2
Pre
8.0667
30
.69149
.12625
Delayed
15.6000
30
.77013
.14061
Pair 3
Immediate
15.9667
30
.80872
.14765
Delayed
15.6000
30
.77013
.14061
In order to determine the difference of the performance of the participants in the second group (group B) on the
pre-test and the two post-tests, we ran another matched t-test to compare the mean scores of the participants in the
three tests. The results of the t-test are shown in table 7.
Table 7. Paired samples test of group B.
Paired Differences
t
df
Sig. (2-
tailed)
95% Confidence Interval
of the Difference
M
Std. Deviation
Std. Error Mean
Lower
Upper
Pair 1
pre-immediate
-7.90000
1.02889
.18785
-8.28419
-7.51581
-42.055
29
.000
Pair 2
pre-delayed
-7.53333
1.13664
.20752
-7.95776
-7.10890
-36.301
29
.000
Pair 3
Immediate-delayed
.36667
1.15917
.21163
-.06618
.79951
1.733
29
.094
The results of the matched t-test, as shown in Table 7, revealed that the t-observed values is higher than the
critical value in two pairs (pre-test vs. immediate post-test and pre-test vs. delayed post-test) indicating that the
difference between the performance of the participants in the pre-test and the two post-tests was statistically
significant. This suggests that the participants in the second group benefited from WCF provided. However no
significant difference was found between the performance of the participants of group B on the immediate and
delayed post-tests. This amounts to saying that participants in the second group were able to do equally well in the
delayed post-test.

451
Elham Eslami / Procedia - Social and Behavioral Sciences 98 ( 2014 ) 445 – 452
6. Discussion
As the results of the study show, there exists a strong connection between written corrective feedback and writing
accuracy. Existing theories can support the findings from different aspects. noticing hypothesis proposed by Schmidt
(1990) declares that only the items which are noticed by the learners are probable to be learnt. As a result, since
error feedback attracts learners' attention towards the erroneous linguistic form, it will assist them in taking the pre-
requisite step to develop their interlanguage system. Assimilation theory suggested by Ausubel (1986) states that it
is more effective to relate new concepts and propositions to existing concepts and propositional framework existed
in learners' interlanguage. in this regard, providing CF can be considered as an effective technique in accelerating
the process of assimilation. Sweller (1988) in his cognitive load theory mentions that the ideal learning situation is
to minimize the load of the working memory as much as possible in order to maximize the alternation in long-term
memory. he also believes that it is important to establish a bond between schematic structures of long term memory
and new data, otherwise the learning won't be lasting and learners will most likely forget the material. Since CF
draws learners' attention to their areas of difficulty and releases their minds to process language content, it can be
advantageous. Finally, CF organizes, structures and modifies the knowledge as a kind of scaffold and prompts the
learners to stick the learnt stuff in their long term memory.
This study was conducted to compare the efficacy of the two types of WCF namely direct red pen and indirect
feedback. as the results show, the indirect feedback group proved to be significantly better than the red pen feedback
group on the delayed post-test. This suggests the lasting effectiveness of the indirect WCF over direct red pen
feedback. this is in contrast with Truscott's claim (1999, 2007) that giving feedback has predictable negative effects
on learners’ writing and if advantageous, it is negligible. On the other hand, the outcome of the current research is in
line with the findings of Sheppard (1992), Frantzen (1995), Fazio (2001), and Chandler (2003) who proved CF to be
a way of improving the accuracy of L2 students' writing.
Furthermore, in harmony with several other studies (Ferris & Helt, 2000; Frantzen, 1995; Lalande, 1982; Lee,
1997; Robb et al., 1986), the results show that indirect written corrective feedback leads to either more or equal
levels of accuracy in the long run, which may imply the superiority of the indirect techniques of error correction
over the course of time. Whether indirect feedback is more beneficial than direct feedback is an important issue to
be considered from the pedagogical point of view. Although most teachers regard coding error types to be slower
than just underlining and correcting, once they get familiar to the process it will become much more easier than
doing both spotting and correcting the errors. In addition, applying indirect methods of error correction will
necessarily call for sufficient linguistic knowledge possessed by students to self-correct errors and also getting used
to self edit their own texts. Therefore, using indirect feedback strategies may strongly demand somewhat focused
error correction especially with low-level-of-proficiency learners.
References
Ausubel, D. P. (1986). Educational psychology; a cognitive view. New York: Holt, Rinehart and Winston.
Bates, L., Lane, J., & Lange, E. (1993). Writing clearly: Responding to ESL compositions. Boston : Heinle & Heinle.
Chandler, J. (2003). The efficacy of various kinds of error feedback for improvement in the accuracy and fluency of L2 student writing. Journal
of Second Language Writing 12 , 267–296.
Ellis, R. (2009). A typology of written corrective feedback types. ELT Journal Volume 63 , 97-107.
Ellis, R. (2009). Corrective feedback and teacher development. L2 Journal , 3-18.
Fazio, L. (2001). The effect of corrections and commentaries on the journal writing accuracy of minority- andmajority-language students. Journal
of Second LanguageWriting, 10, 235–249.
Ferris, D. R. (1995). Can advanced ESL students be taught to correct their most serious and frequent errors? CATESOL Journal, 8, 41-62.
Ferris, D. R., & Hedgcock, J. S. (1998). Teaching ESL composition: Purpose, process, and practice. Mahwah, NJ: Lawrence Erlbaum Associates.
Ferris, D. R., & Helt, M. (2000). Was Truscott right? New evidence on the effects of error correction in L2 writing classes. Paper presented at the
American Association of Applied Linguistics Conference. Vancouver, BC, March, 11–14.
Ferris, D.R., & Roberts, B. (2001). Error feedback in L2 writing classes: How explicit does it need to be? Journal of Second Language Writing,
10 , 161– 184.

452 Elham Eslami / Procedia - Social and Behavioral Sciences 98 ( 2014 ) 445 – 452
Frantzen, D. (1995). The effects of grammar supplementation on written accuract in an intermediate Spanish content course. Modern Language
Journal, 79, 329-344.
Hendrickson, J. (1978). Error correction in foreign language teaching: Recent theory, research and practice. Modern Language Journal, 62, 387-
398.
Hendrickson, J. (1980). The treatment of written work. Modern Language Journal, 64, 216-221.
James, C. (1998). Errors in language learning and use. Essex, UK: Addison Wesley Longman.
Lalande, J.F. (1982). Reducing composition errors: an experiment. Modern Language Journal 66, 140–149.
Lee, I. (1997). ESL learners’ performance in error correction in writing. System, 25, 465–477.
Reid, J. (1998). Responding to ESL student language problems: Error analysis and revision plans. In P. Byrd & J. Reid (Eds.), Grammar in the
composition classroom (pp. 118-137). Boston: Heinle and Heinle.
Robb, T., Ross, S., & Shortreed, I. (1986). Salience of feedback on error and its effect on EFL writing quality. TESOL Quarterly 20, 83–93.
Schmidt, R. (1990). The role of consciousness in second language learning. Applied Linguistics,11, 129-158.
Sheen, Y. (2007). The effect of focused written corrective feedback and language aptitude on ESL learners’ acquisition of articles. TESOL
Quarterly 41, 255–283.
Sheppard, K. (1992). Two feedback types: Do they make A difference? RELC Journal , 103-110.
Sweller, J. (1988). Cognitive load during problem solving: Effects on learning. Cognitive Science, 12, 257–285.
Truscott, J. (1996). The case against grammar correction in L2 writing classes. Language Learning 46:2 , 327-369.
Truscott, J. (1999). The case for “The Case Against Grammar Correction in L2 Writing Classes”: A response to Ferris. Journal of Second
Language Writing , 111-122.
Truscott, J. (2007). The effect of error correction on learners’ ability to write accurately. Journal of Second Language Writing 16 , 255–272.
Walz, J. C. (1982). Error correction technique for the foreign language classroom. Language in education: Theor y and Practice Series.
Washington DC: Centre for Applied Linguistics.