Authors
1 Assistant Professor, Sharif University of Technology
2 MA Graduate of TEFL, Sharif University of Technology
Abstract
Keywords
Introduction
While input, output, and interaction are considered important factors in learning a second language, the role of corrective feedback (CF) on learners’ oral and written linguistic performances in facilitating L2 learning is unquestionable. As a result, CF strategy is an integral topic of teacher training programs. Usually, instructions on feedback strategies depend on the teaching methodology which is used in a particular context, because different L2 teaching methodologies prescribe different error correction strategies (Han, 2002). Feedback practices in a real classroom setting are patterned based on the instructions provided by institutions during their teacher training courses and teachers’ experiences. On the other hand, students’ preferences for error correction cannot be ignored. Although studies which investigated students’ preferences for corrective feedback are scarce, a quick review of literature attests the mismatches between students’ preferences, teachers’ view and real classroom CF moves. While learners consider CF to be an important and essential classroom activity, they evinced disparate views about what type of errors must be corrected and how, when and by whom they expect to be corrected. This difference between students’ expectations and actual classroom practices can eventuate in learners’ demotivation and dissatisfaction and it can eventually impede learning. Therefore, students’ preferences are important factors for deciding on error correction approaches. Based on this, many studies compare students’ and teachers’ views about CF. The gap between teachers’ views or students’ expectations and classroom practices is another area of interest. Learners’ attitude toward error correction techniques can be determined by many factors. For example, studies show that proficiency level affects students’ error correction preferences. Individual differences and cognitive style can be very effective to determine error correction preferences. Given that, this study examined Iranian second language learners’ preferences for oral CF and the extent to which learners’ gender affect learners’ views. For this purpose, this study focuses on the corrective feedback preferences among students in a TOEFL speaking course. The following research questions are raised to be answered:
Background
In learning a second language, corrective feedback (CF) is defined as the information provided by teachers in response to learners’ linguistic errors in their oral or written performance in a second language (Sheen & Ellis, 2011). Based on this information, a learner can confirm, add to, overwrite, tune or restructure the information in the memory (Hatti & Timpeley, 2007). CF has always been an important issue in second language learning and teachers’ education. Different language teaching methodologies prescribe different error treatment strategies. While in the Audiolingual Method which was based on Behaviorist psychology, error correction was emphasized at all costs, naturalistic views of language learning considered error correction to beineffective and even harmful. Communicative Language Teaching, on the other hand, with a focus on fluency and communication of meaning, underestimate the importance of form and accuracy (Han, 2002). However, it does not refute the need for error correction (Russell, 2009). During the 1990s, many experts confirmed the importance of error correction on second language development (Aljaafreh & Lantolf, 1994; Doughty & Varela, 1998; Ellis, 1993, 1994 as cited in Russell, 2009), while many questions regarding error correction remained pending to be answered. Some language learning theories support and explain the role of error correction in L2 learning. These hypotheses include Interaction Hypothesis (Long, 1983, 1996), Output Hypothesis (Swain, 1985, 1995), Noticing Hypothesis (Schmidt, 1994, 2001) and Sociocultural Theory. Each hypothesis supports different forms and strategies for error correction. All of these theories try to find answers to basic questions regarding CF. Questions such as whether errors should be corrected and when, what kind of, how and by whom errors should be corrected.
Regarding the question referring to how errors should be corrected, different CF types have been identified. Based on timing, feedback can be either delayed or immediate. They are also categorized regarding the fact that they address errors either implicitly or explicitly. In the case of implicit feedback, there is no overt indicator that an error has been committed, whereas in explicit feedback types, there is (Ellis, Loewen, & Erlam 2006). Corrective feedback can be either input-providing, i.e. the learner is provided with the correct form, or output-prompting, i.e. the correct form will be elicited from the learner (Lyster & Saito, 2010). As for oral error correction which is the focus of this study, CF may take any of the above-mentioned forms (Sheen, 2010).
The other basic question regarding corrective feedback is whether learners should be corrected. Each teaching methodology has its own views regarding this question. In terms of oral correction, methodologists ascertain the need for oral correction. However, they consider the effect of correction as both beneficial and detrimental (Sheen & Ellis, 2011). Russell and Spada (2006) conducted a meta-analysis on 15 CF studies including 10 oral CF studies in which they reported a very large effect size for the beneficial effect of CF at 1.16. On the other hand, Truscott (1999) claimed that oral CF is harmful because it breaks the linguistic production flow and causes unpleasant feelings. In terms of the CF such as recast, Truscott casts doubt on its effectiveness because of the low possibility of being noticed by the learners. His findings highlighted the importance of the learners’ awareness about the correction moves. Roberts (1995) identified two preconditions for error correction to be effective. First, learners should be aware and recognize that they are being corrected. Secondly, they should understand the nature of their errors so that they can correct them effectively. In another study, Lasagabaster and Sierra (2005) showed a video in which a teacher corrects a student’s error in a classroom setting to both a group of students and a group of teachers. They used retrospection to investigate if the teachers and the students could detect the error correction moves by the teacher in the video. The results showed that students realized 28 percent of the teachers’ correction moves, while the teachers realized 48 percent of the corrections. They concluded that a great portion of the error corrections in the class remains unnoticed by the students. This result indicated the importance of raising students’ awareness of error correction methods and investigating their preferences for error correction strategies. As far as the question of feedback timing is concerned, the kind of oral task in which the learners are involved is important. As a result, for communicative speaking tasks which focus on fluency, delayed feedback, and for the task with more accuracy focus, immediate feedback is considered more appropriate by experts and teachers.
The other question to be addressed regarding CF is the types of errors that need to be corrected. Lyster (1998) investigated the relationship between error types and feedback types. Lyster distinguishes three categories of lexical, phonological, and grammatical errors as the most corrected errors. The results showed that generally teachers were more serious to correct lexical and phonological errors but they were more tolerant about grammatical errors. Lyster asserted that teachers are more inclined to use negotiation of form to correct lexical errors and to use recast to correct phonological and grammar errors. Based on the concerns about what errors to correct, ‘focused corrective feedback’ strategy has been proposed in which a specific error type is to be targeted at a time.
Whether errors should be corrected by teachers, peers or the learners themselves
(self-correction) is another issue that has been investigated. While learner-centered methods advise more learners’ contribution, many studies showed that students prefer to be corrected by their teachers. Learners generally consider error correction as the basic responsibility of teachers. Self-correction, on the other hand, is not always possible because it can only occur when the learner is proficient enough to correct their own mistakes and is not favored by learners (Sheen & Ellis, 2011).
The other aspect of CF which interests L2 researchers has been the discrepancy between teachers’ and students’ perceptions and preferences regarding oral corrective feedback. Schulz (2001) reported a mismatch between teachers’ and students’ belief systems regarding CF and consider this discrepancy as detrimental to learning. In this study, while the majority of the students expected their teachers to correct their oral errors, only half of the teachers under the investigation deemed oral error correction as necessary. Schulz maintained that teachers rely on their institutional prescriptions, their own experiences and their own language learning experiences to develop their patterns of error correction which do not necessarily take students’ preferences into consideration. Schulz claimed that these differences may result in learners’ dissatisfaction and eventually negative effect on learning.
In another study, Lee (2013) investigated the teachers’ and students’ preferences for CF in an advanced level classroom. Lee reported that while recast comprises 92 percent of the error correction, students favored explicit and immediate feedback. Although students expected their teachers to correct them in the middle of their conversation, teachers strongly disagreed with the correction of all oral mistakes. The data for this study were obtained through classroom observation, a questionnaire and follow-up interviews. The interviews revealed interesting reasons for students’ preferences. According to the results, based on in-depth interviews, students preferred explicit correction because when their teachers indicated their problems directly they can clearly understand and quickly correct their mistakes. Moreover, they considered this correction more helpful and accurate for finding correct answers. On the other hand, they described clarification requests as a vague and unclear technique which results in unpleasant feelings and hamper their desires to continue their speech.
Lasagabaster and Sierra (2005), on the other hand, identified some overlaps between students’ and teachers’ perception of error correction. They reported that both groups agreed that quick corrections were not very effective and more time should be allotted to each error to be corrected. Moreover, they favored more selective correction so that the conversation flow would not be interrupted so frequently. Chunhong and Griffiths (2012) investigated students’ preferences for corrective feedback both qualitatively and quantitatively through questionnaires and interviews. The results showed that despite attesting the efficacy of error correction on their language learning progress, students preferred immediate over delayed and explicit over implicit correction and teacher correction over peer or self-correction. They perceived error correction in front of their classmates embarrassing, though.
Iranian Studies on CF Preferences
In the Iranian setting where many private institutes try to adopt Communicative Language Teaching and more learner-centered methods are used, CF is taking on a more significant role. Kaivanpanah, Alavi, and Sepehrinia (2012) explored learners’ preferences in corrective feedback and compared these preferences with teachers’ views in an Iranian context. They further examined how proficiency level might affect students’ attitudes towards feedback. The results showed that proficiency level and affective reaction and beliefs influence learners’ views about feedback. Results revealed that students at higher proficiency levels prefer more self-correction and elicitative methods for error correction, whereas, lower levels prefer more metalinguistic correction. Students did not have a negative attitude towards peer correction. On the other hand, their teachers believed that students had more negative attitudes toward peer correction. Iranian EFL learners’ preferences for oral corrective feedback have been studied by Mohseni and Edalat (2012). In this descriptive qualitative study, students showed more preferences for immediate correction of pronunciation and delayed correction of grammatical errors. Furthermore, they evince more positive views for teacher and self-correction over peer correction. Moreover, in this study, learners favor explicit correction in which teachers provide students with the correct forms. The study identified a significant discrepancy between students’ preferences and teachers’ views and error correction practices. This mismatch between students’ expectation and teacher views and practices has been regarded as a threat to learning.
Given the above review on CF preferences, it can be concluded that individual differences and preferences should be taken into account to make appropriate CF moves in second language classrooms. Course objective is also a factor that affects learners’ preferences and consequently, feedback practices in general. The variables that may affect students’ views and preferences should be identified and taken into consideration. From among many individual differences that may affect learners’ CF preferences, the effect of learners’ gender remains unknown. For this reason, this study will examine student preferences for corrective feedback in a TOEFL speaking course. The purpose of this study is to investigate how course objectives and students’ gender affect CF preferences among Iranian EFL learners.
Methodology
EFL learners may have different preferences for different aspects of CF. These aspects may include feedback types they most favor to receive or feedback timing they prefer. Moreover, they may have different ideas regarding by whom and for what kind of errors they like to be corrected. In order to understand the CF preferences of Iranian EFL learners and the effect of gender and course and task objectives on their view, a 30-item-questionnaire was used. Students marked their answers on a five-likert scale ranging from 5 ‘totally agree’; 4 ‘agree’; 3 ‘agree to some extent’; 2 ‘disagree’; 1 ‘totally disagree. The results were analyzed quantitatively using both descriptive and inferential statistics to probe the students’ preferences for corrective feedback. Moreover, an independent Sample t-test was run to compare the mean scores of male and female students and to find out the effect of gender on EFL learners’ preferences.
The Participants
A total of 32 male and female EFL learners between 20-27 years of age participated in this study. Non-random procedure was used for sampling and students in TOEFL speaking course from an intact class participated in the study. According to the related CF literature, proficiency level proved to be an effective factor in determining EFL learners’ CF preferences. However, the participants for this study were mainly from intermediate and advanced proficiency levels and therefore this variable was controlled.
Table 1. Participants’ Information
|
|
Male |
Female |
total |
|
Count |
22 |
10 |
32 |
|
Percentage |
68.8 |
31.3 |
100.0 |
Instrument
A 30-item questionnaire (see Appendix) was used to elicit EFL learners’ preferences about oral CF. The developed questionnaire came in two parts. The first part elicited demographic information including their age, gender, proficiency level, and length of experience in language courses. In part two, the items focused on four main aspects of CF including what errors need to be corrected, who should correct errors and how and when errors should be corrected. Students answered each item on a five likert scale ranging from 5 ‘totally agree’; 4 ‘agree’; 3 ‘agree to some extent’; 2 ‘disagree’; 1 ‘totally disagree. The items in this section were adopted from an established questionnaire developed and used by Kaivanpanah et. al. (2012). In their study, the reliability of this questionnaire was estimated at 0.70, using the ‘Cronbach’s alpha. Some of the items in this questionnaire were eliminated and some were modified based on the review of the literature and the purpose and context of the study. The questionnaire was translated into Persian and both English and Persian versions are appended.
The items in the questionnaire can be classified into 3 scales:
Results
The data were analyzed using SPSS software. In order to answer the first research question ‘What are the Iranian EFL learners’ preferences for oral corrective feedback?’ descriptive statistics analysis was used. The mean, standard deviation of each questionnaire item, and in some cases, the scale of the items revealed the learners’ preferences for CF. In order to answer the second research question: ‘Does age affect Iranian EFL learners’ oral corrective feedback preferences?’ an Independent Sample T-test was run to compare the mean scores of male and female participants to probe the effect of gender on EFL learners’ preferences for CF.
Descriptive analysis of the data related to the first scale ‘who should provide corrective feedback’ (Table 2) shows that teachers’ correction is the most favored one followed by peer correction and self-correction, respectively.
Table 2. Descriptive Statistics of Students’ Ideas of Who Should Provide Feedback
|
Self-correction |
Mean |
Std. Deviation |
|
Everyone should care about correcting his/her own errors |
3.78 |
1.211 |
|
I prefer the teacher/classmates to repeat the utterance up to the erroneous part and wait for self-correction |
3.19 |
1.061 |
|
I think self-correction is not a good idea because I often can’t find my errors, if I can, I also don’t know how to correct it. |
2.87 |
1.238 |
|
Total |
3.28 |
|
|
Teacher Correction |
|
|
|
Only the teacher has the knowledge to give feedback |
2.91 |
1.058 |
|
Learning is more effective when the teacher corrects the errors |
4.09 |
.818 |
|
I prefer the teacher to explain the errors the classmates have pointed out |
4.06 |
.982 |
|
When the teacher corrects the errors, one is less stressed than when the classmates |
3.22 |
1.211 |
|
Total |
3.57 |
|
|
Peer Correction |
|
|
|
The classmates can provide better feedback as they might know points I might be unaware of |
3.31 |
.931 |
|
The classmates are sincere in correcting others’ errors |
3.44 |
.982 |
|
Then the classmates correct the errors, one does not feel humiliated |
3.53 |
1.295 |
|
The classmates have the competence needed for correcting others’ Errors |
3.00 |
.950 |
|
Total |
3.32 |
|
|
N=32 |
|
|
On the scales, related to the method of corrective feedback (Table 3), explicit corrective feedback as well as a delayed explicit corrective feedback were the most favored. In contrast, students show the least preference for immediate corrections.
Table 3. Descriptive Statistics on Feedback Types
|
Types |
Mean |
Std. Deviation |
|
I prefer classmates/teacher to correct the error and explain about the error |
4.09 |
.928 |
|
I prefer the classmates/teacher to provide some wrong and correct examples like ‘‘he go or he goes’’ and ask me to choose the correct answer |
3.47 |
1.270 |
|
I prefer the classmates/teacher to repeat the erroneous part with a rising intonation helping one notice the error |
3.31 |
1.176 |
|
I prefer the classmates/teacher to repeat the erroneous part of my utterance with an interrogative tone, so that I understand and correct my mistake |
3.28 |
1.198 |
|
I prefer the teacher/classmates to point out the errors and prompt for self-correction |
3.41 |
1.073 |
|
I prefer the teacher/classmates to repeat the whole utterance but stress the erroneous part for easier noticing |
3.37 |
1.100 |
|
I prefer classmates/teacher to correct my errors immediately |
2.69 |
1.281 |
|
The teacher/classmates should explain about my grammatical errors at the end of conversation |
3.97 |
.933 |
|
N=32 |
|
|
Table 4 shows that students believe that all types of errors need to be corrected. Students considered the grammatical errors as the most important errors that need to be corrected followed by vocabulary errors and pronunciation errors.
Table 4. Descriptive Statistics of Types of Errors that Should be Corrected
|
What Errors Should be Corrected |
Mean |
Std. Deviation |
|
All Grammatical errors should be corrected |
4.47 |
.842 |
|
All Pronunciation errors should be corrected |
4.06 |
.914 |
|
All Vocabulary errors should be corrected |
4.44 |
.619 |
|
Only The errors that impede the flow of communication should be corrected |
3.50 |
1.414 |
|
The errors which are regularly repeated by the students should be corrected |
4.63 |
.833 |
|
The errors which their language forms have already been taught should be corrected |
4.34 |
.745 |
|
The errors which would be fossilized if not corrected should be corrected |
4.72 |
.581 |
|
The frequent slips of the tongue should be corrected |
3.25 |
1.270 |
|
Correcting all types of errors. |
3.63 |
1.185 |
|
No error correction while speaking |
2.66 |
1.494 |
|
N=32 |
|
|
In order to investigate the relationship between gender and error preferences, an independent Sample T-Test was run to compare the mean scores of male and female participants on the three different scales of preferences ‘who should provide the corrective feedback?’, ‘preferred corrective feedback’ and ‘types of error that needed to be corrected’.
The results of the independent Sample T-Test show that there is a significant difference in the mean scores for males (M= 3.08, SD= .816) and females (M=3.73, SD= .979); regarding self-correction view with females having a more positive attitude towards self-correction; t (30) = 1.986, p = .05, 2-tailed).
In terms of teacher correction, there is a significant difference between male (M=3.74, SD= .374) and female (M=3.20, SD= .734) preferences; t(30)= 2.773, p=.009, 2-tailed). Male students showed stronger interest for teacher correction.
On the other hand, comparing male and female students mean scores on their preferences for peer correction showed that there is a significant difference between male (M= 3.13, SD=.596) and female (M=3.75, SD=.717) preferences; t(30) = 2.581, p= .030, with females having more positive attitudes towards their peers’ correction.
In terms of differences between male and female preferences regarding corrective feedback types, which is manifested through items 13 to 20, there is no significant difference between male (M=3.51, SD=.512) and female students (M=3.33, SD=.531); t(30)=.902, p=.38, 2-tailed).
Regarding male and female preferences or the type of errors that need to be corrected there is a significant difference between male (M= 3.86, SD=.44) and female (M=4.20, SD=.43); t(30)=2.03, p=.05) with females showing stronger inclination to be corrected more often than male students.
Discussion and Conclusion
Students in a TOEFL preparation course are usually at intermediate and advanced levels. It seems that in TOEFL speaking tasks, the main goal is to show an acceptable oral proficiency level with great emphasis on accuracy. Accordingly, students in these courses care much more about their accuracy while the fluency for these students is of secondary importance. Based on the purpose of the course, corrective feedback is of great importance to the students and they expect that CF can help them to develop an oral skill to perform the tasks on the TOEFL speaking section. It could be the importance of accuracy that makes the students welcome corrective feedback either in the form of teacher correction, self-correction, or peer correction. Students generally showed more reliance on teacher correction which is consistent with the previous research by Mohseni and Edalat (2012). On the other hand, there are differences between male and female learners’ views regarding who should correct them. While female students favor self-correction and peer correction, males are more dependent on their teachers to correct them.
In terms of correction type, explicit and delayed corrective feedback are the most preferred error correction forms. Students’ preference for explicit error correction is inconsistent with Mohseni and Edalat’s (2012) findings that students favored explicit correction for which the teacher provided them with additional explanation. However, immediate correction was the least favorite CF among the students for they do not want an interruption in the flow of their speech which Truscott (1999) described as harmful for language acquisition and what causes unpleasant feeling in learners. The results are also in agreement with Lasagabaster and Sierra’s (2005) findings that both students and teachers agreed that quick corrections are not very effective. Lee (2013), on the other hand, reported that in an advanced level classroom, students favored explicit and immediate feedback. While in this study explicit feedback received the same credit, immediate feedback was not regarded as effective by the students. This can be attributed to the nature of TOEFL speaking tasks in which tasks need to be done within certain time limitations and students preferred not to be interrupted when they were practicing for the tasks. In terms of the feedback forms, male and female students did not differ significantly and they showed the same preferences for the correction type. Therefore, while gender can affect learners’ preferences as to who provides them with the corrective feedback, it does not affect students’ views about the feedback types.
On the other hand, students believed that all their errors must be corrected. They consider their grammatical errors as the most important one to be corrected followed by vocabulary and pronunciation errors. Learners’ preferences in this regard are in conflict with the actual classroom feedback practices, because based on Lyster’s (1998) investigation generally teachers were more serious to correct lexical and phonological errors but they are more tolerant about grammatical errors. The mean score for the items related to types of error to be corrected recorded the highest average which indicates the importance of accuracy for the students attending the TOEFL speaking preparation course. It seems that students do not want any of their errors to be ignored, because rather than success in the communication of meaning, they are more concerned about accuracy to ace their tests.
In sum, it could be concluded that students’ oral feedback preferences are affected by the course objectives. Given that, in test-wise language instruction, students’ obsession with accuracy brings about higher demands for feedback on all types of error using a method which best highlights the error and makes the learners aware of the correct form. The findings of this study urged teachers to consider the course and task objective into account when deciding on correction moves. Gender is another factor which can partially affect the student’s preferences. In this study, students’ preferences for who provides them with the corrective feedback was affected by the students’ gender. Attitudes to different feedback types and types of errors that they prefer to be corrected were mostly affected by the nature and the objective of the tasks and the course in general.
This study exclusively focused on TOEFL speaking task preparation course, however, further works need to be done to compare students’ oral corrective feedback preferences across test-wise tasks and communicative tasks. The effect of feedback types on students’ performance can also be examined to determine if considering students’ preferences would eventuate in a better result. Moreover, qualitative investigations such as interviews can be done to discover students’ reasons behind their preferences.
)