Factors influencing EFL teachers’ provision of oral corrective feedback: the role of teaching experience

ABSTRACT

This study highlights critical factors influencing English as a Foreign Language (EFL) teachers’ decisions to provide in-class oral corrective feedback (OCF). It explores the interplay between teachers’ individual differences – namely educational background, teaching experience, and additional training – and their decisions regarding OCF. EFL teachers in three Turkish universities each had three hours of their speaking classes recorded. A thematic analysis of stimulated recall interview data revealed that teachers’ decisions were influenced by learner-related, contextual, and teacher-specific factors. The teachers stated that they provided OCF because they believed that learners expected it. Learner-related factors such as emotions and low proficiency served as deterrents to OCF. Regarding the contextual factors, institutional expectations motivated the provision of OCF, while the nature of activities dissuaded teachers from administering OCF. The teacher factor most commonly cited as a stimulus for providing OCF was the inclination to elicit the correct form from students. The teachers were reluctant to provide OCF because they favoured delayed feedback at the end of the activity or lesson. Regression analysis of the video data revealed that only teaching experience contributed significantly to teachers’ decision-making. The role of teaching experience is discussed, and suggestions are made regarding the provision of OCF.

KEYWORDS:

• Oral corrective feedback
• teachers’ individual differences
• teaching experience
• classroom-based research
• teachers’ decisions
• EFL

1. Introduction

For the purposes of our study, we understand oral corrective feedback (OCF) as ‘comments a teacher or an interlocutor makes on errors that occur in second language learners’ speech production’ (Li 2022: 353). Addressing learners’ errors in the target language poses a crucial quandary for language teachers, given its potential positive and negative impacts. On the one hand, existing studies and meta-analyses posit that OCF contributes to linguistic development (e.g. Li 2010; Mackey and Goo 2007). On the other hand, some teachers often refrain from giving excessive OCF due to apprehensions that it may disrupt the flow of communication and undermine learners’ confidence (Jean and Simard 2011) or impede learner autonomy (Lantolf 2000). Nevertheless, this assumption may not be universally held, as evidenced by recent findings from Ha and colleagues, which indicate that EFL teachers in the Vietnamese context showed favourable attitudes towards OCF (Ha, Nguyen, and Hung 2021; Ha 2023a).

Earlier studies on OCF focused on the type, modality and structure of the feedback provided (e.g. Mackey and Oliver 2002; Nassaji and Kartchava 2019). Other studies have investigated the characteristics of the learners who receive it; for example, their proficiency level (e.g. Mackey and Philp 1998), age (e.g. Mackey and Oliver 2002), beliefs (Ha 2023b; Ha, Nguyen, and Hung 2021), or motivation, attitudes and anxiety (e.g. DeKeyser 1993). Scant attention has been devoted to exploring the teachers, the providers of feedback, and their associated characteristics (e.g. Gurzynski-Weiss 2014; Long 2017; Ha 2023a; Ha and Murray 2023; Rahman and Singh 2023).

Even fewer studies have addressed the factors that affect teachers’ decisions about providing OCF (e.g. Atai and Shafiee 2017; Gurzynski-Weiss 2016; Ha 2023a&b). Compared to other dimensions of feedback, this area has a significant dearthof research. Although recent research on OCF highlights the need for the investigation of various factors that may affect the feedback provision of teachers, among other aspects, such as teachers’ and students’ attitudes and perspectives about OCF (Ha 2023a&b; Nassaji and Kartchava 2019), individual differences such as the ways in which additional training, teaching experience and educational background might impact the teachers’ provision of OCF remain under-researched (Ha and Murray 2021). Motivated by the lack of previous research on this topic, in addition to the calls by Lyster, Saito, and Sato (2013) and Brown (2016), this study examines the factors influencing the decision-making processes of teachers regarding OCF in English as a foreign language (EFL) classes in three Turkish universities. This study is also one of only a few studies investigating the effects of individual differences on the provision of OCF, including teaching experience, additional training, and educational background.

2. Literature review

Recent research on OCF predominantly adopts an interactionist perspective of second language acquisition, conceptualising feedback as a facilitative instrument for linguistic development (Nassaji and Kartchava 2021). Within this framework, OCF targeting non-target-like performances of learners is posited to enhance the likelihood of learners recognising their errors (Schmidt 2001), thereby fostering linguistic development (Li 2010; Mackey and Goo 2007).

Moreover, a variety of studies examined the teachers and their characteristics. In such research, major areas of focus have been:

• the beliefs of the teachers (Yuksel, Soruc, and McKinley 2021; Yuksel, Soruc, and McKinley 2023; Canals et al. 2020; Gurzynski-Weiss 2010; Ha, Nguyen, and Hung 2021; Hernandez Mendez and Reyez Cruz 2012; Mahalingappa, Polat, and Wang 2021),

• the consistency between the beliefs and practices (Yuksel, Soruc, and McKinley 2021; Yuksel, Soruc, and McKinley 2023; Basturkmen, Loewen, and Ellis 2004; Ha and Murray 2023; Junqueira and Kim 2013; Kamiya 2016; Olmezer-Ozturk, 2019; Rahman and Singh 2023), as well as in two opposing lesson types (Ha 2023a) or

• the individual differences that predict the relationship between teachers’ beliefs and/or practices (Yuksel, Soruc, and McKinley 2023; Gurzynski-Weiss 2010; Long 2017; Mackey, Polio, and McDonough 2004; Polio, Gass, and Chapin 2006).

These studies contribute to a heightened understanding of the significance of factors that shape teachers’ decision-making processes in the administration of OCF. However, further research is warranted to explore the correlation between individual differences among EFL teachers and their decision-making processes (Atai and Shafiee 2017; Gurzynski-Weiss 2016; Sluman 2014).

2.1. Factors that influence the decisions of teachers regarding OCF

A frequently referenced study on teachers’ decision-making in the context of providing OCF is by Gurzynski-Weiss (2016), wherein the influencing factors were categorised into three distinct domains: contextual, learner, and teacher factors. The study involved 32 university-level teachers instructing Spanish as a foreign language. Thirteen factors encouraging the use of OCF and twelve factors discouraging its implementation were identified. Predominantly, contextual factors were highlighted as the most influential, with linguistic target (mentioned by all teachers) and error interrupting meaning (mentioned by 28 out of 32 teachers) being the most cited encouraging factors. Conversely, the three most cited discouraging factors were errors that did not impede meaning (26 out of 32 – contextual), concern for a specific student's emotional state or anxiety level (21 out of 32 – learner), and students’ low level of preparedness or general ability (20 out of 32 – learner).

In a case study examining an experienced ESL teacher in England, Sluman (2014) delved into the rationale behind the teacher's decisions regarding the dispensation of OCF. The findings highlighted that the teacher's decisions were shaped by various factors, including considerations related to learners (e.g. the current developmental stage of the learner, linguistic needs of the learner), contextual factors (e.g. the mood of the class, type of error), and teacher-related factors (e.g. beliefs about OCF). Conversely, when the teacher opted not to provide OCF, similar factors came into play, including considerations related to learners (e.g. concern for a particular student's feelings), contextual factors (e.g. the flow of the lesson, the mood of the class), and teacher-related factors (e.g. beliefs about OCF, personal learning experiences). In another study, Atai and Shafiee (2017) explored the decision-making processes of three Iranian teachers regarding the dispensation of OCF through stimulated recall interviews. Employing Kumaravadivelu's (2012) framework, they categorised 19 pedagogical thought categories into three distinct types: professional knowledge, procedural knowledge, and personal knowledge. Although Atai and Shafiee's (2017) classification differs from that of Gurzynski-Weiss (2016), they similarly identified contextual factors (e.g. institutional constraints, exposure to content, classroom management), learner factors (e.g. affective considerations, individual differences), and teacher factors (e.g. teacher’s agency) as pivotal components in teachers’ decision-making processes regarding the provision of OCF.

In addition to the factors previously discussed, more recent research has uncovered several other elements influencing teachers’ decisions. Notably, Ha's recent studies on OCF (e.g. Ha and Murray 2021; Ha 2023a) have highlighted the influence of a professional development programme focused on OCF, as well as the instructional emphasis on the lesson (e.g. grammar versus speaking), in shaping teachers’ decisions concerning OCF. In Ha and Murray's (2021) study, Vietnamese high school EFL teachers exhibited shifts in their beliefs regarding OCF following a tailored workshop that incorporated experiential and reflective activities derived from recent research on second language acquisition. Moreover, another study by Ha (2023a) investigated the OCF beliefs and practices of Vietnamese high school EFL teachers in two contrasting lesson types – grammar versus speaking. This revealed that the nature of the lesson plays a pivotal role in enabling teachers to demonstrate greater consistency in providing CF in the grammar-focused lesson compared to the speaking-focused lesson. In the case of grammar instruction, educators exhibited a preference for explicit explanations or corrections, often coupled with metalinguistic elements. Conversely, in communication-oriented lessons, recasts emerged as the favoured approach for addressing pronunciation errors.

A synthesis of these studies indicates considerable divergence among teachers in their decisions on whether or not to offer OCF. Moreover, teachers in each specific setting may harbour distinct cognitive perspectives regarding the provision of OCF, despite the existence of some overarching patterns. Furthermore, teachers’ beliefs about OCF are considered an independent construct, distinct from other facets of language learning (Li 2017). Given the limited number of studies investigating the factors influencing teachers’ decision-making processes in providing OCF, additional research in this domain is warranted (Nassaji and Kartchava 2019; Ha and Murray 2021).

2.2. Effects of teachers’ individual differences on the provision of OCF

Li and Vuono (2019), in their comprehensive review spanning 25 years of CF research, contend that the beliefs of both teachers and learners have not been sufficiently investigated, particularly considering the diverse characteristics and individual differences among participants. Individual differences encompass various traits associated with teachers, including personality, identity, and other demographic variables. These differences are defined as ‘situational variables that account for differences attributable to circumstances such as experience and training’ (Agarwal and Prasad 1999: 362).

In a review of prior research on OCF, Nassaji (2016) contends that teaching experience could influence the effectiveness of feedback delivered by teachers. Among the limited studies in this domain, Polio, Gass, and Chapin (2006) investigated the impact of teaching experience on teachers’ beliefs regarding OCF. Their findings indicated a correlation between teachers’ experience and their cognitive orientations towards OCF. Experienced teachers were observed to elicit more student production, while their less experienced counterparts tended to rely more on themselves and the procedural aspects of the lesson, prioritising these over the developmental needs of the learners. Contrary to this finding, Gurzynski-Weiss (2016) found that teachers with the least amount of experience exhibited greater attentiveness to and reflection upon learner errors. In a recent investigation, Kartchava et al. (2020) posit that the deficiency in integrating theoretical knowledge and practical skills observed among pre-service teachers, due to their lack of experience, might lead them to function more as native-speaking interlocutors rather than as language teachers. The somewhat contradictory and mixed findings underscore the need for additional research to delve into the impact of teaching experience on the factors influencing teachers’ provision of OCF.

Gurzynski-Weiss (2016), in examining the impact of additional training or awareness-raising activities on the provision of OCF, identified training as the most frequently cited individual difference associated with contextual factors. Similarly, Mackey, Polio, and McDonough (2004) reported that training heightened the awareness of incidental focus-on-form among less experienced teachers.

In a series of studies, Ha and colleagues concluded that additional training, particularly in the form of a professional development programme, positively influenced the beliefs of experienced EFL teachers. This is evidenced in Ha and Murray's (2021) study. Subsequent studies (Ha 2022; Ha 2023b) revealed that the professional development programme, or a form of awareness-raising reflection activity, despite its brevity, had a notable and positive influence on the practices of EFL teachers. Contrastingly, a cluster of studies, including those conducted by Gurzynski-Weiss (2010), Junqueira and Kim (2013), and Long (2017), reported no discernible impact of additional training on teachers’ practices of providing OCF. The varied and inconclusive nature of these results underscores the need for further research in this domain.

In summary, as emphasised by Gurzynski-Weiss (2017), the scrutiny of teachers’ characteristics during the delivery of OCF is crucial due to the potential of individual differences to ‘influence the learning opportunities that all learners in a given classroom setting receive during meaningful interaction’ (154). The factors influencing teachers’ decisions and the intricate interactions with individual differences warrant additional research and closer examination. The overview of previous studies provided valuable insights for the design and planning of the current study as follows:

• Only a limited number of studies have explored the factors influencing teachers’ decision-making processes when delivering OCF.

• The conclusions drawn from the studies in the literature are derived from relatively small participant samples.

• Research on the predictive influence of individual differences on the factors shaping teachers’ decision-making processes is notably scarce, if not entirely absent.

While there are research studies that have separately examined either (a) the factors influencing EFL teachers’ provision of OCF or (b) the predictive power of teachers’ individual differences in their beliefs/practices, there is currently a dearth of research that investigates both factors and individual differences in a unified context, therefore, this study makes an original contribution to the existing body of research on OCF.

Research Questions

The current study aimed to address the following research questions:

RQ1: What factors influence EFL teachers’ decisions in the provision of OCF?

RQ2: What is the relationship, if any, between EFL teachers’ individual differences and the three factors (contextual, learner, and teacher) regarding their decision-making processes while providing OCF?

3. Methodology

3.1. Setting

The research was conducted at the Intensive English Schools of three universities in a major city in Turkey. The selection of these three schools was purposeful, not only because they utilised the same textbooks from an international publishing house – which was assumed to minimise potential differences among them – but also because this enabled the researchers to obtain more representative data. The selected textbooks were designed to instruct students in English for academic purposes, focusing on enhancing their academic language competence by teaching the four language skills separately. To ensure uniformity in observations regarding the ‘focus’ aspect of corrective feedback types, only speaking classes were video-recorded by placing a camcorder at the back of each class. This choice was made because speaking classes typically involve more teacher-student or student-student interactions.

3.2. Participants

The participants in the study comprised a convenience sample of 67 Turkish EFL teachers and their students, using a volunteer sampling strategy. Each class had between 12 and 22 students (Mean = 16.7). All the students were enrolled on pre-sessional intensive English classes dedicated to enhancing their English language skills before commencing their degree programmes. These classes, also referred to as English preparatory classes, are taught intensively for one year before students in Turkey begin their major studies at the university, typically before Year 1. The teachers were grouped according to three individual differences, as follows:

• The participants had varying lengths of teaching experience, ranging from 1 year to 27 years (Mean = 12.4). Based on the assumption that teachers’ prior experiences predominantly influence their cognitions and practices in the first seven years of their teaching career (Gurzynski-Weiss 2010), the threshold for teachers’ experience was set at seven years. Consequently, 41 teachers were categorised as less experienced, while the remaining 26 were classified as experienced. ¹

• Regarding the teachers’ educational background, thirty-nine held undergraduate degrees in English language teaching (ELT), while the remainder graduated from various fields such as English language and literature (n = 13), translation and interpretation (n = 8), American culture (n = 6), and linguistics (n = 1). Despite all participants teaching English at the university level, their diverse educational backgrounds may have influenced their beliefs and practices in OCF, especially considering the absence of a uniform pre-service teacher education programme for all participants before they commenced teaching.

• Participants were queried about their engagement in any additional training, awareness-raising activities, or seminars specifically addressing OCF after they commenced teaching. Those who reported undertaking a training course were categorised into the ‘+additional training’ group (n = 36), while those who reported no involvement in any training programme were placed in the ‘–additional training’ group' (n = 31).

3.3. Data collection

Data for this study were obtained in two stages. In the initial stage, detached classroom observations were conducted using video recordings, while the second stage involved collecting cognitive data through stimulated recall interviews. One of the researchers (author 2) orchestrated the data collection process and was present in all recording sessions.

3.3.1. Stage one: detached observation for teachers’ OCF practices

All the teachers (N = 67) were video-recorded three times throughout one academic semester by the first two authors, resulting in a cumulative total of 201 hours of detached observations. Each class session lasted between 52–57 minutes. These observations were conducted in the teachers’ regular classrooms, with the intention of unveiling both their exemplary and typical OCF practices. The extensive number of observations was primarily intended to document the teachers’ most prevalent practices and to mitigate the possibility of teachers engaging in any failsafe activities that they might perform, assuming they were part of a research project. Video recordings were exclusively conducted in speaking classes because it was anticipated that these lessons would involve more interactive activities. However, it is essential to note that these video recordings primarily served as stimuli during the recall sessions for the teachers, as outlined in more detail below.

3.3.2. Stage two: stimulated recall interviews

The primary method of data collection was through stimulated recall interviews. Following the approach outlined by Mackey and Gass (2022), these sessions involve recalling a past activity using a stimulus, such as an audio or video recording. The interview sessions specifically concentrated on the teachers’ practices of OCF and the rationale behind their choices. To aid teachers in recalling the most accurate information and enhance the reliability of their reported thoughts (following Mackey and Gass 2022), the interviews were conducted promptly after the classes. In most cases, the interviews took place on the same day as the videotaped teaching. The interviews occurred in an office situated at the location where the teacher taught and involved both the researchers (the first two authors of this manuscript) and the teacher. Prior to the interviews, the teachers were assured that there were no right or wrong answers to the questions and that their responses would be treated confidentially. When queried about their language preferences during the interviews (i.e. Turkish and English), all teachers expressed a preference for speaking in English. They believed that using English would allow them to articulate key terms or vocabulary more comfortably.

During the stimulated recall interviews, the researcher played the segment depicting the initial learner error until the conclusion of the episode. The video was then paused after the student's response, if one was provided. The stimulated recall interviews incorporated the first twenty errors recorded during the speaking lessons. Each interaction episode comprised (a) a student error, (b) a teacher's response, and (c) the student's subsequent response (following Lyster and Ranta 1997). All errors, whether corrected or not by the teachers, were included in the selected first twenty errors. To prevent distraction, no other instances of classroom discourse beyond these initial twenty errors were presented to the teachers during the recall sessions. The teacher was then asked to think and speak aloud, considering the following issues:

1. what s/he thought at that time,

2. why s/he decided to provide OCF (or not) for that error,

3. if s/he corrected the error, what motivated him/her to choose that particular type of OCF?

All stimulated recall interviews were audio recorded for further analysis, as detailed below.

3.4 Data analysis

Following the transcription of the audio recordings from the stimulated recall interviews, the data analysis commenced, focusing on the three main factors. A thematic analysis of OCF episodes associated with the three factors was conducted. Initially, the first two authors of this article scrutinised instances involving errors, and subsequently, two EFL teachers also engaged in the same analysis. To enhance the reliability of the data analysis, a training manual containing samples of the factors was developed. Additionally, to validate the coding process, two external reviewers were enlisted to independently recode 30% of the transcriptions from the stimulated recall interviews. Regarding inter-rater reliability, Cohen's kappa analysis results indicated good agreement between the researcher and the first external reviewer (k = .887) and between the researcher and the second external reviewer (k = .891) concerning the identification of the factors.

4. Findings

4.1. Factors that influence the decisions of the teachers in the provision of OCF

The results pertaining to Research Question 1, which investigates the decisions influencing teachers’ OCF during speaking lessons, reveal that the 67 teachers in this study were driven by learner, contextual, and teacher factors in their moment-to-moment decisions. Notably, in a single corrective feedback move, they occasionally considered more than one of these factors (Table 1). Additionally, teachers’ personal beliefs, attitudes, and thoughts about learning, teaching, and corrective feedback played a significant role in guiding their decisions.

Table 1. Factors that influenced the provision of OCF.

Factors	Encouraging corrective feedback	Total number of mentions (number of mentions more than once by the same teacher)	Discouraging corrective feedback	Total number of mentions (number of mentions more than once by the same teacher)
Learner	Learners expectations about OCF	47 (23)	Feelings & anxiety	41 (34)
	Learners benefit from OCF	37 (14)	Low level of proficiency	36 (23)
	Learners are not bothered with OCF	34 (17)	Learners’ personality/ vulnerability	32 (18)
	Understanding learner’s intention	29 (8)	Not ready for the target item	23 (17)
	Confused look	26 (11)
	Previous language learning experiences	21 (9)
Contextual	Institutional expectations	58 (46)	Seriousness of error	29 (14)
	Common mistake type	47 (34)	Linguistic focus	20 (11)
	Linguistic focus	44 (31)	Type of the activity	34 (23)
	Seriousness of error	34 (23)	Timing of error	27 (15)
	Fossilisation	32 (18)
	L1 Use	31 (23)	No time for correction	16 (11)
	Pause in speech	28 (19)
Teacher	Desire to hear the correct form from the students	56 (43)	Choose to correct later	41 (36)
	Previous teaching experience	41 (32)	Classroom management issues	38 (24)
	Previous learning experience	34 (26)	Limited knowledge of the teacher	23 (4)
			Fear of too much correction	11 (6)
			Beliefs about OCF	18 (14)

In this section, the emerging patterns from the stimulated recall interviews of the teachers are presented, accompanied by sample comments from the interviews. As outlined in the methodology section, the analysis of the stimulated recall interviews focused solely on the errors that were observed by the teachers during the lessons at the time of the interaction, with a 56% ratio of errors identified by the teachers. Several errors not relevant to the main focus of the study were excluded from the analysis and are not reported here. Teacher decisions were categorised into three broad groups: learner, contextual, and teacher factors. These categories are detailed in Table 1, presented in two sections – factors that motivated teachers to provide OCF and factors that deterred them from doing so. A separate category was included for instances where the same factor was repeated by the same teacher (refer to Table 1 for details).

Learner Factors. As indicated in Table 1, the most frequently mentioned learner factor that prompted the provision of OCF was the perceived expectation of the learner, cited by 47 teachers (70%). During the stimulated recall interviews, 23 teachers elaborated on this perceived expectation of the learner regarding OCF. An example of this sub-category is presented in Excerpt 4 below.

Table

(4)

Teacher 03: Our students are open to learning, but I believe they don't have enough opportunities. Maybe the classroom is the only setting they can improve their English. In terms of feedback, I think they need it because they don’t have many chances outside the classroom. This might be the only chance to hear the correct form of that word.

Another frequently mentioned sub-category was the teachers’ belief that they would benefit from providing OCF due to their openness to learning and development, cited by 37 teachers (55%). Additional learner factors that motivated teachers in providing OCF included the perception that learners would not be bothered by corrective feedback (cited by 34 teachers, 51%), the belief that teachers understood the student's intention and therefore aimed to provide the target structure (29 teachers, 43%), and the interpretation of a student's look or tone as confused (26 teachers, 39%). Learners’ previous experiences in learning English were also noted as a sub-category that encouraged the provision of corrective feedback, cited by 21 teachers (31%).

The teachers in this study refrained from providing OCF when they were concerned about the learners’ feelings. This was cited by 41 teachers (61%). Additional learner factors identified in the stimulated recall interviews included the learner's low level of proficiency in English (36 teachers, 54%), the teacher's interpretation of learners’ vulnerability (32 teachers, 48%), and their perceived low level of readiness for the target linguistic item (23 teachers, 34%).

Table

(5)	Teacher 18: I didn’t want to provide the correct form there. I noticed it but thought about the proficiency level of the student. She is not ready developmentally. It is too much for her. You know what I mean?

Contextual Factors. The contextual factor most frequently cited as motivating the provision of OCF was institutional expectations, mentioned by 58 teachers (87%). This factor was mentioned multiple times by 46 teachers during the stimulated recall interviews, with each instance corresponding to a different error. Another contextual factor frequently mentioned in the stimulated recall interviews was the teacher's classification of common mistake types by Turkish EFL learners, cited by 47 teachers (70%). Additionally, this classification was sometimes linked with the perceived risk of fossilisation, as indicated by 32 teachers (see Exerpt 6 below).

Table

(6)

Teacher 58: Turkish learners have trouble with that ending [-mine]. They get confused, especially in their early development. At this level [A2], it is so common. I wanted other students to hear it too when I insisted on the pronunciation of that word [determine]. I repeated it more than once, not only for that student but also for others. It is a very common mistake.

Additional contextual factors cited by the teachers included the linguistic focus in the syllabus (44 teachers, 66%), errors that changed the meaning of what the student was communicating (34 teachers, 51%), instances where some students used Turkish in a class that required only English (31 teachers, 46%), and situations where a student paused in the flow of speech for more than 3–4 seconds (28 teachers, 42%).

The most frequently cited contextual factor that deterred teachers from providing OCF was the type of activity, specifically speaking or discussion, mentioned by 34 teachers (51%). Some teachers believed that due to the nature of their lessons, they occasionally overlooked certain grammatical mistakes (see excerpt 7). Additional contextual factors that discouraged teachers from providing OCF included errors that did not change the meaning of the intended communication (mentioned by 29 teachers, 43%), errors that occurred in the middle of a student's speech (27 teachers, 40%), errors that were not the linguistic focus of the course syllabus (20 teachers, 30%), or errors that could not be corrected due to a lack of sufficient time (16 teachers, 24%).

Table

(7)	Teacher 09: We are in a speaking lesson. I sometimes ignore minor grammar issues because I don’t want to be obsessed with accuracy. In speaking, what we need is fluency. Even though I noticed that mistake, I ignored it because it wasn't about speaking.

Teacher Factors. The most frequently cited teacher factor influencing the provision of OCF was the teacher's desire to hear the correct form from the students, acknowledged by 56 teachers (84%) (see Exerpt 8). Additionally, in the stimulated recall interviews, teachers’ previous teaching experience was identified as encouraging them to believe that certain errors needed correction for elimination, as mentioned by 41 teachers (61%). The teachers’ previous language learning experience emerged as another important sub-category, highlighting that teachers’ practices of OCF are also influenced by their experiences as language learners themselves, as mentioned by 34 teachers (51%).

Table

(8)	Teacher 51: I believe that mistakes provide very good opportunities for learning. I kind of like it when my students make mistakes. Then I ask them to repeat the correct form because it is how they will learn it. This will quicken the process of development.

The most frequently mentioned teacher factor that discouraged them from providing immediate OCF was their preference for delayed corrective feedback, either when learners finished their turn or when the activity was over (41 teachers, 61%, see example [9]). Other teacher-related sub-categories that influenced teachers’ practices of OCF included classroom management (38 teachers, 57%), their limited knowledge (23 teachers, 34%), and their teaching beliefs about language learning and error correction (18 teachers, 27%). Some teachers also believed that an excess of OCF would be counterproductive to the goals of the speaking course, leading them to refrain from providing feedback, as mentioned by 11 teachers (16%) (see Excerpt 9).

Table

(9)	Teacher 11: I don’t always provide feedback to my students. Sometimes during an activity, I provide feedback at the end of that activity. It might be more beneficial. I wait until the end so the student can finish his sentence. That also increases their fluency.

4.2. The relationship between teachers’ individual differences and their provision of OCF

Research Question 2 explored the predictive power of individual differences, specifically teaching experience, additional training, and educational background, on the factors influencing teachers’ decisions while providing OCF in speaking classes. To investigate the interaction between individual differences and teachers’ decision-making processes during the provision of OCF, descriptive statistics were employed. Mean scores and standard deviations for each individual difference were calculated and are presented in Table 2. The findings indicated that, for instance, less experienced teachers exhibited a greater reliance on contextual factors (M = 6.53) compared to other factors. In contrast, experienced teachers identified contextual factors with the lowest frequency (M = 4.44). The mean values for the three factors predicting teachers’ decisions – educational background, additional training, and teaching experience – are presented in Table 2.

Table 2. Means for each individual difference in three factors that affect the decisions of teachers.

Individual difference	Learner factors	Contextual factors	Teacher factors
	Mean (SD)	Mean (SD)	Mean (SD)
Non-ELT	6.00 (1.466)	5.89 (1.370)	6.07 (1.585)
ELT	5.18 (1.392)	5.03 (1.262)	5.13 (1.189)
-additional training	5.69 (1.424)	6.09 (1.146)	5.84 (1.588)
+additional training	5.38 (1.518)	4.74 (1.238)	5.24 (1.232)
Less experienced	5.73 (1.413)	6.53 (.730)	6.00 (1.486)
Experienced	5.36 (1.515)	4.44 (.998)	5.14 (1.291)

Following the presentation of mean scores, additional multiple regression analyses were performed for each individual difference to ascertain their predictive influence on the three factors. In essence, the aim was to investigate whether, for instance, teachers with less teaching experience were more inclined to rely on any of the factors when providing OCF compared to those with more experience.

Separate multiple regression analyses were performed for each factor, and the results indicated no significant relationship between individual differences and learner factors. Despite the absence of statistical significance, there were variations in the means among groups with different individual differences, as outlined in Table 2. Specifically, the learner factor exhibited no significant association with educational background (p = .058), additional training (p = .895), and teaching experience (p = .485) (refer to Table 3). The adjusted R² indicated that these three predictors collectively accounted for 8.5% of the variance in learner factors.

Table 3. Results of multiple regression analysis of teachers’ individual differences and learner factors.

	R	R^2	F
	.291	.085	1.911
	B	T	Sig.
Constant	6.115	18.872	.000
Educational background	−.790	−2.121	.058
Additional training	.057	.133	.895
Teaching experience	−.292	−.702	.485

In a separate multiple regression analysis examining the influence of teachers’ individual differences on contextual factors, the results indicated a significant prediction by teaching experience (p = .000). As shown in Table 4, less-experienced teachers reported considerably higher mean scores for contextual factors (M = 6.53) compared to their experienced counterparts (M = 4.44). Nevertheless, no significant predictive effect of educational background (p = .071) and additional training (p = .325) was identified for contextual factors. The adjusted R² indicated that the three individual differences collectively explained 63% of the variance in contextual factors.

Table 4. Results of multiple regression analysis of teachers’ individual differences and contextual factors.

	R	R^2	F
	.798	.636	36.172t
	B	T	Sig.
Constant	6.857	36.062	.000
Educational background	−.533	−2.435	.071
Additional training	−.248	−.992	.325
Teaching experience	−1.877	−7.744	.000

Another multiple regression analysis was conducted to assess the predictive power of individual differences on the teacher factor (Table 5). The results indicated that none of the individual differences significantly predicted teacher factors (educational background, p = .091; additional training, p = .994; teaching experience, p = .059). The adjusted R² demonstrated that the three individual differences collectively explained 17% of the variance in teacher factors.

Table 5. Results of multiple regression analysis of teachers’ individual differences and teacher factor.

	R	R^2	F
	.415	.172	4.299
	B	T	Sig.
Constant	6.418	21.287	.000
Educational background	−.834	−2.406	.091
Additional training	−.003	−.008	.994
Teaching experience	−.744	−1.925	.059

5. Discussion

This study confirmed that teachers’ OCF decisions were systematic and organised, primarily grounded in sensible judgments, principles, and beliefs about OCF, aligning with observations in other studies (e.g. Ha and Murray 2023). Three key factors – learner, contextual, and teacher – were identified as influential in shaping EFL teachers’ decisions when delivering OCF (Gurzynski-Weiss 2016). Furthermore, the study explored 14 factors that motivated EFL teachers to offer OCF and another 14 factors that dissuaded them from doing so. Notably, some of these sub-factors resonated with findings from the works of Atai and Shafiee (2017), Gurzynski-Weiss (2016), and Sluman (2014). Nevertheless, prior research had also indicated that some distinct sub-factors were influenced by contextual constraints (Ha and Murray 2023) or lesson types, including grammar and speaking lessons (Ha 2023a).

The primary results suggest that EFL teachers in this study tended to deliver OCF due to factors associated with learners. They felt that learners expected it (the most prevalent reason) (e.g. Davis 2003), believed that learners benefited from it (e.g. Li 2010, for reviews and meta-analysis), and thought that learners were not bothered by it (Rassaei 2015). Additionally, teachers in our study acknowledged learners’ intentions or perceived their confused looks as signals for assistance, prompting them to opt for providing OCF (Yang 2016). Indeed, teachers’ past experiences as English language learners, particularly in high school, emerged as a factor motivating the provision of OCF, aligning with findings from previous studies (e.g. Hernandez Mendez and Reyez Cruz 2012). The recurrence of these sub-factors emphasises their significance across different contexts.

Similarly, learner-related factors such as emotions (Rassaei 2015), low proficiency level (Yang 2016), vulnerable personality (Lemak and Valeo 2020; Rassaei 2015), and low developmental readiness for the target linguistic item (Yang 2016) emerged as discouraging factors for teachers in providing OCF in our study, aligning with findings from previous research (e.g. Hernandez Mendez and Reyez Cruz 2012). These factors underscore the complex interplay of learner characteristics in shaping teachers’ decisions regarding OCF. Some of these reasons were also reported in Gurzynski-Weiss (2016) and Sluman’s (2014) studies with different frequencies and order of importance as reported by their participants. Furthermore, Hernandez Mendez and Reyez Cruz (2012) have documented comparable apprehensions related to the deterrents preventing teachers from administering OCF, such as low proficiency levels and developmental readiness. Our research contributes to the existing body of literature by offering more elaborate justifications and a more rigorous analysis grounded in the assumptions of a relatively extensive cohort of educators.

Concerning contextual factors, the most commonly cited motivation for delivering OCF was institutional expectations. Storch (2018) delves into this matter extensively, highlighting it as a constraint in teachers’ practices related to written corrective feedback. Other contextual factors were the teacher's classification of common mistake type (Ha and Murray 2023), the linguistic target in the syllabus (Basturkmen, Loewen, and Ellis 2004), an error that changed the meaning/message (Kamiya 2016), use of L1 (Sato 2019) or a pause in the flow of speech lasting more than 3–4 seconds (Bao 2019). Nevertheless, the contextual factors identified in this study, as well as those in Gurzynski-Weiss’ investigation (2016), exhibited a diverse range. The motivation for teachers to offer feedback stemmed from institutional expectations, which may be attributed to high expectations regarding the linguistic progress of learners within the particular universities where the present study was conducted.

Contextual factors that deterred the provision of OCF included the nature of the activity, such as speaking or discussion (Yuksel, Soruc, and McKinley 2021), errors that did not change the learner’s message (Hernandez Mendez and Reyez Cruz 2012), errors that were not part of the linguistic target outlined in the syllabus for that day (Basturkmen, Loewen, and Ellis 2004), and the timing of the error (Yilmaz and Sagdic 2019), particularly when it occurred in the midst of student speech or when there was insufficient time to provide corrective feedback during the lesson.

In Gurzynski-Weiss' (2016) study, teachers also identified four of the factors mentioned in our study, except for activity type, which emerged as the most frequently reported factor in our study. Likewise, in Ha and Nguyen's (2021) study, Vietnamese teachers demonstrated a tendency to be more discerning in their selection of error types compared to their students. They asserted that prioritising significant errors, particularly those related to the focus of their lessons, was essential. However, the experienced ESL teacher in Sluman's (2014) study expressed unfavourable attitudes towards offering OCF in fluency-based speaking classes, mirroring the sentiments conveyed by most of our participants.

The teacher-related factor most commonly mentioned as encouraging the delivery of OCF was the inclination to hear the correct form from the students. This inclination may be rooted in the notion of ‘habit formation', which posits that learners are prompted to produce target-like utterances after making a mistake (Smith and Loewen 2018). We contend that further research on the general beliefs of teachers in language learning is needed to delve deeper into this aspect. Additional teacher-related factors highlighted in this study included the impact of teachers’ prior teaching experience, which instilled a belief that certain errors required correction for elimination (Hernandez Mendez and Reyez Cruz 2012), as well as the influence of teachers’ previous language learning experiences (Junqueira and Kim 2013).

Teacher-related factors that hindered the immediate provision of OCF included the preference of teachers for delayed feedback, as observed in studies by Fallah and Nazari (2019) and Olmezer-Ozturk (2019). Additionally, issues deemed more crucial to classroom management, limited knowledge, or performance on the part of the teacher (Gurzynski-Weiss 2016), and the influence of teaching philosophy and beliefs regarding language learning or error correction were reported as deterrents. Some teachers expressed the notion that an excess of corrective feedback might be counterproductive to the goals of the speaking course, as noted in studies by Atai and Shafiee (2017) and Hernandez Mendez and Reyez Cruz (2012).

Teachers’ decisions regarding the provision of feedback on learners’ errors are shaped by factors beyond the errors themselves. Equally significant are the learning context, the students involved, and the teacher's own beliefs about the process of language acquisition. These underlying factors interact with ingrained expectations, originating from both students and the institution, to influence the way teachers provide feedback. This leads us to consider the broader question of the teacher's overarching philosophy on language learning and how corrective feedback practices alignwiththat framework. While this study delved into the immediate factors influencing feedback practices, it unveils a compelling avenue for further research: understanding how the interaction between teachers’ pedagogical beliefs and broader expectations manifests in real classroom situations.

The second research question in this study aimed to find out the predictive influence of individual differences – specifically, teaching experience, additional training, and educational background – on the factors that impact teachers’ decision-making in the context of administering OCF during speaking classes. Essentially, the objective was to explore, for instance, whether teachers with limited teaching experience exhibited a greater tendency to depend on any of the identified factors in delivering OCF, in contrast to their counterparts with more extensive teaching experience.

The findings revealed that, except for teaching experience, there was no statistically significant correlation between individual differences and the aforementioned factors. The results revealed that less experienced teachers demonstrated a higher dependence on contextual factors (mean = 6.53) compared to other influencing factors, meaning that they leaned more on contextual factors to navigate the complexities of providing OCF and that they relied heavily on the immediate classroom environment and situational cues rather than drawing from their broader pedagogical repertoire. Conversely, experienced teachers identified contextual factors with the lowest frequency (mean = 4.44), which meant that they relied on their established pedagogical frameworks and showed reduced sensitivity to immediate situational cues, underscoring their ability to maintain a pedagogical consistency that transcends specific classroom circumstances – an indicative of their deeper pedagogical foundation and a more strategic approach that they adopt when providing OCF. This result substantiates the conclusions drawn by Polio, Gass, and Chapin (2006) and Yuksel, Soruc, and McKinley (2023), revealing a significant correlation between teachers’ years of experience and the cognitive frameworks influencing their attitudes towards providing corrective feedback in oral communication.

6. Conclusion

It is crucial to consider the limitations of the study when interpreting both the findings and their implications. Although the researchers adhered to a stringent schedule for interview planning in order to mitigate the impact of time on stimulated recall, conducting interviews immediately after the recorded hour, preferably on the same day, could have enhanced the efficacy of the stimulus (video recordings). Moreover, the coding of teachers’ individual differences, particularly teaching experiences, may be subject to varied interpretations, as different researchers classify the ‘year of teaching experience’ in diverse ways. The coding process can exhibit variability. Therefore, additional studies in different contexts are warranted to investigate this aspect further.

We believe this study has made a significant contribution to the existing body of literature by examining the factors influencing EFL teachers’ decisions regarding OCF practices, as well as the influence of teachers’ individual differences on their decisions in this regard. Furthermore, the study revealed a notable correlation between teachers’ years of experience and contextual factors. Our research has underscored the fact that teachers do not haphazardly provide corrective feedback on learners’ errors; instead, their decisions are intentional and systematic, guided by certain principles (Ha and Murray 2023). Our findings also revealed that many instructors in our study provided OCF because of the expectations of the learners and institutions. More importantly, contextual factors were more prevalent among less experienced instructors. These two major implications require further investigation as it might be assumed that rather than believing in its effectiveness pedagogically, some of the instructors provide OCF because they think that they should do it. This finding would also imply that a more comprehensive treatment of CF practices might be needed in pre-service teacher education programmes so that even less experienced instructors would rely more on their own pedagogical frameworks while providing OCF rather than following what their learners and institutions demand.

Acknowledgements

Dogan Yuksel's time on this output was supported by a UKRI Future Leaderships Fellowship awarded to Anna Kristina Hultgren [grant number MR/T021500/1].

Disclosure statement

No potential conflict of interest was reported by the author(s). We, hereby, confirm that the work described has not been published previously, that it is not under consideration for publication elsewhere, that its publication is approved by all authors and tacitly or explicitly by the responsible authorities where the work was carried out, and that, if accepted, it will not be published elsewhere in the same form, in English or in any other language, including electronically without the written consent of the copyright-holder.

Data availability statement

Data arenot currently available in a depository. However, the authors would happily share it via contact email if requested.

Notes

1 Years of teaching experience vary in the literature. Some take the cut-off year between the inexperienced and experienced status as two (McNeill 2005), three (Mok 1994), five (Tsui 2003) or seven (Cundale 2001; Gurzynski-Weiss 2010).