Understanding the development and implementation of teachers ’ beliefs of written corrective feedback: A study of two novice transborder teachers in

This study explored how two novice transborder teachers in mainland China ’ s public secondary schools developed their written corrective feedback (WCF) beliefs and applied them in practice. Utilising WCF samples, interviews, teachers ’ voice memos, and documents, it was found that teachers ’ prior learning experiences with WCF and contextual factors significantly influenced their WCF beliefs. Notable belief-practice inconsistencies were observed in the provision of direct or indirect WCF, highlighting the adjustments these teachers made in response to contextual factors. The study offers insights for enhancing second language (L2) writing pedagogy training in language teacher education.


Introduction
Teachers' written corrective feedback (WCF) is defined as a written response aimed at correcting inaccurate usage or providing information on where errors have occurred, the cause of the error, and how it may be corrected (Bitchener & Storch, 2016).WCF has been widely accepted as an important method that contributes to the improvement of students' second language (L2) writing accuracy and writing skills (Ferris, 2011;Lee, 2017).Although WCF remains a contentious topic in the literature, recent research has shifted the focus from assessing its effectiveness to investigating the specific type(s) of WCF that can generate the best pedagogical results (Brown et al., 2023;Thi & Nikolov, 2021).
Despite previous research on various approaches to WCF, there has been limited scholarly attention given to a critical factor-the teachers themselves (Ferris & Kurzer, 2019).Little has been known about the beliefs that guide teachers' provision of WCF and how these beliefs shape their practices in authentic work contexts (Riazi et al., 2018;Yu, 2021;Zheng et al., 2022).As active agents in the pedagogical decision-making process, teachers exert a profound influence on their students' learning experience (Borg, 2006(Borg, /2015;;L. Li, 2020).Incorporating teachers' perspectives into existing research can contribute to the development of more effective WCF pedagogy to support L2 writing instruction (Ferris & Kurzer, 2019).
Within the limited body of research examining WCF from teachers' perspectives, novice transborder teachers have been an understudied group of teachers, despite their growing numbers in the field of English language teaching (ELT) over the years.In specific, in English-dominant contexts, formal language teacher education programmes, especially those offered at a postgraduate level (MATESOL), are characterised by a large number of English L2 speakers coming from the world (Hennebry- Leung et al., 2019).These student teachers, upon graduating and returning to their home countries to teach English, are known as transborder teachers (Kamhi-Stein, 2009).During their initial three years of teaching (Farrell, 2012), they are referred to as novice transborder teachers.
Generally speaking, this group of teachers face two main challenges, adding layers to their teaching complexities.First, they encounter common difficulties experienced by novice teachers transitioning from student teachers to real-world teachers (Farrell, 2019;Yu et al., 2020).Second, they also confront the intricacies of adapting what they have learnt in one pedagogical context to another (Hennebry-Leung et al., 2019;Trent, 2020).However, little is known about the real-life situations they encounter at work and how they manage their beliefs and practices in those situations.This could result in a limited grasp of the current ELT practices, which may further hinder teacher educators in their efforts to prepare student teachers from similar backgrounds (L.Li, 2020).
In order to bridge the identified gaps, the present study was designed to draw on the language teacher cognition framework (Borg, 2006(Borg, /2015) ) to investigate two novice transborder teachers' beliefs and E-mail address: xl487@exeter.ac.uk.

Contents lists available at ScienceDirect
Teaching and Teacher Education journal homepage: www.elsevier.com/locate/tatehttps://doi.org/10.1016/j.tate.2024.104519Received 23 June 2023; Received in revised form 5 January 2024; Accepted 10 February 2024 practices of WCF in secondary school settings in mainland China.It aims to (a) understand how novice transborder teachers develop their beliefs of WCF, (b) explore how they apply their beliefs into practice, and (c) examine the factors influencing the implementation of their WCF beliefs in their professional contexts.It is important to note that this research did not intend to generalise findings to this group of teachers or the complex teaching realities in mainland China.Instead, its goal is to provide rich details for readers to evaluate the relevance of this research to their situations and potentially benefit from the research findings.

Novice transborder teachers
Novice transborder teachers have two features that distinguish them from others.First, they are English L2 speakers, which means their linguistic repertoire includes at least one other language besides English (Mahboob, 2017).Although teachers who are English L2 speakers play an important role in the field of ELT across the world (Kamhi-Stein, 2009), there is a notable paucity of studies examining their beliefs and practices in their professional settings, especially in EFL contexts where English is learnt "both for linguistic and functional reasons" (L.Li, 2017, p. 5).
Second, transborder teachers are distinguished by their educational backgrounds, which involve acquiring MATESOL education from universities in ESL contexts, such as the United States, or in regions where English holds official status and the teacher education is significantly influenced by ESL contexts, exemplified by Hong Kong.These teachers' educational experiences are characterised by a transition across "geographical, educational and societal boundaries" (Trent, 2020, p. 316), which may starkly contrast with their own schooling experiences and their future teaching settings (Hennebry-Leung et al., 2019;Liyanage & Bartlett, 2008;Nguyen & Walkinshaw, 2018).Despite the growing presence of those teachers, the literature remains scant on how their unique educational backgrounds shape their pedagogical beliefs and practices in real-world teaching contexts.This study was designed to bridge this research gap and empower this important yet underrepresented group of teachers, with the aim of advancing the development of language teacher education.

Teachers' written corrective feedback
WCF provides teachers with a high degree of flexibility, allowing them to give students "individualized attention that is otherwise rarely possible under normal classroom conditions" (K.Hyland, 2003, p. 177).In general, WCF can be divided into three categories: WCF strategy, WCF focus, and WCF scope.
WCF strategy can be broadly divided into direct WCF and indirect WCF (Lee, 2017).In direct WCF, a teacher not only locates students' errors but also gives correct answers.In indirect WCF, a teacher points out the presence of errors without providing answers.Besides, Ellis (2009) proposed a third WCF strategy, metalinguistic explanation (ME), which involves providing metalinguistic clues to explain the nature of an error, including brief descriptions of grammatical rules, and the use of code.Despite its value in organising WCF categories, a survey of the literature indicates that there is no consistency in how ME was conceptualised and operationalised in previous studies (see Bitchener & Knoch, 2010;Cheng et al., 2021, for example), which is likely due to the ambiguity of the concept of ME (Mao & Crosthwaite, 2019).In view of the purpose of this research, this study followed Cheng et al. (2021) and Mao and Crosthwaite (2019) by only adopting direct and indirect WCF as two WCF strategies.
WCF focus refers to error types a teacher decides to address in students' writing (Ferris, 2011).It can be divided into two types.One is global issues, defined as "those that interfere with the overall message of the text" (Ferris, 2011, p. 87), including errors related to ideas (i.e., personal views or intentions), content (i.e., the conveyed information), and organisation (i.e., the overall structure, paragraphs, or passages; Mao & Crosthwaite, 2019).The other is local issues, defined as those that "do not inhibit a reader's comprehension" (Ferris, 2011, p. 87), including errors relating to grammar (i.e., morphological or syntactic errors), language expression (i.e., lexical errors), and mechanics (i.e., spelling, punctuation, or capitalisation; Mao & Crosthwaite, 2019).
WCF scope refers to the extent to which a teacher gives WCF on students' writing (Lee, 2017).It can be classified into comprehensive/unfocused and selective/focused based on "the number of error categories or target structures" (S.Li & Vuono, 2019, p. 100).In specific, comprehensive/unfocused feedback refers to feedback that targets a wide range of errors, while selective/focused feedback refers to feedback that targets limited types of errors (Cheng et al., 2021).It is worth noting that the terms "a range of" and "limited" are not precisely defined.The extent can vary, ranging from feedback that focuses only on a specific error or linguistic structure to feedback that addresses all linguistic features (S.Li & Vuono, 2019).Thus, in this study, unfocused/comprehensive WCF refers to feedback that targets both global and local issues, while focused/selective WCF refers to feedback that targets either global or local issues, or specific subcategories within each error type.

L2 teachers' beliefs and practices of WCF
Teachers' beliefs refer to "what teachers know, believe and think" (Borg, 2003, p. 81).Previous research on language teacher cognition has shown that teachers' beliefs and practices are under the influence of teachers' schooling experiences, professional coursework, classroom practice, and relevant contextual factors (Borg, 2006(Borg, /2015)).Although EFL teachers account for approximately three-quarters of ELT teachers across the world (Zhang & Zhang, 2021), little is known about their beliefs and practices of WCF in EFL contexts, where the majority of learners of English are located globally (Mao & Crosthwaite, 2019;Zheng et al., 2022).Moreover, most of the available research was carried out in EFL higher education settings, resulting in a lack of understanding about the beliefs and practices of WCF among secondary school EFL teachers (Lee, 2017).
Within the limited body of research concerning secondary school EFL teachers, Lee's (2004Lee's ( , 2008) ) research offered valuable insights on the relationship between teachers' WCF beliefs and practices.Drawing on the data from questionnaires, interviews, documents, and WCF samples, Lee found that teachers' WCF practices were not always in line with their beliefs or even official guidance due to contextual factors such as local institutional cultures, the exam-oriented culture, and expectations from other stakeholders.
Research in mainland China has predominantly examined university teachers' beliefs and practices.For example, Cheng et al. (2021) identified congruences and disparities in novice university teachers' approaches to WCF.These disparities were attributed to various contextual factors such as teachers' prior learning experiences, their limited work experience, Chinese traditional culture, students' proficiency levels, and examination stress.Similarly, Mao and Crosthwaite (2019) administered questionnaires and conducted interviews with five university teachers with teaching experience ranging from seven to 26 years.They discovered inconsistencies between the participants' beliefs and practices concerning WCF focus and strategy, primarily due to contextual factors such as time constraints, heavy workloads, and teachers' perceptions of students' attitudes.Additionally, a comparison of these two studies offers insights into the possible impact of teaching experience on teachers' WCF beliefs and practices.
Despite the insights from the existing research, there is a notable variation in the educational backgrounds and/or teaching experience of the participants.The specific development and application of WCF beliefs by novice transborder teachers in mainland China's secondary schools remain underexplored.This study aims to address this research gap by focusing on two such teachers to gain deeper insights.The

Participants and context
A purposeful sampling strategy (Patton, 2015) was used to select participants that can provide rich information for in-depth study.The selection criteria included: (a) Chinese L1 speakers completing MATE-SOL education in ESL contexts, (b) teachers taught English in mainland China's public secondary schools within three years of graduation, and (c) those who agreed to take part in the study.
Nina and Daniel (pseudonyms) volunteered to participate in this study.Both earned undergraduate degrees in English-related majors from top-tiered universities in mainland China before pursuing MATE-SOL education in the United States.As L2 writers who were educated in the same pedagogical context, their experiences with writing in English began in elementary school.However, Nina mentioned that it was not until high school, particularly senior high school, that she had many opportunities to "actually write something in English," such as journal writing and completing writing tasks for official tests.Nina noticed significant improvement in her English writing skills when she was in college.She attributed this progress to the design of her major, which required most assignments to be done in English, and her commitment to practising writing in English for further education abroad.After completing a 2-year MATESOL programme and writing "numerous course papers," Nina expressed confidence in her L2 writing ability.
As for Daniel, he noted that it was not until college that he had sufficient chances to practise L2 writing: "English writing was a somewhat neglected aspect during my time in elementary school and even in high school."During college, by completing assignments in English and preparing for further education abroad, Daniel commented that his L2 writing skills improved significantly.Like Nina, Daniel also remarked that after completing his MATESOL learning, he felt confident in writing in English.
The MATESOL education Nina and Daniel completed was a 2-year programme in the United States.This programme provided mandatory courses covering both the theoretical and practical aspects of ELT, including ELT methodology, assessment, curriculum design, and second language acquisition.It also offered the flexibility to choose elective courses as per individual interests.However, at that time, it did not provide courses for teaching L2 writing.Additionally, the programme included a one-semester teaching practicum at local educational institutions, providing valuable hands-on experience.During their practicums, Nina assisted with teaching material preparation and classroom management at a local high school, while Daniel taught public speaking skills at an English language centre.Neither participated in L2 writing instruction.
During the data collection period, despite being in her second year of teaching, Nina taught at an EFL public secondary school for the first time.Before that, she taught the speaking section for the Test of English as a Foreign Language (TOEFL) at a private educational institution.The school Nina worked in was an experimental high school admitting highperforming students based on their High School Entrance Exam (HSEE) scores.In mainland China, the HSEE is a requisite for junior high school (Grades 7-9) students to advance to senior high school (Grades 10-12), with English being a core subject of assessment.The writing section generally requires students to compose a brief essay, typically 60-80 words, in response to a prompt that often involves persuasive or letter writing tasks.
Nina taught two Grade 10 classes, preparing students to take the College Entrance Exam (CEE) to pursue higher education in mainland China.She described most of her students as "beginner-level L2 writers," who had gained their writing experience primarily through preparation for the HSEE.Given that the CEE demands a higher level of L2 writing skills compared to the HSEE, it was no surprise to Nina that her students faced difficulties with the L2 writing tasks at this stage, which were derived from previous mock CEE papers.Specific to this study, the second writing task Nina assigned posed a greater challenge than the first.
Daniel was also in the second year of teaching.He taught two Grade 10 classes at the international division of a public high school, where students intended to pursue higher education overseas.He was responsible for preparing students for the International English Language Test System (IELTS) and had a demanding schedule, teaching at least one writing class per day.Daniel similarly noted that his students' earlier exposure to L2 writing stemmed from their preparation for the HSEE.As the IELTS writing requires advanced proficiency in academic writing, Daniel said that, as anticipated, most students found it "extremely challenging and difficult."Like Nina, Daniel assigned a more demanding writing task in the second set, sourced from mock IELTS papers.Appendix A shows the writing tasks assigned by the participants.

Data collection and data analysis
After obtaining the participants' consent, the data collection commenced.Due to the global pandemic, all data were collected online, following the guidance of the author's institution.Data were collected over one semester from three sources: teachers' WCF samples, two rounds of interviews, and teachers' voice memos.The interviews and voice memos were the primary data sources, yielding detailed insights into the participants' interpretations of their WCF beliefs and practices.Some documents pertinent to this study were also collected as complementary data, including local and school-level educational policies, participants' teaching plans, assessment rubrics, and writing prompts.By cross-referencing information gathered from other sources, this helped to create a more comprehensive understanding of the participants' professional contexts, which also informed the tailored development of interview questions for each individual.Overall, gathering data from various sources served two main purposes: (a) triangulating data to ensure the trustworthiness of this research (Duff, 2008) and (b) offering comprehensive contextual details for each case (Rossman & Rallis, 2017).Prior to the main study, a pilot study was conducted to revise and refine the interview questions (Appendix B).Table 1 presents a summary of the dataset used in this study.
In detail, two sets of WCF samples (10 samples per set) were collected from each participant, with the first set collected before midterm and the second one collected before the end of the semester.The samples were coded based on feedback points, defined as all writing interventions

Table 1
Summary of the dataset.
conducted by the teacher (F.Hyland, 2003).Feedback points were first categorised based on their strategy and focus (Table 2).Then, each sample was coded according to its WCF scope.To ensure coding reliability, one university EFL teacher was invited to check the coding.She and the author randomly selected 12 samples from the 20 samples collected from two participants in the first set.A high level of agreement was achieved (WCF strategy: 96.72%; WCF focus: 97.01%; WCF scope: 100%).After resolving any disagreements through discussion, the author then coded the samples independently.
When the coding results of the first set of WCF samples were ready, the first round of in-depth semi-structured interviews (Duff, 2008) was arranged at the convenience of each participant.This was to make sure that they could explain their beliefs in a contextualised way.Interviews were conducted in Chinese and audio-recorded with the participants' permission using Microsoft Teams.During each interview, the participants had access to their WCF samples if they wanted to provide examples to support their points.The same procedure was followed in the second interview.The recordings were transcribed and translated verbatim by the author and sent to the participants for member checking.Pseudonyms were used to protect privacy in all transcriptions.
Drawing on Borg's (2006Borg's ( /2015) ) language teacher cognition framework, the first interview (3 h) was designed to gather information about each participant's language learning experiences, professional coursework, classroom practices, and their professional contexts.The guiding interview questions consisted of two sections, with the first section exploring the participants' beliefs and practices regarding ELT in general and the second one focusing on WCF in specific.
The second interview (1.5 h) was conducted at the end of the semester.Following the scheme adopted by Junqueira and Payant (2015), personalised follow-up questions were asked to (a) capture any insights that the participants may have developed over the semester and (b) explore any new topics that may be found from other data sources.
In addition, during the interval between the two rounds of interviews, the participants were suggested to record at least one voice memo and send it to the author before the second interview.The content of voice memo could be anything that they thought was relevant to this research.This aimed to capture any insights that might have been overlooked during the time between interviews (Creswell & Creswell, 2018).Also, any new topics that were found from the voice memos could be discussed during the second interview.Nina recorded two voice memos while Daniel recorded one, each lasting 2-3 min.The voice memos were transcribed and translated verbatim by the author and sent to each participant for member checking.
Collectively, the transcriptions of interviews and voice memos were analysed via reflexive thematic analysis (Braun & Clarke, 2022), which involves "familiarisation," "coding," "generating initial themes," "reviewing and developing themes," "refining, defining and naming themes," and "writing up" (p.40).The current study employed this analysis approach initially with a deductive orientation according to the predetermined WCF categories and language teacher cognition framework (Borg, 2006(Borg, /2015)).Meanwhile, the coding process also remained open to new themes.Table 3 shows examples of temporary code labels used when analysing teachers' interpretations of their beliefs and practices.

Researcher positioning
It is important for researchers to be self-aware of their perspectives and potential biases that could influence the research process.I shared the same L1 and some parallel educational experiences with the participants in this study.These included our schooling in mainland China and professional learning in a MATESOL programme in an ESL context.The shared aspects could help us establish rapport and trust, as the participants may view me as a "peer" (Duff, 2008, p. 120) rather than an outsider.This can lead to more effective data collection.
That said, I also recognised the importance of maintaining a professional relationship with the participants to ensure the trustworthiness of this study.This is because our relationship and mutual knowledge may influence the co-construction of data during interviews and other interactions (Macalister, 2023;Talmy, 2011).To mitigate these effects, several strategies were employed: data triangulation, member checking, and provision of the guiding interview questions (Appendix B).Additionally, quotations from the participants were used to ensure accurate representation of their perspectives.These steps were taken to offer transparency, allowing readers to assess my involvement and the trustworthiness of this qualitative research.

Teachers' WCF beliefs and sources of their beliefs
The participants' WCF beliefs can be categorised into four aspects, namely, the necessity and effectiveness of WCF, WCF focus, WCF strategy, and WCF scope.The sources of their beliefs are discussed below each subheading.

The necessity and effectiveness of WCF
Both participants strongly believed that it was necessary to provide WCF to students.They saw this as one of the key roles of a teacher-to inform students of any errors they were unable to identify on their own.As in Nina's words, "Students make errors because most of the time they do not know those are errors.Otherwise, they will not write in that way.If I don't give WCF, I'll worry that they might become stuck in their errors."Similarly, Daniel believed that teachers remained a reliable source of learning for students, despite the abundance of learning resources available, which highlighted the necessity of providing WCF on students' writing.
However, Nina and Daniel showed different levels of confidence in the effectiveness of WCF.Nina believed that WCF was effective in helping students improve L2 writing skills.Such a belief came from her own positive learning experiences with it, as she noted that some of her teachers' WCF "did help me understand what L2 writing should be."She also ascribed this belief to her observations on the progress her students made, which reinforced her confidence in the effectiveness of WCF.
Regarding Daniel, he was sceptical about the effectiveness of WCF, which stemmed from his negative learning experiences with it: "I had

WCF focus
Both Nina and Daniel believed that global and local issues deserved equal attention.One reason for their belief was the emphasis on exams.Although their students took different exams, both global and local issues were measured in the correspondent assessment rubric.Therefore, they preferred to provide WCF on both global and local issues.
Another reason was their personal learning experiences with WCF.Specifically, Nina viewed it as "very beneficial," as it helped her "know how to write in English."Her positive learning experiences motivated her to continue using this approach as a teacher.By contrast, Daniel used this approach due to his negative learning experiences with receiving WCF that mainly addressed local issues, especially grammatical errors.Since Daniel was good with grammar, he seldom received WCF.Nonetheless, achieving a satisfying test score remained a challenging task for him: I did not receive much WCF from my teachers.However, this did not mean that I nailed it.… I was eager to know what kind of rubrics they used when grading our writing.… and what aspects were measured in the exam, so that I could work on them.But no one gave me an answer.
Motivated by his negative learning experiences, Daniel opted to take the opposite approach.That is, he focused on giving WCF on both local and global issues.

WCF strategy
Nina and Daniel proposed that the WCF strategy should vary according to error types.Specifically, they agreed that direct WCF was appropriate for local errors, while indirect WCF was better suited for global errors.
Despite the same belief they had, their sources of belief differed.Regarding Nina, her belief was influenced by her students' current English language proficiency levels and her personal language learning experiences.Specifically, Nina mentioned that many of her students had relatively limited English language skills, which made it challenging for them to self-identify local issues such as "misspellings, problematic word choices, or misuse of punctuation."Considering this, Nina believed that providing direct WCF on local issues was beneficial for her students.Regarding global issues, Nina referred to her own negative learning experiences where her teachers "poured what they thought was right on me without thinking about what I was trying to express."Influenced by such unpleasant learning experiences, Nina decided to give indirect WCF on global issues.
Regarding Daniel, his belief in providing direct WCF on local issues came from his personal learning experiences.As mentioned before, Daniel's teachers were mainly concerned with local issues, particularly grammatical errors, and gave direct WCF on those errors.Daniel commented that "Although this method did not work for me, it worked for other students."Regarding addressing global issues, Daniel's support of indirect WCF was influenced by his student-centred teaching philosophy, which was developed during MATESOL education.Daniel translated this teaching philosophy to his provision of WCF by giving indirect WCF on global issues, with the aim to "give students the directions" rather than the answers.

WCF scope
Both Nina and Daniel expressed their strong preference for unfocused WCF.They believed that it was meaningful to focus on every error they could identify rather than selectively focusing on certain error types.Nina and Daniel presented different reasons.As for Nina, she attributed this belief to her understanding of the responsibility of being a teacher.She also mentioned that the progress her students made further solidified her conviction in delivering unfocused WCF.Daniel's belief in unfocused WCF originated from his negative learning experiences, which made him question the value of focused WCF.He expressed concern that focused WCF, which was employed by his own teachers, might inadvertently lead students to believe that "writing assessment is subjective."It was Daniel's intention to enlighten his students about the existence of a defined rubric for assessing written work, rather than it being subject to the personal views or biases of teachers.Another factor shaping Daniel's belief was exam pressure.He believed that giving unfocused WCF could better help students get higher test scores.

Teachers' WCF practices and their relationship to teachers' WCF beliefs
Both participants provided WCF in practice, which was in line with their belief in the necessity and effectiveness of WCF.Thus, the following sections focus on the participants' beliefs and practices regarding WCF focus, WCF strategy, and WCF scope.

WCF focus
A close examination of the WCF samples suggested that the participants' practices of WCF focus reflected their beliefs in general.Three findings are worth mentioning here.First, as shown in Table 4, in each set, Nina and Daniel tackled both local and global issues, with a greater emphasis on the former (Nina: 82.05% in S1 and 89.27% in S2; Daniel: 80% in S1 and 72.48% in S2).Second, the amount of WCF provided increased from the first set to the second set, primarily due to the increased amount of local WCF.Third, within each error type, the issue that received most WCF changed from one set to another.

WCF strategy
As mentioned earlier, Nina and Daniel expressed their preference for providing direct WCF on local issues and indirect WCF on global issues.When examining their practices of WCF strategy (Table 5) in relation to their beliefs, both consistencies and inconsistencies were found.
Regarding Nina, consistencies were found when she addressed local issues in both sets, where she primarily employed direct WCF (S1: 100%; S2: 98.10%).Also, when addressing global issues in the second set, her practice aligned with her belief in employing indirect WCF (57.89%).However, when tackling global issues in the first set, she primarily used direct WCF (90.48%), which contradicted her belief.
Regarding Daniel, consistencies were found when he addressed global issues in both sets, where he primarily employed indirect WCF (S1: 100%; S2: 96.67%).Consistency was also found when he addressed local issues predominantly with direct WCF in the first set (93.75%).However, in the second set, he mainly used indirect WCF (58.23%), which contradicted his belief.

WCF scope
As indicated in Table 6, both Nina and Daniel primarily utilised unfocused WCF (Nina: 80% in S1 and 100% in S2; Daniel: 70% in S1 and 80% in S2).Such practices were consistent with their beliefs, despite only a few samples being coded as focused WCF.

Teachers' interpretations of WCF practices and factors leading to modifications in their implementation process
As mentioned before, both participants exhibited consistencies and inconsistencies between their beliefs and practices of WCF strategy and WCF scope.An analysis of their interviews and voice memos revealed a X. Liu complex set of factors that influenced their decisions on whether to enact their beliefs or adopt alternative methods in their authentic work contexts.This section first explores the factors that influence both participants, and then examines the unique variables impacting each individual.
The participants considered students' actual writing performance as a significant influencing factor.Furthermore, they delved deeper into the factors affecting students' writing and tailored their WCF methods accordingly.The underlying factors mentioned by them included students' proficiency in L2 writing, the complexity of the writing task, students' familiarity with the task type, and students' learning attitudes and needs.
Regarding Nina, even though she preferred to give indirect WCF on global issues, these factors accounted for the inconsistencies in her approach.This was especially noticeable in the first set, where she offered more direct WCF on global issues.More specifically, after the first writing task, Nina observed challenges among students with the use of conjunctions and idea organisation.She was concerned that the second task would make those issues more noticeable.This was because the second task, with its higher word count and less content and language support, would demand more advanced writing and comprehension skills.This was also due to the students' lack of familiarity with the type of the second writing task.Considering this, she provided more direct WCF on global issues in the first set to help students "build a solid foundation of English."This approach was intended to equip them better for the complexity of the second assignment.
The belief-practice inconsistencies identified in Daniel's case could also be explained by these factors.Daniel's second writing task was a step up in complexity, offering less content and language support and requiring a higher word count than the first.Despite this, Daniel was confident, noting that "Task 2 shouldn't be a challenge, as we've practised similar tasks multiple times and students are very familiar with this type of task."However, he did observe that some students demonstrated negative attitudes towards English writing, leading to a higher occurrence of errors, particularly grammatical ones, in the second task.
Moreover, as a transborder teacher with learning experience from overseas universities, Daniel was acutely aware of the potential challenges his students might face with L2 writing in international academic settings.He compared the IELTS writing tasks to the more demanding essays students would face in college, questioning, "If they can't handle the IELTS, how will they complete future assignments?"Disappointed by his students, Daniel resolved to become "stricter" by using indirect WCF more extensively.He believed this approach could motivate students to correct their own local errors more actively and pay closer attention to their writing.This was evident in the higher amount of indirect WCF he gave on local issues in the second set.
Another influencing factor was heavy workload, which was identified in Daniel's case.It influenced how he prioritised addressing global and local issues.Although Daniel believed that both issues deserved equal attention, he prioritised checking for global issues, especially content and ideas, in practice.Specifically, he only provided local WCF on essays where content and ideas were appropriate for the assigned topic.Although this sequence was not explicitly stated by Daniel, it could be inferred from his emphasis on the importance of staying on topic in writing: "If it [an essay] is off topic, it will be meaningless."Rather than continuing to point out individual errors in an off-topic essay, Daniel expressed his preference for other feedback forms such as a face-to-face meeting, considering his busy schedule.

Discussion
This section discusses the research findings in relation to the research questions.The first subsection explores the participants' beliefs and the sources of those beliefs (RQ1).The following section addresses RQ2 and RQ3 concurrently, as the analysis of the relationship between the participants' WCF beliefs and practices (RQ2) establishes a foundation for and is logically connected to the examination of the potential factors influencing the participants' implementation of their beliefs into practice (RQ3).

Novice transborder teachers' WCF beliefs and the sources of their beliefs
This research explored the participants' WCF beliefs from four aspects.Regarding the necessity and effectiveness of WCF, while the participants had varying levels of confidence in its effectiveness, they both strongly agreed on the necessity of providing WCF.They believed that it was an important responsibility of a teacher to bring to students' attention any errors they might have missed.This sense of responsibility to provide WCF was also reported in previous research (e.g., Bitchener, Note.S1: The first set of WCF samples.S2: The second set of WCF samples.2012; Cheng et al., 2021;Mao & Crosthwaite, 2019) and was associated with the goal of preventing the fossilization of errors (Ferris, 2011).Also, this finding echoes Bitchener and Storch's (2016) observation that although some teachers are sceptical of the value of WCF, most teachers agree that some students can benefit from some of the feedback and that "some improvement is better than no improvement or no opportunity for improvement" (p.3).
Regarding the specific approaches to WCF provision, the participants expressed similar belief in addressing both global and local issues (i.e., WCF focus).They also agreed on providing direct WCF for local issues and indirect WCF for global issues (i.e., WCF strategy).These findings are different from previous research where most participants preferred to prioritise global issues (e.g., Cheng et al., 2021;Junqueira & Payant, 2015) and favoured direct WCF (e.g., Mao & Crosthwaite, 2019) or indirect WCF (e.g., Saeli & Cheng, 2021).
Moreover, it is worth noting that the purpose of this qualitative study is not to generalise how teachers provide WCF.Instead, it places greater value on examining the development of teachers' corresponding beliefs, with a focus that extends beyond the specific WCF approaches that they adopt.Drawing on the data from interviews and teachers' voice memos, two important factors were found to significantly influence the development of the participants' WCF beliefs.
First, it was found that the participants' L2 writing learning experiences with WCF had a considerable influence, with positive and negative experiences playing distinct roles.The influence of positive learning experiences was identified in the case of Nina.It contributed to her beliefs in and applications of the WCF approaches that enhanced her learning experiences.This finding aligns with broader language teaching research that suggests teachers often model their classroom teaching after their own positive learning experiences (e.g., Channa, 2020;Davin et al., 2018;Rabbidge, 2017).
Regarding negative learning experiences, it was found that the participants critically examined their prior learning experiences and generally opted for different approaches than the ones that had led to their negative learning experiences.Specifically, Daniel's uncertainty about the effectiveness of WCF, preferences for providing WCF on global and local issues, and support for unfocused WCF were all influenced by his negative learning experiences.Similarly, due to her negative learning experiences with receiving WCF on her ideas, Nina opted for giving indirect WCF on students' global errors.Moreover, the participants seemed to evaluate their learning experiences as either positive or negative based on their personal feelings and perceptions.This is consistent with prior research (e.g., Davin et al., 2018;Rahman et al., 2020), which suggested that teachers' evaluation process was subjective.
Collectively, the findings contribute to the limited research on the schooling experience of L2 teachers (Yigitoglu & Belcher, 2014) by illustrating how the "apprenticeship of observation" (Lortie, 1975) influenced novice transborder teachers' decision-making process regarding WCF.The findings also corroborate Moodie's (2016) argument that language teachers were motivated to diverge from the teaching models that resulted in their own negative learning experiences as students.
Moreover, it was observed in Daniel's case that, in addition to personal preferences, he also factored in his classmates' learning achievements when deciding whether to adopt the WCF approaches used by his teachers.That is, even if a WCF approach resulted in negative learning experiences or outcomes for him, he would still be willing to utilise it if it benefited his classmates' learning.This finding provides a novel perspective on the factors influencing novice teachers' choice of traditional teaching methods.The existing literature has highlighted a tendency among novice teachers to conform to conventional teaching approaches, often as a response to practical limitations (e.g., Cheng et al., 2021;Kang & Cheng, 2014;Urmston & Pennington, 2008).This study, while acknowledging these patterns, suggests that the pedagogical choices of novice teachers might also be shaped by their critical assessment of the outcomes of traditional teaching methods.They likely chose the ones they deemed effective, either for themselves or for most students.
Another identified influencing factor was contextual variables, including exam pressure and students' English language proficiency levels.Similar reasons were also reported in previous research (e.g., Cheng et al., 2021;Mao & Crosthwaite, 2019).However, instead of referring to those contextual factors as constraints, the participants interpreted how they developed their beliefs based on those factors.For example, they mentioned that they favoured providing WCF on local and global issues, as these were evaluated in the official writing rubrics.They also referred to their students' current limited English proficiency levels when interpreting their WCF beliefs.They displayed an understanding of the challenges their students grappled with in L2 writing, suggesting their awareness of what they can expect from students at this stage of English learning.
Additionally, in Daniel's case, he mentioned that the student-centred teaching philosophy he developed in the MATESOL programme influenced his preference for providing indirect WCF on global issues.Corroborating previous research in general language teaching (e.g., D. Li & Edwards, 2013;L. Li, 2012), this finding provides specific evidence on how teachers' professional education could positively influence their WCF beliefs.However, compared to the influence of teachers' prior learning experiences with WCF, this study suggests that the influence of teacher education was limited.One potential explanation may be the insufficient training provided for teaching and assessing L2 writing, an issue that persists in current language teacher education and has been warranted by many L2 writing researchers (Bhowmik & Kim, 2021;Crusan et al., 2016;Lee, 2021;Yu et al., 2022).Such a finding offers important insights for teacher educators, which will be discussed in the Conclusion section.

Teachers' application of their WCF beliefs and factors influencing this process
Different from previous research where many tensions between teachers' beliefs and practices were reported (e.g., Cheng et al., 2021;Junqueira & Payant, 2015;Mao & Crosthwaite, 2019), this study found that the participants' practices generally reflected their beliefs.A potential factor contributing to this finding may stem from the participants' adeptness at contextualising their beliefs, which was noted in the study conducted by Sanchez and Borg (2014) on experienced teachers.
In specific, rather than stating their ideal beliefs in a decontextualised way, the participants in this study described their beliefs and how they formed those beliefs in relation to their authentic work contexts, even though they were still novice teachers.Moreover, it is relevant to point out that while this study refers to novice teachers as those with less than three years of teaching experience after graduation, the participants were in their second year of teaching.This suggests that the potential influence of their initial year of teaching could be considered.However, this influence was not observed in their decision-making process regarding WCF.Instead, it was seen to affect their overall teaching, such as classroom management, which is beyond the scope of this study.
The identified belief-practice consistencies could also be explained by how teachers' practices were interpreted in this study.For example, in this study, both participants provided more local WCF than global WCF, despite their belief in the equal importance of global and local issues.However, these practices were not viewed as inconsistent with their beliefs, as some previous research indicated.In some earlier studies, the number of feedback points given on a particular type of error was often linked to the level of importance a teacher placed on it (e.g., Cheng et al., 2021;Junqueira & Payant, 2015;Lee, 2008;Mao & Crosthwaite, 2019).By contrast, this study offers a new perspective.It suggests that when teachers value global and local issues equally and support unfocused WCF, the feedback points assigned for different error types are mainly determined by students' writing performance.In such cases, it might be inadequate to evaluate the alignment between teachers' beliefs and practices solely based on the number of global and local WCF they provided.
Additionally, as noted by Junqueira and Payant (2015), a single feedback point may be sufficient to address multiple global issues, whereas this is not always the case for local issues.This also accounts for the higher frequency of local WCF compared to global WCF in teachers' WCF practices.Taken together, this study suggests that the quantity of feedback points given on a specific error type may not precisely reflect or fully encapsulate teachers' beliefs, particularly if teachers emphasise the equal importance of addressing global and local errors in their beliefs.
Apart from the consistencies, this study also found certain beliefpractice inconsistencies and explored possible reasons.The identified inconsistencies were mainly related to how the participants adapted their WCF strategies in response to contextual factors such as students' writing performance and exam pressure.While similar contextual factors were reported in earlier research (e.g., Ferris, 2014;Lee, 2008;Zhao & Zhang, 2022), this study provided additional analysis on the underlying reasons for students' writing performance, suggesting that it was further influenced by students' English language proficiency levels, the complexity of the writing task, students' familiarity with the task type, and students' learning attitudes and needs.This study also discussed how the participants adapted their WCF strategies in response to changes in those factors as the students' learning progressed.For instance, even though Nina advocated for providing indirect WCF on global issues, she gave more direct WCF on global issues in the first set, intending to equip students for more difficult and demanding writing assignments.
In addition to these factors, in Daniel's case, his workload and students' future use of English were also found to have an impact.For example, due to his heavy workload, he prioritised addressing global issues especially content rather than local issues.Additionally, while Daniel preferred addressing local issues with direct WCF, he recognised the impending challenges his students might face with the intensive use of L2 writing in international educational settings.This led him to provide more indirect WCF in the second set to more actively engage his students in the learning process.Furthermore, it seems plausible to infer that Nina's awareness of her students' progression to higher education within mainland China-a context where advanced English writing skills are not universally requisite across all majors-might also influence her WCF approaches, as Nina appeared to focus primarily on preparing students for the writing tasks of the CEE, although no explicit evidence was found.Overall, the participants' practices of WCF strategy coincide with researchers' suggestions that teachers should consider students' individual needs and characteristics when providing feedback (Bitchener & Storch, 2016;Ferris, 2011;McLellan, 2021).
Moreover, the inconsistencies found in this study align with previous research indicating that novice teachers may possess beliefs that are not yet fully established and can be influenced by external factors, such as classroom dynamics (Cheng et al., 2021;Huang et al., 2021;Yu et al., 2020).One possible explanation is that novice teachers may not have enough teaching experience to effectively handle unexpected demands or changes that may arise in their work context, as compared to their more experienced colleagues (Borg, 2006(Borg, /2015;;Farrell, 2019;Tsui, 2009;Yu et al., 2020).This finding points to the significance of motivating novice transborder teachers to reflect on their beliefs and practices, which can increase their understanding of how context can shape their beliefs and impact their decision-making process.Any belief-practice mismatches that surface can serve as a springboard for novice transborder teachers to dive into professional development, which can equip them to tackle potential obstacles arising from a lack of teaching experience.Moreover, this kind of reflection can bolster the confidence of novice transborder teachers, encouraging them to apply the knowledge gained from their teacher education in ESL settings into their everyday EFL teaching.The importance of reflective practice is well-established in language teacher education (Farrell, 2019;L. Li, 2020), while this study highlights its value particularly for novice transborder teachers.

Conclusion
Drawing on the language teacher cognition framework (Borg, 2006(Borg, /2015)), this study sheds light on the often-overlooked WCF beliefs and practices of novice transborder teachers transitioning from MATE-SOL programmes in ESL contexts to EFL secondary school settings.The findings suggest the intricacies of their professional contexts and the adaptation of their WCF practices to these contexts.The implications of this research extend to enabling teachers with similar educational backgrounds to critically reflect on their beliefs and practices.Moreover, it offers insights for teacher education and recommendations for practitioners in the field of ELT.
Different from previous studies, this study found that, in general, the participants' WCF practices echoed their beliefs, which were attributed to their abilities to shape the beliefs not only through critical reflections on their prior L2 learning experiences but also in response to their current teaching contexts.However, a main belief-practice inconsistency was identified regarding WCF strategy, suggesting the strong influence of contextual factors such as students' writing performance and teachers' workload on teachers' implementation of their WCF beliefs.By comparison, the impact of MATESOL education was implicit, which highlights the concern that L2 writing may not be adequately addressed in language teacher education.Consistent with prior research (e.g., Carless & Winstone, 2020;Crusan et al., 2016;Lee, 2017Lee, , 2021;;Yu et al., 2022), this study stresses the urgent need for teacher educators to further enhance their curriculum design.A practical way to achieve this is by incorporating more L2 writing pedagogy.Doing so allows teacher educators to better equip their student teachers for professional success.
Teachers at different career stages can also draw on these research findings as a starting point for self-reflection, especially on how authentic teaching contexts can influence teachers' beliefs and practice (Farrell, 2019).Student teachers might want to consider how the teaching context where they are learning might differ from their future professional context, and what challenges these differences could pose for their classroom teaching.It would also be beneficial to critically assess how relevant the theories and methods taught in their teacher education programmes are to their future work context.In-Service teachers, particularly novice teachers, are encouraged to contemplate their beliefs and practices, utilising any inconsistencies as catalysts for professional advancement (Borg, 2009;Mao & Crosthwaite, 2019).School leaders and administrators can also collaborate with teacher educators to support the transition of novice transborder teachers from ESL training settings to their subsequent roles as EFL teachers.
However, it is important to acknowledge that this study has limitations due to its small sample size.To further our understanding of this group of teachers, future research could include more novice teachers teaching on different educational levels as informants.Additionally, inviting both novice and experienced teachers to explore the similarities and differences between them could shed light on the influence of teaching experience on teachers' WCF beliefs and practices.

Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

X. Liu
What are novice transborder teachers' beliefs of WCF, and how do they develop such beliefs?RQ2.How do teachers provide WCF, and to what extent are their practices aligned with their beliefs of WCF? RQ3.How do teachers interpret their practices of WCF, and what are the factors that lead to any modifications in their implementation process?

Table 2
Coding schemes for WCF strategy and WCF focus.

Table 3
Examples of temporary code labels for teachers' interpretations.questions about how to improve my English writing skills, but I don't think any of them was answered by my teachers' feedback."While Daniel personally did not find WCF to be effective, he considered the experiences of his classmates when evaluating its effectiveness: "But I guess maybe it was just in my case.As far as I can remember, at least some of my classmates benefited from my teachers' feedback."This insight contributed to his belief in providing WCF instead of abandoning it. many

Table 4
Teachers' practices of WCF focus.

Table 5
Teachers' practices of WCF strategy.

Table 6
Teachers' practices of WCF scope.