The interplay of oral corrective feedback and L2 willingness to communicate across proficiency levels

Abstract

This study aims to explore the impact of oral corrective feedback types on English as a foreign language (EFL) learners’ willingness to communicate across proficiency levels. It also investigates how EFL learners view different types of feedback in relation to their willingness to communicate. Sixty Iranian EFL learners were tracked in four proficiency levels. Initially, the participants filled in a questionnaire to measure their attitudes to oral CF and their willingness to communicate. Subsequent to the teachers’ employment of explicit correction, recasts, and prompts, the learners’ willingness to communicate was measured anew. A semi-structured interview was also conducted. The results revealed learners’ high preference for prompts. A two-way mixed between-within ANOVA demonstrated a significant effect for both oral corrective feedback and proficiency level on willingness to communicate. Furthermore, elicitative types of feedback were ranked as the most contributory feedback type to L2 willingness to communicate.

Keywords

L2 willingness to communicate learners’ views oral corrective feedback proficiency level

I Introduction

Throughout the history of English language teaching, communication has been the prime motive for learning languages. People learn languages to exchange information with speakers of other languages. However, not all learners are equally communicatively competent. Despite having high levels of linguistic knowledge, learners are sometimes reluctant to take part in communication. On the other hand, some students with far less linguistic competence efficiently participate in the exchange of information in various communicative situations (MacIntyre, Clément, Dornyei & Noels, 1998). Willingness to communicate (WTC) is a construct that accounts for speakers’ inclination to initiate a conversation. WTC, first developed in first language communication research (McCroskey & Baer, 1985), was defined as the probability to engage in communication when given the choice (McCroskey & Richmond, 1990). This construct was later applied to a second language context to examine the factors that contribute to learners’ psychological readiness to initiate communication (MacIntyre & Charos, 1996). Studies exploring the multifaceted nature of WTC led to a multilayered pyramid-shaped model of WTC proposed by MacIntyre et al. (1998). L2 WTC was defined as ‘a readiness to enter into discourse at a particular time with a specific person or persons, using an L2’ (MacIntyre et al., 1998, p. 547). In explaining the social situation of their Pyramid model, MacIntyre et al. (1998) enumerate five central components, namely participants, setting, purpose, topic, and channel of communication as particularly relevant. Due to the pivotal role that teachers and peers play inside the classroom in leading the students to take part in communicative acts (Kang, 2005), in this study, we focused on the participants component of the model.

Previous research has shown that topic, interlocutors, and conversational context (Kang, 2005) affect L2 WTC. In the classroom, the interlocutors, that is, teachers and peers, can influence numerous aspects of the language learning process including L2 WTC. One of the ways whereby teachers can affect learners’ L2 WTC is their reaction to the learners’ erroneous utterances in the form of corrective feedback (CF). The way teachers treat their students’ errors has been found to influence the learners’ L2 WTC (Kang, 2005). Although many studies (Li & Vuono, 2019; Saito & Lyster, 2012; Yang & Lyster, 2010) have confirmed the facilitative role of CF, Krashen (1982) and Truscott (1999) challenged its efficacy in improving learners’ linguistic abilities. While Truscott (1999) claimed that correction interrupts classroom activities and disturbs ongoing communication processes, Krashen (1982) believed in the entire elimination of correction in communicative-type activities. As an attempt to fill the gap in the effect of oral corrective feedback (OCF) on L2 WTC, this study, drawing on Lyster and Ranta’s (1997) classification of OCF types, explores the interaction between types of OCF, that is, explicit correction, recasts, and prompts (including metalinguistic feedback, clarification request, repetition and elicitation) and Iranian EFL learners’ WTC across proficiency levels (PLs). It also intends to find out how Iranian EFL learners view various aspects of OCF, per se, and in relation to L2 WTC.

II Literature review

1 Learners’ views about OCF

Learners’ perception of OCF can affect their responses to OCF and the strategies for L2 learning (Kartchava & Ammar, 2004). Horwitz (1988) noted that learners’ beliefs underlie learner behavior to a large extent. Concerning the importance of CF preferences, Lyster, Saito and Sato (2013) assert that they inform practitioners of the learner’s perspectives and this might result in more effective teaching practice. Chenoweth, Day, Chun and Luppescu (1983), in an early study of CF preferences, reported that ESL learners (in the United States) held a generally positive attitude toward error correction and claimed to prefer even more correction than their native speaker friends did. Similarly, Schulz (1996, 2001) found a strong preference for CF expressed by both American and Colombian ESL students. In Japan’s EFL context, Katayama (2007) also administered a questionnaire to 588 students at several Japanese universities and found that students demonstrated strongly positive attitudes toward teacher’s correction of errors. In another study, Lasagabaster and Sierra (2005) concluded that learners (and teachers) shared the view that the most efficient corrections occurred when more time, longer explanation, and the use of different correction strategies were utilized. Furthermore, in exploring the teachers’ and learners’ preferences of CF types in Japanese as a foreign language classroom in Australia, Yoshida (2008) reported that most learners preferred to have an opportunity to think about their errors and correct forms before receiving the correct forms through recasts. In another comparative study of EFL teachers’ and students’ attitudes toward OCF, Roothooft and Breeze (2016) found that students were much more positive about explicit types of CF than their teachers and that they claimed to experience positive emotions when receiving CF. In addition, Kaivanpanah, Alavi and Sepehrinia (2015) compared Iranian learners’ and teachers’ views about OCF and reported that the more proficient learners favored elicitative types of feedback that required self-correction. Two other studies examined learners’ views about OCF in relation to PL. Brown (2009) compared first- and second-year university students and found that the second-year more advanced students had a stronger preference for indirect types of CF than did the first-year students. Also, juxtaposing the role of PL with the cultural background and the types of errors in shaping the learners’ OCF preferences, Yang (2016) concluded that, as opposed to the beginning level learners, the intermediate level learners considered clarification requests more effective on pronunciation errors. Finally, in a recent study, Zhu and Wang (2019) investigated 2670 Chinese EFL learners’ views about OCF and indicated that participants had an overall positive attitude toward CF. Learners showed preferences for immediate CF over delayed CF and output-promoting CF (e.g. prompts) over input-providing CF (e.g. explicit correction).

2 PL and L2 WTC

MacIntyre’s (1994) model suggested that WTC is caused by a combination of communication apprehension and perceived competence, which in turn have their origins in introversion and self-esteem. Informed by this and similar studies which focused on perceived competence, researchers drew on the connection between PL and L2 WTC. For instance, Matsuoka (2005), exploring WTC among Japanese college students, reported a complicated relationship between PL and L2 WTC. She concluded that L2 WTC predicted proficiency, though PL was not a significant predictor of L2 WTC. In another study, Liu and Jackson (2009) argued that higher-PL learners were most willing to engage in conversations in English classes as opposed to lower-PL learners. They ascribed the results to the less importance lower-PL learners attached to interpersonal interaction and speech communication in the classroom. Though two studies conducted in the Iranian EFL context, that is, Alemi, Tajeddin and Mesbah (2013) and Khajavy, Ghonsooly and Fatemi (2014), maintained that higher-PL learners demonstrated higher levels of L2 WTC, the latter study highlighted an indirect path from PL to communication confidence and then to L2 WTC. They noted that learners with higher PL perceived themselves as more competent and less anxious to communicate in English, which, in turn, affected their L2 WTC.

3 Research on OCF

A plethora of studies to date has investigated the effects of teachers’ CF on L2 development and has demonstrated that OCF facilitates L2 development (Li, 2010; Li & Vuono, 2019; Lyster & Saito, 2010; Russell & Spada, 2006). However, there are certain variables that mediate the effectiveness of CF, including CF types (Nassaji, 2009) and individual differences (Rassaei, 2015; Sheen, 2008; Zhang & Rahimi, 2014). Of the CF types, recasts have been the most studied type. In their often cited study, Lyster and Ranta (1997) reported that recasts accounted for more than 55% of all the CF provided by the teachers. They maintained that despite being the most popular, recasts are the least likely type of CF to lead to uptake of any kind. This has been attributed to the fact that ‘learners do not necessarily notice the modification’ (Lyster and Ranta, 1997, p. 57). Thus, changes in recasts are either imperceptible or perceived as merely an alternative (Ammar, 2008; Panova & Lyster, 2002). Contrary to recasts, which provide learners with the correct form, prompts provide negative evidence and set the scene for self-correction (Lyster et al., 2013). Individual differences also mediate the effectiveness of CF and interact with CF in L2 development. Li (2010) and Lyster and Saito (2010) argued that the facilitative role of OCF may be constrained by contextual factors and individual learner differences. Similarly, Sheen and Ellis (2011) stated that ‘individual difference factors [including PL “my parentheses” comes after “including PL” rather than after page number. That is, (including PL, my parentheses).] and contextual factors will clearly influence whether, how and when OCF is effective’ (p. 604).

4 OCF and L2 WTC

As the literature indicates, when considering the role of interlocutors in developing EFL learners’ WTC, researchers have rarely targeted the role of OCF offered by the teachers and its impact on the students’ L2 WTC in the classroom context. In their study, Tavakoli and Zarrinabadi (2016) explored the effects of implicit (recast) and explicit (explicit correction and metalinguistic feedback) CF on learners’ L2 WTC and concluded that while implicit CF did not influence L2 WTC, explicit CF increased it through promoting learners’ self-confidence. Some other studies, though not initially intended to examine the connection between OCF and L2 WTC, have yielded interesting results as to how OCF and L2 WTC are related. For instance, Kang (2005), in her qualitative study, found that situational L2 WTC emerged from the joint effect of three interacting psychological conditions of excitement (a feeling of elation about the act of talking), responsibility (a feeling of obligation or duty to deliver and understand a message, or to make it clear), and security (feeling safe from the fears that nonnative speakers tend to have in L2 communication). Explicating the influence of interlocutors on security, Kang (2005) asserted that interlocutors were key determinants of security. She maintained that the learners’ insecurity originated from the participants’ fear of losing face by making mistakes. Besides, MacIntyre, Burns and Jessome (2011) asked the French immersion students to describe the situations in which they were the most and the least willing to communicate through the focused essay technique. They found that error correction was a major and influential factor in L2 WTC, although they failed to draw a general inference on the facilitative or debilitative role of CF in L2 WTC. Furthermore, focusing on the effect of teachers on learners’ WTC, Zarrinabadi’s (2014) qualitative study depicted that error correction, teachers’ support, decision on the topic, and the wait time are the main ways teachers influence learners’ WTC. Thus, some studies have found positive effects of (some types of) OCF on WTC, while others are less clear about the nature of the effect of OCF on WTC. A careful look at the CF literature reveals that previous studies have mostly focused on either the effect of some (not all) OCF types or the relationship between OCF and other linguistic and cognitive aspects of second language learning, but very few studies have looked at the relationship between OCF and WTC.

In addition, in line with the recent changes of focus in the Iranian national EFL syllabus from grammar and reading comprehension to communication, the strategies to foster learners’ communicative competence have taken priority. All the educational activities in EFL classes are supposed to be directed at reaching higher levels of communication on the part of the students (Iranian Ministry of Education, 2011). To achieve this goal, OCF as an integral part of EFL classes and its impact on learners’ L2 WTC has received a central role in the recent syllabus (Iranian Ministry of Education, 2011). Intending to bear practical implications for the communication-oriented EFL syllabi in Iran and elsewhere, the present research aims to explore how Iranian EFL learners view different types of OCF and how OCF types affect their WTC across the PLs.

The following research questions are addressed:

What are the views of EFL learners about teachers’ OCF and the tie between OCF and L2 WTC inside the classroom?

Do OCF types (explicit correction, recasts, and prompts) significantly affect L2 WTC of EFL learners in the starter, elementary, pre-intermediate, and the intermediate levels of language proficiency, and is there an interaction effect between OCF types and PL?

III Method

1 Participants

Sixty Persian-speaking female participants were recruited from an English language institute in the west of Iran. The learners were grouped by the institute into four PLs; starter, elementary, pre-intermediate, and intermediate, based on their scores on the Oxford placement test (Allan, 2004). Their age varied from 11 to 22 years (M = 16.8). An intact class from each PL was randomly selected and the three OCF types were respectively provided to the students (15 in each class) during the semester. The Touchstone series (McCarthy, McCarten, & Sandiford, 2006) was taught at the institute. The overall lesson structure and content were similar for the most part. For instance, similar topics were chosen and presented to the learners in all groups in the free discussion section of the course. The starter-level students were in their third semester of the institute English syllabus and had earlier studied course books such as First Friends (Lanuzzi, 2015), Get Ready (Hopkins, 2003) and English Time (Rivers & Tayama, 2015) prior to the present course. They studied Touchstone One during the semester. They had at least 2 to 7 years of experience in studying English at the state schools, 2 to 4 hours a week. The students at the elementary, pre-intermediate, and the intermediate levels were respectively in their 6th, 9th and 12th semester of the English program followed by the institute. The program was predominantly conversation-centered, and the oral exchange of information, and consequently OCF, was an integral part of it. A male teacher with a master’s degree in TEFL joined the first author in teaching the learners. To maintain consistency in offering OCF, each instructor focused on two specific PLs. That is, the first author instructed the starter and the elementary level learners, and the other instructor engaged in teaching the pre-intermediate, and the intermediate level learners. Prior to the study, consents were obtained from the learners and their parents.

2 Instruments

a Survey questionnaire

To obtain the students’ views about OCF, a modified version of Katayama’s (2007) questionnaire was adopted (Appendix A). The original questionnaire, utilizing a 5-point Likert scale, was designed to investigate the students’ attitudes toward OCF inside the classroom, their preferences for different types of OCF and for the linguistic target of OCF. To meet the aims of the present study, some modifications were made to the questionnaire. The wording of the directions for answering the questions was slightly changed. Items tapping into the immediate/delayed and the implicit/explicit aspects of OCF were added (items 4 to 9). In part B of the original questionnaire, the learners were asked to indicate how often they liked their errors in different linguistic targets such as grammar, pronunciation, vocabulary, inappropriate expressions and organization of discourse (the last two subcomponents were deleted in the modified questionnaire) to be corrected from 1 (never) to 5 (always). The numbers between the two ends (in the original questionnaire) were substituted by some frequency adverbs so that it would be easier for the students to answer and more meaningful for the researchers to analyse the data. To identify what OCF types the students rank as the most/least contributory to their L2 WTC (item 20) and to find out in what components of language OCFs are more contributory to their L2 WTC (item 21), two items were added to the questionnaire. The questionnaire was piloted with 20 students of the language institute where the study was conducted and it was also validated and reviewed by two university professors familiar with the literature. The reliability of the questionnaire was estimated using Cronbach’s alpha (0.82). To ensure understandability, the learners completed the Persian version of the questionnaire. Table 1 shows the questionnaire constructs and the corresponding items.

Table 1.

Constructs and the corresponding items of the questionnaire.

Constructs	Item numbers
Teachers’ corrective behavior	1, 2, 3
Explicit/implicit OCF	4, 6, 8
Immediate/delayed OCF	5, 7, 9
Linguistic target of OCF	10, 11, 12, 21
OCF types and their effect on WTC	13, 14, 15, 16, 17, 18, 19, 20

Notes. OCF = oral corrective feedback. WTC = willingness to communicate.

b Semi-structured interviews

Two students from each class (eight in total) were randomly selected to be interviewed face-to-face, with reference to the interviewees’ WTC scores after each OCF type and their responses to the questionnaire items. The interview questions (asked in Persian; Appendix B) were aimed at tapping into areas of controversy concerning the tie between OCF and L2 WTC in the literature. The goal was to explore what cognitive and affective processes the students went through while receiving OCF and how OCF influenced their L2 WTC. Each interview session took 10 to 15 minutes. The responses were coded and categorized to detect the prospective themes.

c WTC scale

Initially, learners’ WTC was assessed by a WTC scale (MacIntyre et al., 2001; Appendix C). This scale was chosen as it was specifically designed to assess foreign language (French in the original study) learners’ WTC. The scale consisted of 35 items, all of which concentrated on the students’ willingness to participate in communicative tasks inside and outside the class time. For the purposes of the present research, we considered the first 27 items which focused on WTC inside the classroom. The highest and the lowest possible scores could be 135 and 27, respectively. The students were asked to indicate their willingness on a scale from 1 to 5 (where 1 = almost never willing, 2 = sometimes willing, 3 = willing half of the time, 4 = usually willing, and 5 = almost always willing). The scale covered the four skill areas (speaking: 8 items; reading: 6 items; writing: 8 items and comprehension 5 items). The reliability estimates (alpha level) for these skill areas were .81, .83 .88, and .83, respectively. Zarrinabadi and Abdi (2011) used and validated this scale in the Iranian EFL context. Some modifications were made to the original version of this scale to suit it for the objectives of the present research. In items 8 of speaking, 3 of the reading section and 4 and 5 of the comprehension section, the word ‘French’ was replaced by ‘English’ throughout. Also ‘not’ in item 2 of the comprehension section was removed.

d Intervention

To compare and contrast the effect of OCF types (explicit correction, recasts, and prompts) on the learners’ L2 WTC, the teachers divided the 18-session semester into three phases, each phase consisting of six sessions. In the first phase (i.e. sessions one to six), the teachers employed explicit correction to react to their students’ erroneous utterances. Similarly, recasts and prompts were employed by the teachers to correct the learners’ oral errors during sessions seven to 12 and 13 to 18, respectively (Figure 1). The examples of OCF types adapted from Katayama (2007) and provided over the course are presented below.

Example of Grammatical (G) error:

Teacher: ‘Where did you go yesterday’?

Student: ‘I go to the park.’

Example of Pronunciation (P) error:

Teacher: ‘What kind of flowers do you like best’?

Student: ‘I like loses best.’

(Recast)

G: ‘I went to the park.’ or ‘Went.’

P: ‘I like roses best.’ or ‘Roses.’

(Explicit correction)

G: ‘Go is wrong. You should say went.’

P: ‘Loses is wrong. You should say roses.’

(Clarification request)

Teacher indicates that an error occurred by nonverbal behavior, such as gesture and facial expressions or by saying words such as ‘sorry?’

(Repetition)

G: ‘I go to the park.’

P: ‘I like loses best.’

(Elicitation)

G: ‘I . . .’

P: ‘I like . . .’

(Metalinguistic feedback)

G: ‘Go is the present tense. You need the past tense here.’

P: (Using a picture of a mouth) ‘When you pronounce r for roses, your tongue should not touch the roof of the mouth. It should . . .’

Figure 1.

The schematic representation of the intervention phase.

The first researcher of the study provided the other male teacher with requisite instruction as to how to give feedback on the errors made by the students. As the aim was to focus on teachers’ OCF, the researcher asked the other teacher to explain to students, at the outset of the course, that in case of error occurrence of any kind it is the teacher’s responsibility to give OCF and rectify the error rather than the classmates. Hence, the interference of peer correction and its prospective effects on learners’ ultimate L2 WTC could be minimized. The intervention period for each PL took six weeks, three sessions per week.

3 Procedure

First, one class from each PL was randomly selected. All the groups completed the WTC scale before the commencement of the intervention. Each group received OCF in the form of explicit correction, recasts, and prompts provided by the teachers during the semester. The learners took the WTC scale every six sessions (four times altogether, pre-intervention WTC plus three times post-OCF). The survey questionnaire was then administered to all the students. The administration of the survey questionnaire at the outset of the study was meant to ensure that the learners’ answers were not affected by OCF types they received during the intervention. The other teacher was also provided with the requisite instruction as to how he should provide OCF to the learners’ errors. In the last phase of the study, eight students from four PLs were randomly selected and interviewed. The interview sessions were recorded, transcribed, and subsequently typed into Microsoft Word. The interview sessions complemented the data obtained through the questionnaire in extracting the themes most and least contributory to the learners’ L2 WTC. Finally, the learners’ performance on the WTC scale, the survey questionnaire, and the interviews were compared and the learners’ differing perceptions of OCF types and their effect on L2 WTC were studied.

To answer the first research question, frequency distributions were calculated to extract the learners’ views about the teachers’ corrective behavior, the linguistic target of OCF, the implicitness or explicitness, delayed or immediate OCF as well as the connection between teachers’ OCF types and their learners’ L2 WTC. Besides, qualitative analysis of the interviews indicated the learners’ reflection on what they went through while receiving OCF and complemented the data collected through the survey questionnaire and the intervention phase. The obtained themes were interpreted in tandem with the data gathered via the intervention to investigate the situations wherein the learners considered an OCF type more or less contributory to their L2 WTC. As for the second research question, pre-intervention and post-OCF WTC measures of the learners were compared in each PL. The overall higher score of the learners in a class on post-OCF WTC scale was indicative of the more contributory role of OCF type employed in that class. A two-way mixed between-within ANOVA was run to see how learners in each PL performed on WTC scales and which type(s) of OCF led to the highest WTC.

IV Results

1 Learners’ views about OCF and L2 WTC

The students’ answers to the questionnaire items were analysed to identify the learners’ attitudes toward OCF and how OCF affected their L2 WTC. Of the initial cohort of 60 questionnaires, five incomplete questionnaires were removed, and we analysed the rest. Drawing on Boone and Boone’s (2012) distinction between Likert type items and Likert scale (or construct in our study) data, we first investigated the single items and then moved on to see how the predetermined constructs were reflected in them (Table 2). The first construct was concerned with the corrective behavior of the teacher in the classroom. The learners unanimously believed that they liked their teacher to correct their errors. Concerning the frequency of the teachers’ OCF, 98.2% stated that the teachers should correct all the errors. However, a disagreement arose as to whether the teacher should only correct the errors which thwart communication (45.5% agreement vs. 36.4% disagreement).

Table 2.

Results of the survey questionnaire: frequency (percentage in parentheses).

Items	1^a	2	3	4	5
I. Teacher’s corrective behavior:
1. I want the teacher to correct my errors in the classroom.	0	0	1 (1.8)	20 (36.4)	35 (61.8)
2. Teachers should correct all the errors learners make in the classroom.	0	2 (3.6)	5 (9.1)	25 (45.5)	23 (41.8)
3. Teachers should correct only the errors that interfere with communication.	4 (7.3)	16 (29.1)	10 (18.2)	14 (25.5)	11 (20)
II. Immediate and delayed OCF:
4. I like the teacher to correct my grammatical errors immediately.	0	6 (10.9)	3 (5.5)	26 (47.3)	20 (36.4)
6. I like the teacher to correct my vocabulary errors immediately.	0	1 (1.8)	6 (10.9)	30 (54.5)	18 (32.7)
8. I like the teacher to correct my pronunciation errors immediately.	0	1 (1.8)	2 (3.6)	28 (50.9)	24 (43.6)
III. Explicit and implicit OCF:
5. I like the teacher to correct my grammatical errors explicitly.	0	2 (3.6)	8 (14.5)	29 (52.7)	16 (29.1)
7. I like the teacher to correct my vocabulary errors explicitly.	0	0	7 (12.7)	28 (50.9)	20 (36.4)
9. I like the teacher to correct my pronunciation errors explicitly.	0	0	6 (10.9)	26 (47.3)	23 (41.8)
IV. Linguistic target of OCF:
10. How often do you like your errors in grammar to be corrected?^b	4 (7.3)	5 (9.1)	5 (9.1)	14 (25.5)	27 (49.1)
11. How often do you like your errors in pronunciation, accent and intonation to be corrected?^b	3 (5.5)	5 (9.1)	8 (14.5)	11 (20)	28 (50.9)
12. How often do you like your errors in vocabulary usage to be corrected?^b	6 (10.9)	9 (16.4)	6 (10.9)	12 (21.8)	22 (40)
V. OCF types and their effect on WTC:
13. Teacher ignores the student’s error.	28 (50.9)	19 (34.5)	7 (12.7)	0	1 (1.8)
14. Teacher presents the correct response or part of the response. (Recast)	11 (20)	19 (34.5)	13 (23.6)	11 (20)	1 (1.8)
15. Teacher points out the error, and provides the correct response. (Explicit correction)	0	4 (7.3)	15 (27.3)	24 (43.6)	12 (21.8)
16. Teacher indicates that an error occurred by nonverbal behavior, such as gesture and facial expressions or by saying words such as ‘sorry?’ (Clarification request)	8 (14.5)	10 (18.2)	11 (20)	20 (36.4)	6 (10.9)
17. Teacher repeats the erroneous utterance. (Repetition)	1 (1.8)	5 (9.1)	19 (34.5)	18 (32.7)	12 (21.8)
18. Teacher repeats the student’s utterance up to the error, and waits for self-correction. (Elicitation)	2 (3.6)	4 (7.3)	8 (14.5)	25 (45.5)	16 (29.1)
19. T explains why the response is incorrect. (Metalinguistic feedback)	1 (1.8)	1 (1.8)	14 (25.5)	21 (38.2)	18 (32.7)

Notes. ^a1 = Strongly disagree, 2 = Disagree, 3 = Neutral, 4 = Agree, 5 = Strongly agree. ^b 1 = Never, 2 = Seldom, 3 = Sometimes, 4 = Usually, 5 = Always. OCF = oral corrective feedback. WTC = willingness to communicate.

As for the second construct, that is, the explicit/implicit dichotomy of OCF, the learners embraced (agreed or strongly agreed) explicit OCF in dealing with grammatical (81.8%), lexical (87.3%) and pronunciation errors (89.1%). Prior to administering the questionnaire, the first author expounded on items 4 through 9 (explicit/implicit and immediate/delayed OCF constructs) to the learners. He explained that these items focus on dichotomies and as such strong agreement with immediate OCF in grammar, for example, equals strong disagreement with delayed OCF. It also held true for explicit/implicit OCF construct with the difference that we clarified which OCF types are considered explicit and implicit. We categorized OCFs into implicit or explicit according to Sheen and Ellis’s (2011) taxonomy of OCF strategies. Thus, of the 6 OCF types used in our study, explicit correction, metalinguistic feedback, and elicitation were explicit, and recasts, clarification request, and repetition were implicit. With regard to the third construct, the immediate/delayed OCF, there was a clear preference (agreed or strongly agreed) for immediate correction among the learners in correcting grammatical (83.7%), lexical (87.2%) and pronunciation errors (94.5%).

As is evident, receiving explicit and immediate OCF tended to be preferred among learners, especially in rectifying pronunciation errors. With regard to the fourth construct, the linguistic target of OCF and the way OCF in those linguistic targets contributed to learners’ WTC, OCF on pronunciation was the most preferred (50.9%), closely followed by grammar (49.1%) and lexicon (40%). The respondents ranked OCF on pronunciation (Table 3) as the most contributory to their L2 WTC (47.2%), followed by grammar (38.2%) and lexicon (25.5%).

Table 3.

Contribution of linguistic targets to L2 WTC: frequency (percentage in parentheses).

Item	Most contributory	2nd	Least contributory
Respectively, in which part is oral corrective feedback more contributory to your L2 WTC?
Pronunciation	26 (47.2)	19 (34.6)	10 (18.2)
Grammar	21 (38.2)	16 (29.1)	18 (32.7)
Lexicon	14 (25.5)	19 (34.5)	22 (40)

Notes. L2 = second language. WTC = willingness to communicate.

Interesting results were obtained in the main part of the questionnaire where the students were asked to show their general dis/agreement with specific OCF types. The results revealed that 85.4 percent of the learners strongly disagreed or disagreed with ignoring the errors. Around 55 percent of them also strongly disagreed or disagreed with the use of recasts. Further, other OCF types respectively favored (agreed or strongly agreed) were elicitation (74.6%), metalinguistic feedback (70.9%), explicit correction (65.4%), repetition (54.5%), and clarification request (47.3%). The learners also ranked OCF types according to their relative contributory role to L2 WTC (Table 4). Elicitation (43.6%) was ranked the most contributory to L2 WTC, followed by metalinguistic feedback (29.1%), repetition (14.5%), explicit correction (12.7%), clarification request (5.4%), and recast (1.8%).

Table 4.

Relative contributory role of oral corrective feedback (OCF) types to second language (L2) willingness to communicate (WTC): frequency (percentage in parentheses).

Item	Most contributory	2^nd	3rd	4th	5th	Least contributory
20. Rank the OCF types from 1 (most contributory) to 6 (least contributory to L2 WTC).
Elicitation	24 (43.6)	12 (21.8)	13 (23.6)	3 (5.5)	3 (5.5)	0
Metalinguistic feedback	16 (29.1)	15 (27.3)	8 (14.5)	6 (10.9)	3 (5.5)	7 (12.7)
Repetition	8 (14.5)	19 (34.5)	10 (18.2)	13 (23.6)	3 (5.5)	2 (3.6)
Explicit correction	7 (12.7)	6 (10.9)	9 (16.4)	5 (9.1)	18 (32.7)	10 (18.2)
Clarification request	3 (5.4)	6 (10.9)	9 (16.4)	7 (12.7)	18 (32.7)	12 (21.8)
Recast	1 (1.8)	1 (1.8)	6 (10.9)	13 (23.7)	11 (20)	23 (41.8)

From the questionnaire results, it can be seen that most students preferred prompts such as elicitation and metalinguistic feedback to explicit correction and recasts. Elicitation was also considered to contribute the most to students’ WTC. In the interviews, some reasons for this preference for prompts were identified. Having been corrected via prompts, the learners stated that prompts made them resort to their current linguistic knowledge to rectify the errors and as such prompts engaged them in finding the right answer by resorting to their internal resources. Thus, prompts fostered motivation when they led to self-correction, which in turn boosted self-confidence. For instance, Mary, aged 14 years, said ‘When I am notified of the error and fix it myself, it feels good. I get motivated to do more.’ From this perspective, feedback was rather viewed as practice than correction.

With regard to explicit correction, it was seen as less conducive to WTC than other OCF types, even though over 65% of the students claimed to like it. One of the low PL level students, Elena, aged 11 years, for example, pointed out, ‘after that (being explicitly corrected), some classmates laughed at me and made fun of me outside. I didn’t like it.’ For the younger learners at the initial stages of language learning, explicit correction was even considered a threat, especially when the relationship between the teacher and the learner as well as among the learners was far from being intimate. However, explicit correction was believed to contribute to the higher-PL learners’ L2 WTC. Interestingly, commensurate with the rise in PL and age, sometimes, learners not only came to terms with explicit correction but also they tended to like the way it aided them to notice their errors, repair them, and consequently develop L2 WTC. Negar, aged 19 years, said, ‘I exactly know what was wrong. So, I would probably not repeat it again.’ In higher PLs, the learners advocating explicit correction maintained that explicit correction paved the way for discovering similar errors in similar contexts and ways to obliterate them from their interlanguage. This view was well expressed in Minoo’s (aged 21) comment, ‘well I like it explicit. I think that is a kind of teaching not for me but for everyone in the class.’ Such statements highlighted when and how explicit correction carried pedagogical implications in the language classroom. In other words, these comments suggest that explicit correction might go beyond its corrective intent and, especially when it was followed by a brief explanation of the error and the corrected form, could function as a pedagogical tool for not only the producer of the erroneous utterance but also for the whole class.

As for the relationship between PL and the differing perception of OCF’s effect, three students believed that their PL had nothing to do with their perception of OCF types and the impact of OCF on their L2 WTC. Seemingly, improvement in PL does not make a specific OCF type more or less significant or more or less contributory to learners’ L2 WTC. On the other hand, two interviewees maintained that PL affected their mindset toward OCF and the relation between OCF and L2 WTC. In this regard, Minoo (aged 21) said ‘I vividly recall that I did not like to be corrected much when I was an elementary learner, but it is quite reverse now and I deem correction necessary.’ However, three interviewees believed that the situational factors wherein OCF occurred (e.g. class size, the relationship between the teacher and the student, and individual difference factors such as shyness) mediated their perception of PL’s effect.

2 Effect of OCF types

Table 5 shows that the learners’ L2 WTC differed as a result of receiving different types of OCF. WTC decreased in the learners receiving explicit correction at lower levels (i.e. starter and elementary), but increased in the learners at higher levels (i.e. pre-intermediate and intermediate). The provision of recasts, on the other hand, led to a rise in WTC associated with the learners at all levels, while prompts promoted the biggest change in WTC for all the learners.

Table 5.

Descriptive statistics of willingness to communicate (WTC) scales.

WTC scales	Proficiency level (PL)	Mean	Standard deviation	n
Pre-intervention	Starter	84.5333	7.92705	15
	Elementary	89.4667	9.10939	15
	Pre-inter	90.1333	9.50839	15
	Intermediate	94.1333	8.46605	15
	Total	89.5667	10.84149	60
Post-explicit correction	Starter	78.1333	8.09644	15
	Elementary	85.0000	10.62342	15
	Pre-inter	92.1333	7.77685	15
	Intermediate	103.6667	9.13257	15
	Total	89.7333	8.08044	60
Post-recast	Starter	89.5333	8.99100	15
	Elementary	92.3333	9.03696	15
	Pre-inter	92.3333	7.66874	15
	Intermediate	101.1333	9.16669	15
	Total	93.8333	9.90950	60
Post-prompt	Starter	102.0000	8.46843	15
	Elementary	107.1333	10.35006	15
	Pre-inter	104.6000	8.27043	15
	Intermediate	112.0667	7.95291	15
	Total	106.4500	11.53686	60

A two-way mixed between-within ANOVA was conducted on the influence of the two independent variables (OCF and PL) on L2 WTC. OCF included three types (explicit correction, recasts, and prompts) and PL consisted of four levels (starter, elementary, pre-intermediate, and intermediate). All the differences were statistically significant at the .05 level of significance. Mauchly’s test indicated that the assumption of sphericity was violated (χ2(2) = 114.24, p < .001); therefore, the degrees of freedom were corrected using Greenhouse–Geisser estimates of sphericity (ε = 0.46). The main effect of OCF yielded an F ratio of (F(1.39, 78.21) = 83.77, p < .001, χ² = 0.599), indicating a significant difference between explicit correction (M = 89.73, SD = 8.08), recasts (M = 93.83, SD = 9.90), and prompts (M = 106.45, SD = 11.53). The effect size using Partial Eta Squared was 0.59, which was a large effect. The observed power was 1, which was very strong. Also, there was a significant main effect of PL (F(3, 56) = 4.014, p = .012, η² = 0.177) on L2 WTC of the learners. The effect size using Partial Eta Squared was 0.17, which indicated a large effect with a strong observed power of .81. The interaction effect between OCF types and PL was also significant (F(4.190, 78.21) = 4.382, p = .003, η² = 0.190). The effect size using Partial Eta Squared was large (i.e. 0.19) and the observed power was strong (i.e. .92).

Moreover, the pairwise comparisons showed that the performance of the learners from pre-intervention WTC (M = 89.56, SD = 10.84) to post- recast (M = 93.83, SD = 9.90) and from pre-intervention to post-prompt (M = 106.45, SD = 11.53) was significant. Significant differences were also observed from post-explicit correction (M = 89.73, SD = 8.08) and post- recast (M = 93.83, SD = 9.90) to post- prompt (M = 106.45, SD = 11.53). By and large, PL had a significant effect on the L2 WTC of the learners. However, a Bonferroni post hoc test revealed that only the difference between the least (starter) and the most (intermediate) proficient learners was significant which indicates that, generally, higher-PL can lead to higher L2 WTC. Nevertheless, comparing the PLs reveals that considerable differences in the PL are needed to cause variation in one’s L2 WTC rather than minor ones as observed, for instance, between starter and elementary levels. Therefore, both OCF types and PL affected the L2 learners’ WTC inside the classroom, with certain degrees of variability within each group and level (Figure 2).

Figure 2.

The interaction of oral corrective feedback (OCF) and proficiency level (PL).

V Discussion

This study set out to explore the effect of OCF types on L2 WTC. It also sought to investigate EFL learners’ views about OCF and the relationship between OCF and L2 WTC. With respect to the first research question, which considered the learners’ views about OCF, echoing previous research (Chenoweth et al., 1983; Katayama, 2007; Schulz 1996, 2001), this study found that the overwhelming majority of the learners liked their errors to be corrected. Furthermore, consistent with previous studies (Lyster, 2004; Sheen, 2007; Yilmaz, 2012), explicit OCF is strongly preferred, especially in pronunciation. In contrast to the data obtained through WTC scales, where explicit correction decreased the L2 WTC of the lower-PL learners, explicit correction was found to be highly preferred in the survey questionnaire data, even though prompts (in particular elicitation and metalinguistic feedback) were rated even higher. Likewise, the interview data suggested that the learners drew a distinction between the facilitative role of OCF types in L2 acquisition and how they contribute to one’s L2 WTC. In other words, in spite of the fact that explicit correction decreased L2 WTC of lower-PL learners, it was still popular as a way to explicitly point out where the learners have problems and to rectify their errors. Therefore, the learners apparently would rather sacrifice the transient WTC decrease to the longer lasting pedagogical merits of explicit correction. An issue worth noting is the learners’ special fondness to be corrected in pronunciation and grammar. A probable reason can be that many learners intend to develop their communicative skills by attending the conversation courses, and since they value pronunciation as a prerequisite to successful communication, receiving OCF in pronunciation is highly preferred. Their preference to be grammatically corrected, on the other hand, seems to have emanated from the decades-long tradition of teaching grammar-based courses in the Iranian EFL context. In line with the WTC data, the questionnaire results indicated that elicitation and metalinguistic feedback were the two most valued OCF types. Likewise, they were ranked as the two OCF types which were thought to contribute the most to L2 WTC. The reason why the learners in the present study preferred prompts such as elicitation may be because they welcome the opportunity to think about an error before receiving the correct form, like the Japanese as a foreign language students in Yoshida’s (2008) study. The preference for metalinguistic feedback on the other hand, is in line with findings from Lasagabaster and Sierra (2005), since the students in their study preferred longer explanations of errors, and it also confirms the preference for this type of CF expressed by the Spanish EFL students in Roothooft and Breeze’s (2016) study.

As for the second research question, that is, the effect of OCF and PL and their interaction effect on L2 WTC, the results of the present research conform with previous studies which reported higher levels of L2 WTC for the higher-PL learners (Alemi et al., 2013; Khajavy et al., 2014; Liu & Jackson, 2009). Our findings are partly consistent with Tavakoli and Zarrinabadi (2016), who maintained that explicit correction increased L2 WTC. It should be clarified, however, that this increasing effect is observed only among the higher-PL learners. Besides, our findings are at odds with Tavakoli and Zarrinabadi’s (2016) claim that implicit OCF did not affect learners’ L2 WTC. Though the provision of explicit correction differentially affected the learners with differing PLs, recasts and prompts increased learners’ L2 WTC across all the PLs, albeit with some variation.

A potential explanation for the differential effect of explicit correction on L2 WTC of the learners with varied PLs is that for the lower-PL learners, explicit correction was interpreted as an obvious sign of not mastering the point in question and as such it greatly hindered communication in the classroom. In other words, due to the tender age and an underdeveloped language ego, explicit correction may have triggered embarrassment for the younger less proficient learners. As Kang (2005) rightly pointed out, explicit correction brings about insecurity and a sense of losing face for the lower-PL learners. They might suffer even more severely in cases in which the relationship between the peers and with the teacher is not intimate enough. Kang (2005) urged teachers to create a ‘safe environment’ (p.290) especially for the lower-PL learners who experienced high levels of insecurity. On the contrary, at higher-PLs, explicit correction was viewed as a strategy to assist the learners in noticing an error, and through comparing the error and the corrected form, to prevent recurring errors in similar contexts. Therefore, explicit correction could act more as a pedagogical strategy for the higher-PL learners than just an indication of flaws in speech.

Contrary to the differing effects of explicit correction, recasts increased the learners’ L2 WTC across all the PLs. Though, overall, OCF types significantly affected L2 WTC, post-recast mean WTC scores were not markedly different from those of pre-intervention WTC. The relatively high gain score of the starter level learners is probably due to the implicit nature of recasts. It is likely that the starter level learners would hardly notice that their utterances were erroneous and hence recasts were seen as an implicit and supportive suggestion (Lyster& Ranta, 1997). As Lyster (1998) aptly noted, the lower-PL learners may have confused recasts with feedback on content. On the other hand, the intermediate level learners, due to the higher linguistic competence, could possibly have recognized the differences in the erroneous utterances and the corrected form and managed to rectify the error. As Philp (2003) and Trofimovich, Ammar and Gatbonton (2007) noted, higher-PL learners notice the corrective intent of recasts more easily and, thus, are able to correct their errors. This is a probable reason why their post-recast WTC scores rose by several points. Despite the frequent use of recasts in most contexts, in this study they were not really favored by students. More than half of the participants disagreed with the use of recasts (55%). The same trend was also observed where the students ranked recasts as the OCF type which contributed the least to L2 WTC. Though the lower-PL learners’ disagreement with recasts can be related to its imperceptibility, the higher-PL learners who might have been better able to detect the corrective nature of recasts, considered prompts as a more engaging alternative to recasts.

As with recasts, prompts enhanced L2 WTC in all the language learners. The learners’ post-prompt gain scores were not only higher than the post-recast scores but also the highest among all four WTC scales. It appears that for higher-PL learners, prompts contribute to L2 WTC to the extent that they engage the learners in self-correction. Apparently, it is through the challenges that prompts present to the learners to elicit the correct form from their interlanguage that the higher-PL learners find the highest level of inclination to communicate in the classroom. In this respect, our findings corroborate previous studies which reported a preference for elicitative types of OCF among higher-PL learners (Brandl, 1995; Kaivanpanah et al., 2015). The learners stated the teachers demonstrated through prompts that the utterance was problematic and, to some extent, indicated where the problem lay. This cue, when followed by the proper wait time (Zarrinabadi, 2014) provides learners with an ideal opportunity to correct their errors. The effort to rectify the error via one’s own resources can add to the learners’ motivation and self-esteem. The succession of self-correction, motivation (and less anxiety), and self-esteem may lead to seeking more opportunities to communicate with the teacher and the peers.

A noteworthy point was that while the learners ruled out the role of PL in how they perceived the effect of OCF on their L2 WTC, PL proved to mediate the effectiveness of OCF types on one’s L2 WTC in practice. Put differently, despite the fact that some learners dismissed PL as unimportant in their perception of OCF’s effect on L2 WTC, their actual WTC scores underwent significant changes from one PL to another as was documented in post-OCF WTC measures. That is, although some learners perceived PL to be irrelevant to their WTC, PL (in combination with OCF type) did trigger fluctuation in learners’ L2 WTC.

VI Conclusions

This study revealed that both OCF and PL affected L2 WTC. It also highlighted the views of the learners about OCF and the connection between OCF and L2 WTC. Our findings point to the importance of tailoring OCF to the learners’ PL inside the language classroom. The findings of this study broaden our understanding of how teachers can encourage learners to communicate and of the social situation in the pyramid model of L2 WTC. This study depicted that L2 WTC inside the language classroom, among other factors, hinges upon the type of OCF offered by the teachers and PL of the learners. The study also showed that learners are generally disposed to OCF. Explicit (including explicit correction, elicitation, and metalinguistic feedback) and immediate OCF types, especially in pronunciation, were highly preferred. OCF in pronunciation and grammar was perceived to have the greatest contribution to L2 WTC. In addition, elicitation and metalinguistic types of feedback were the most popular OCF types, both generally and according to their contribution to L2 WTC. This study has pedagogical implications for EFL teachers, especially those teaching the Iranian EFL context where the newly developed communicative syllabus seeks to offer more opportunities for the learners to communicate independently. Though it is important for teachers to know their students’ views about different OCF types, teachers need not be overly concerned about the perceived negative influences of OCF as the learners mostly welcome it. They are expected, however, to tailor their OCF to the learners’ PL. Based on our results, we would advise teachers to use recasts and prompts with lower-PL learners and a combination of explicit correction, recasts, and prompts with higher-PL learners to promote the learners’ L2 WTC. This study attempted to provide a relatively comprehensive picture of how OCF affects L2 WTC, which may lead to a better understanding of the factors that play a role in WTC.

A number of caveats need to be noted regarding the present study. First, as with most studies with repeated measures, administering the same scale for four times might have affected the learner’s responses. Second, incorporating male participants into the study could have added to the thoroughness of the research. Third, despite our efforts to control for extraneous variables, factors other than OCF and PL might have had an impact on the learners’ L2 WTC. Future research may focus on how OCF types affect learners with different personality types as well as different entry L2 WTC levels. Furthermore, comparing OCF types as negative evidence with the teachers’ positive reactions (in the form of expressions such as very good or well done) given after the correction of an error might yield interesting results as to how they might affect L2 WTC. Further research can also explore the differential effects of different types of prompts or recasts on L2 WTC. Last but not least, it may be interesting to replicate our study in an online context.

Supplemental Material

Appendix – Supplemental material for The interplay of oral corrective feedback and L2 willingness to communicate across proficiency levels

Supplemental material, Appendix for The interplay of oral corrective feedback and L2 willingness to communicate across proficiency levels by Mostafa Zare, Zohreh Gooniband Shooshtari and Alireza Jalilifar in Language Teaching Research

Footnotes

Acknowledgements

We wish to thank Mr. Azim Sabiza and Miss Ghazal Norouzi for their generous assistance in collecting data and the learners of Shokuhe-Andishe Language Institute for participation in this study. We are also grateful to Dr. Goudarz Alibakhshi for his help with statistical analysis and the two anonymous Language Teaching Research reviewers for their invaluable comments and suggestions.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Mostafa Zare

Zohreh Gooniband Shooshtari

Supplemental material

Supplemental material for this article is available online.

References

Alemi

Tajeddin

Mesbah

(2013). Willingness to communicate in L2 English: Impact of learner variables. Journal of Research in Applied Linguistics, 4, 42–61.

Allan

(2004). Oxford placement test 2. Oxford: Oxford University Press.

Ammar

(2008). Prompts and recasts: Differential effects on second language morphosyntax. Language Teaching Research, 12, 183–210.

Boone

H.N.

Boone

D.A.

(2012). Analyzing likert data. The Journal of Extension, 50, 1–5.

Brandl

(1995). Strong and weak students’ preferences for error, feedback options and responses. The Modern Language Journal, 79, 194–211.

Brown

A.V.

(2009). Students’ and teachers’ perceptions of effective foreign language teaching: A comparison of ideals. The Modern Language Journal, 93, 46–60.

Chenoweth

N.A.

Day

Chun

A.E.

Luppescu

(1983). Attitudes and preferences of ESL students to error correction. Studies in Second Language Acquisition, 6, 79–87.

Hopkins

(2003). Get ready. Oxford: Oxford University Press.

Horwitz

E.K.

(1988). The beliefs about language learning of beginning university foreign language students. The Modern Language Journal, 72, 283–294.

10.

Iranian Ministry of Education. (2011). National curriculum. Tehran: Ministry of education.

11.

Kaivanpanah

Alavi

S.M.

Sepehrinia

(2015). Preferences for interactional feedback: Differences between learners and teachers. The Language Learning Journal, 43, 74–93.

12.

Kang

(2005). Dynamic emergence of situational willingness to communicate in a second language. System, 33, 277–292.

13.

Kartchava

Ammar

(2014). Learners’ beliefs as mediators of what is noticed and learned in the language classroom. TESOL Quarterly, 48, 86–109.

14.

Katayama

(2007). Students’ perceptions toward corrective feedback to oral errors. Asian EFL Journal, 9, 289–305.

15.

Khajavy

G.H.

Ghonsooly

Fatemi

A.H.

(2014). Willingness to communicate in English: A microsystem model in the Iranian EFL classroom context. TESOL Quarterly, 50, 154–180.

16.

Krashen

(1982). Principles and practice in second language acquisition. Oxford: Pergamon.

17.

Lanuzzi

(2015). First friends. Oxford: Oxford University Press.

18.

Lasagabaster

Sierra

J.M.

(2005). Error correction: Students’ versus teachers’ perceptions. Language Awareness, 14, 112–127.

19.

(2010). The effectiveness of corrective feedback in SLA: A meta-analysis. Language Learning, 60, 309–365.

20.

Vuono

(2019). Twenty-five years of research on oral and written corrective feedback in System. System, 84, 93–109.

21.

Liu

Jackson

(2009). Reticence in Chinese EFL students at different proficiency levels. TESL Canada Journal, 26, 65–81.

22.

Lyster

(1998). Negotiation of form, recasts, and explicit correction in relation to error types and learner repair in immersion classrooms. Language Learning, 48, 183–218.

23.

Lyster

(2004). Differential effects of prompts and recasts in form-focused instruction. Studies in Second Language Acquisition, 26, 399–432.

24.

Lyster

Ranta

(1997). Corrective feedback and learner uptake: Negotiation of form in communicative classrooms. Studies in Second Language Acquisition, 19, 37–66.

25.

Lyster

Saito

(2010). Oral feedback in classroom SLA: A meta-analysis. Studies in Second Language Acquisition, 32, 265–302.

26.

Lyster

Saito

Sato

(2013). Oral corrective feedback in second language classrooms. Language Teaching, 46, 1–40.

27.

MacIntyre

P.D.

(1994). Variables underlying willingness to communicate: A causal analysis. Communication Research Reports, 11, 135–142.

28.

MacIntyre

P.D.

Charos

(1996). Personality, attitudes, and affect as predictors of second language communication. Journal of Language and Social Psychology, 15, 3–26.

29.

MacIntyre

P.D.

Burns

Jessome

(2011). Ambivalence about communicating in a second language: A qualitative study of French immersion students’ willingness to communicate. Modern Language Journal, 95, 81–96.

30.

MacIntyre

P.D.

Baker

S.C.

Clément

Conrod

(2001). Willingness to communicate, social support, and language-learning orientations of immersion students. Studies in Second Language Acquisition, 23, 369–388.

31.

MacIntyre

P.D.

Clément

Dornyei

Noels

K.A.

(1998). Conceptualizing willingness to communicate in a L2: A situational model of L2 confidence and affiliation. Modern Language Journal, 82, 545–562.

32.

Matsuoka

(2005). Willingness to communication among Japanese college students. Paper presented at the 10th Conference of the Pan-Pacific Association of Applied Linguistics. Available at: http://www.paaljapan.org/resources/proceedings/PAAL10/pdfs/matsuoka.pdf (accessed May 2020).

33.

McCarthy

McCarten

Sandiford

(2006). Touchstone. Cambridge: Cambridge University Press.

34.

McCroskey

J.C.

Baer

J.E.

(1985). Willingness to communicate: The construct and its measurement. Paper presented at the annual meeting of the Speech Communication Association, Denver, CO, USA. ERIC Document Reproduction Service No. ED265604.

35.

McCroskey

J.C.

Richmond

V.P.

(1990). Willingness to communicate: Differing cultural perspectives. Southern Communication Journal, 56, 72–77.

36.

Nassaji

(2009). Effects of recasts and elicitations in dyadic interaction and the role of feedback explicitness. Language Learning, 59, 411–452.

37.

Panova

Lyster

(2002). Patterns of corrective feedback and uptake in an adult ESL classroom. TESOL Quarterly, 36, 573–595.

38.

Philp

(2003). Constraints on ‘noticing the gap’: nonnative speakers’ noticing of recasts in NS–NNS interaction. Studies in Second Language Acquisition, 25, 99–126.

39.

Rassaei

(2015). Oral corrective feedback, foreign language anxiety and L2 Development. System, 49, 98–109.

40.

Rivers

Tayama

(2015). English time. Oxford: Oxford University Press.

41.

Roothooft

Breeze

(2016). A comparison of EFL teachers’ and students’ attitudes to oral corrective feedback. Language Awareness, 25, 318–335.

42.

Russell

Spada

(2006).The effectiveness of corrective feedback for the acquisition of L2 grammar. In Norris

Ortega

(Eds.), Synthesizing research on language learning and teaching (pp. 133–162). Amsterdam: John Benjamins.

43.

Saito

Lyster

(2012). Effects of form-focused instruction and corrective feedback on L2 pronunciation development of /ɹ/ by Japanese learners of English. Language Learning, 62, 595–633.

44.

Schulz

R.A.

(1996). Focus-on-form in the foreign language classroom: Students’ and teachers’ views on error correction and the role of grammar. Foreign Language Annals, 29, 343–364.

45.

Schulz

R.A.

(2001). Cultural differences in student and teacher perceptions concerning the role of grammar instruction and corrective feedback: USA–Colombia. The Modern Language Journal, 85, 244–258.

46.

Sheen

(2007). The effects of corrective feedback, language aptitude and learner attitudes on the acquisition of English articles. In Mackey

(Ed.), Conversational interaction in second language acquisition: A collection of empirical studies (pp. 301–322). Oxford: Oxford University Press.

47.

Sheen

(2008). Recasts, language anxiety, modified output, and L2 learning. Language Learning, 58, 835–874.

48.

Sheen

Ellis

(2011). Corrective feedback in language teaching. In Hinkel

(Ed.), Handbook of research in second language teaching and learning: Volume 2 (pp. 593–610). New York: Routledge.

49.

Tavakoli

Zarrinabadi

(2016). Differential effects of explicit and implicit corrective feedback on EFL learners’ willingness to communicate. Innovation in Language Learning and Teaching, 12, 247–259.

50.

Trofimovich

Ammar

Gatbonton

(2007). How effective are recasts? The role of attention, memory, and analytical ability. In Mackey

(Ed.), Conversational interaction in second language acquisition (pp. 144–171). Oxford: Oxford University Press.

51.

Truscott

(1999). What’s wrong with oral grammar correction. Canadian Modern Language Review, 55, 437–455.

52.

Yang

(2016). Learners’ oral corrective feedback preferences in relation to their cultural background, proficiency level and types of error. System, 61, 75–86.

53.

Yang

Lyster

(2010). Effects of form-focused practice and feedback on Chinese EFL learners’ acquisition of regular and irregular past tense forms. Studies in Second Language Acquisition, 32, 235–263.

54.

Yilmaz

(2012). The relative effects of explicit correction and recasts on two target structures via two communication modes. Language Learning, 62, 1134–1169.

55.

Yoshida

(2008). Teachers’ choice and learners’ preference of corrective feedback types. Language Awareness, 17, 78–93.

56.

Zarrinabadi

(2014). Communicating in a second language: Investigating the effect of teacher on learners’ willingness to communicate. System, 42, 288–295.

57.

Zarrinabadi

Abdi

(2011). Willingness to communicate and language learning orientations in Iranian EFL context. International Education Studies, 4, 206–214.

58.

Zhang

L.J

Rahimi

(2014). EFL learners’ anxiety level and their beliefs about corrective feedback in oral communication classes. System, 42, 429–439.

59.

Zhu

Wang

(2019). Investigating English language learners’ beliefs about oral corrective feedback at Chinese universities: a large-scale survey. Language Awareness, 28, 139–161.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.11 MB