Developing EFL Teachers’ Beliefs and Practices in Relation to Learner Autonomy Through Online Teacher Development Workshops

The development of autonomy is acknowledged as a significant educational goal in language education, but it seems challenging in many schools. One of the reasons for this discrepancy is the lack of professional development addressing teacher beliefs and practices about learner autonomy (LA) as valid educational concerns. However, in teacher education research, limited attention has been paid to language teachers’ understandings and practices of promoting LA. Hence, this research project aims to develop in-service teachers’ beliefs, understandings, and practices about LA through teacher development workshops. First, an in-depth interview was used to gather five teachers’ beliefs about autonomy and teaching practices in secondary schools. Second, two workshops were conducted based on the literature and the initial interview to develop teachers’ LA-related beliefs, understandings, and practices. Third, a post-workshop evaluation was conducted through a focus group, the written reflection, and the researcher’s self-reflective journals to examine to what extent teachers’ knowledge and beliefs about LA have changed and developed. The findings revealed that the online workshops have made making teachers more aware of the concept of LA and its significance in language learning. Their LA-related beliefs and understandings of their teaching objectives, prior learning experiences, and teaching competence influenced their LA-related classroom practices. Some practical suggestions were provided for designing and implementing professional development workshops.


Introduction
Learner autonomy (LA) has emerged as a focal area of foreign language learning for over 40 years.It is widely acknowledged as a crucial capability of language learners and a desirable goal for successful language teaching and learning (Benson, 2007;Holec, 1981;Lin & Reinders, 2019;D. G. Little, 1991;D. Little et al., 2017).In terms of its rationale, the findings of the LA-related research (e.g., Benson, 2013;Borg & Al-Busaidi, 2012;Cotterall, 1995Cotterall, , 2000Cotterall, , 2017) ) show that fostering LA helps improve the quality of language teaching and learning and facilitates the individual life-long learning process.
Despite this generally positive perspective, promoting LA seems challenging in contexts that highlight more traditional approaches to language teaching (Benson, 2013;Littlewood, 1999).This indicates that although the development of autonomy is acknowledged as a prominent educational goal in foreign language education, there is a discrepancy between the ideal and reality.According to Borg and Alshumaimeri (2019), teachers' conceptualization of LA reflects their pedagogical attitudes and beliefs about effective classroom practice, influenced by various socio-cultural, professional and institutional forces.Likewise, inadequate teacher development has been associated with the problematic enactment of autonomy in classroom practices (Manzano Va´zquez, 2018).
Although developing LA is gaining momentum as an educational goal and various research has sought solutions to enhance LA in Asian countries (Barnard & Li, 2016), there is still comparatively little empirical study on how teachers' beliefs about LA are enacted in teaching practices.Besides, most of the research has been conducted in tertiary institutions and self-access centres that primarily rely on quantitative data such as tests and surveys (e.g., Borg & Alshumaimeri, 2019).Furthermore, there is an apparent absence of the topic in teacher development.While there are enormous professional development workshops concentrating on classroom management and classroom instruction, there are very few on enhancing teachers' beliefs and practices of LA.To fill these gaps, this study aims to explore and develop secondary school teachers' understandings, beliefs and practices about LA by drawing on the theoretical and empirical materials on LA.Thus, the study can shed light on what LA means to language teachers that could drive actual practices within classrooms.Further, this study also focuses on the effective design and implementation of LA-focused professional development workshops.Such suggestions can generate practical implications regarding what support is required to facilitate teachers' professional development to promote LA in their classrooms.

LA and Teachers' Related Beliefs
According to Holec (1981, p. 3), autonomy is described as ''the ability to take charge of one's own learning.''He mentions that being an autonomous learner means one is able ''to have and to hold the responsibility for all the decisions concerning all aspects of this learning.''D. G. Little (1991, p. 4) views autonomy through a psychological lens, conceptualizing ''autonomy'' as ''a capacity for detachment, critical reflection, decision making and independent action.''In this respect, the foci in Holec's and Little's definitions differ, with the former on the exercise of autonomy and the latter on the underlying psychological capacities that drive learners to exercise.Apart from that, Benson (2013, p. 58) defines autonomy as ''a capacity to take control over one's own learning.''In explaining the definition, he stresses that control over learning may employ various forms concerning different dimensions of the learning process.Such a process may occur in interactions with other learners and teachers in collectively negotiating learning goals and tasks, highlighting the social dimension of autonomy as ''interdependence.''Despite the common usage of Holec's definition, it overlooks two aspects.On the one hand, as D. G. Little (1991) states, autonomy can take different forms and involve multiple ways to express oneself.Holec emphasizes the independent exercise of LA while neglecting its socio-psychological nature in all foreign language education contexts.On the other, from Holec's perspective, fostering LA and developing language proficiency seem separable.However, many researchers argue that LA will likely be perceived as a desirable educational goal instead of an added-on ability (Benson & Toogood, 2002;Dickinson, 1995;Ille´s, 2012;D. G. Little, 1991;D. G. Little et al., 2017;R. Smith, 2008).In other words, the development of LA is stimulated by the social-interactive process on which effective interaction between teachers and their learners relies.More importantly, the teacher's job is to harness their pre-existing capacity for autonomous behaviour by building a desirable classroom environment where learning opportunities can be maximized.
Although academics and researchers in LA-related fields generally acknowledge the above understandings (Benson, 2013;Benson & Huang, 2016;Crabbe, 1993;D. Little, 1995;Voller, 1997), the possibility that classroom teachers have the same understanding has remained unclear.According to Phipps and Borg (2009), teachers' beliefs can shape teachers' pedagogical behaviours and, consequently, learners' learning chances.The relationship between teachers' beliefs and classroom practices seems interactive and dialectic.However, there can be a lack of congruence between teachers' beliefs and teaching practices.Many researchers suggest contextual factors constrain the two (e.g., Tapinta, 2016;Wang & Ryan, 2020;Yasmin & Sohail, 2018).Institutional factors, curriculum policies, learner characteristics, and teaching philosophies contribute to the complexity and mismatch between teachers' stated beliefs and their actual operation.
As the value of LA and the significance of incorporating it into language education become widespread, the explorations of EFL teachers' beliefs about LA have increased as well (Barnard & Li, 2016;Borg & Al-Busaidi, 2012;Borg & Alshumaimeri, 2019).Recently, Borg and Alshumaimeri (2019) investigated 359 teachers' beliefs and practices about LA in a Saudi Arabian university.All showed that most EFL teachers from different contexts believed that LA was crucial for students.Although most teachers were positively disposed to the notion of LA and involving students in the decisionmaking process, they were less optimistic regarding the practicality of this concept.In other words, there is a noticeable discrepancy between the extent of desirability teachers felt and the extent of teachers' conceptions about the feasibility of engaging students in making decisions.The noted gap highlights the complexity of relations between beliefs and practices, which, as Benson (2016) indicates, might be a productive focal point for professional development.
Teacher Development for LA D. Little (1995) proposes that for teachers to promote autonomy in their learners, they need to become autonomous and reflective researchers of their classroom practices.It highlights the key role of teacher education in raising teachers' awareness of the concept and preparing them with adequate knowledge, skills, and confidence in developing LA in their classrooms (R. Smith & Vieira, 2009).According to Benson (2013), there is a shift in teacher education from a focus on teachers as knowledge transmitters to a focus on teachers as self-directed learners and reflective practitioners.As justified further by Manzano Va´zquez (2018), active engagement in and reflection on LA-related pedagogy seems effective for developing LA.Therefore, LA-focused teacher development should enable teachers to take an evaluative stance on the various contextual factors involved in the learning and teaching process.In other words, critical reflection on the pedagogical knowledge and experience implicit in professional assumptions should be promoted.
For instance, an autonomy-oriented action research project was conducted at the University of Minho (Portugal) (Vieira, 2020).The programs featured a collaboration between teacher educators and teachers in a negotiated structure, starting with a reflection on approaches to teacher education and pedagogical knowledge and moving on to preparation, implementation and evaluation of action research projects.Although different tools were used to collect data, a common thread in these initiatives is presented: reflective, inquiry-based and collaborative teacher education is an effective approach to promote professional autonomy in seeking to develop LA.As Jime´nez Raya and Vieira (2015) argue, such a focus enables teachers to deal with the existing constraints and use the chances afforded to find spaces for manoeuvre.
To this end, reflective tools such as cases and action research should be recommended as they help teachers become reflective researchers and practitioners in making sense of their pedagogical decisions during the process of developing LA (Jime´nez Raya & Vieira, 2015, 2021;Moreira, 2009;Vieira, 2020).Taken together, there seems to be a consensus that inquiry-based teacher development integrating critical reflection into teachers' lived experiences is more likely to promote conceptual change and pedagogy transformation.

The Chinese Context
LA was first introduced into the EFL field in China as a vital element of learner-centred learning in 1985 (Y. Wang & Wang, 2016).However, research in this area was not on the rise until the beginning of this century, as many Chinese researchers suggested LA as a western notion potentially related to Chinese education (Hua, 2002;Pang, 2001).As Y. Wang and Wang (2016) put it, these works have enriched the baseline of contemporary LA in the Chinese context, through which the western notion of LA became known (e.g., Benson & Voller, 2013;Dickinson, 1987;Holec, 1981;D. G. Little, 1991).
According to Benson and Voller (2013), although LA has become well-known in its western form in foreign language education, it has a deep historical root in eastern philosophies.For instance, Xu and Zhu (2014) state that the concept of self-direction and self-discovery is affected by Confucian culture, where key elements for learning are situated in goal-setting and being reflective while learning.Building on ancient wisdom, modern Chinese educators take it a step further.For example, Ye Shengtao's statement of ''to teach is not to teach'' (Qin, 2008, p. 3) and Chen Heqin's maxim of ''learning in doing'' (Zhou & Zhang, 2007, p. 38).Although these two educators discussed LA from different perspectives, they both emphasized the key responsibility lies in teachers as guides, facilitators and supporters.
Recently, the newly issued English curriculum standards for compulsory education (2022 Edition) (Chinese Education Ministry [COE], 2022) in China have designated a shift of goals from developing students' Overall Language Ability to cultivating students' Core Competencies.The curriculum explicitly stated that one teaching aim is to develop students' correct values, essential character traits, and key skills necessary for lifelong learning and social development.In this case, fostering LA become a prerequisite for achieving these goals.Meanwhile, it also stressed the importance of refining EFL teachers' classroom practices and their underlying pedagogical beliefs and transferring power from teachers to students to boost their capability for language use and autonomous learning (COE, 2022).
To this end, the study intends to explore and develop teachers' beliefs and practices about LA.The following research questions were formulated based on the review of the literature and consideration of the context: 1. What are secondary school EFL teachers' beliefs and understandings about LA and its impact on language learning?a) How desirable and feasible do teachers think it is to foster LA in their contexts, and what challenges are they facing in fostering LA?? b) What are the teachers' understandings about their role and their actual classroom practices in relation to the promotion of LA? 2. How effective are professional development workshops in developing in-service teachers' understanding, beliefs and practices regarding the development of LA?
3. What improvements can be made to the design and delivery of these workshops?

Overall Design
The research is divided into three stages.At stage 1, preliminary interviews were conducted to elicit in-service teachers' perspectives on LA and retrospective descriptions of their LA-related classroom practices.The semistructured interview schedule in Borg and Al-Busaidi's (2012) project was modified slightly to suit specific contextual parameters.
Following the empirical data collection of the interview, stage 2 involved participants in two professional development workshops to bridge the teachers' existing beliefs and attitudes about LA promotion and facilitate LA in their teaching practices through effective classroom talk and action research.At stage 3, to answer research questions 2 and 3 about the effectiveness and future improvement of the workshops, different data collection procedures were used, including focus group discussion, participants' written feedback, and the researcher's reflective journals.

Participants
The participants selected for this study were identified through purposeful and convenience sampling (Merriam, 2009;Patton, 2015).First, the author's former co-workers and known EFL teachers were contacted via phone.Second, within the convenience sample, participant selection was also purposefully narrowed down to meet two criteria: participants should be 1) A Chinese EFL teacher with at least three years of teaching experience; and 2) A full-time classroom teacher in public secondary schools in Xinjiang province of China.This type of purposeful sampling is what Patton (2015) termed ''homogeneous sampling'', which aims to depict a specific subgroup indepth.As Mills and Gay (2019) note, qualitative inquiry focuses on depth and breadth of description and understanding, not the ability to generalize findings, and samples are not necessarily representative of a larger population.The underlying principle of participant selection thus mirrors what Patton (2015, p. 242) describes as the selection of ''information-rich cases,'' from which rich and varied insights into the phenomenon under investigation can be provided to maximize what can be learnt (Do¨rnyei, 2007).
Five teachers were selected to participate in the preliminary interview of the eight potential participants contacted.All names were replaced with pseudonyms.Besides, all teachers were invited to participate in the following workshops and focus group discussions.Of these five, one did not participate in the following workshops and the discussion, and another teacher did not participate in the focus group for personal reasons.To protect participants' anonymity, all names were replaced with pseudonyms.The Information Sheet and Consent Form were sent to all participants after receiving verbal consent.A summary of the participant profiles is provided in Table 1 below.

Data Collection
Semi-Structured Interviews.Each online face-to-face interview took 60 to 80 minutes and was conducted via Zoom or Ding Talk.The aim of the initial interviews focused on individual teachers' conceptions of LA, their LArelated teaching practice, and their professional development needs.An interview guide (Supplemental Appendix A) was prepared before the initial session.However, neither the exact wording of the questions nor the probes in the guide were not necessarily asked in the same order.The author and the participants' shared language (Mandarin) was used in all the interviews, and all the participants double-checked the transcripts and interview notes.Any mismatch was further discussed until the final agreement was reached.To triangulate the data and enhance the trustworthiness, the participants' teaching documents (e.g., teaching plans, classroom photos and students' homework materials) were collected when necessary.end of the first workshop, the participants were expected to comment on the workshop.However, the discussion ended improperly due to an abrupt network termination.

Written Feedback (WF). The initial plan did not include written feedback (Supplemental Appendix H). At the
As Barnard and Burns (2012) note, researchers need to be flexible if circumstances differ from the original plan.Therefore, the written feedback was considered the author's fallback plan.Participants were asked to give one day-delayed written feedback via WeChat (a chatting app) to answer the last question of the first workshop ''How does the input so far resonate with your experience?.''This prompt allowed participants to express their opinions, which they might feel reluctant to verbalize in the group discussion.
Focus Group Discussion (FG) and the Researcher's Positioning.According to Ho (2006), focus group participants are interviewed in groups on the principle that the synergistic effect of group interaction can generate more insightful responses than individual interviews.Hence, a focus group discussion was conducted one week after the second workshop (Supplemental Appendix F).Furthermore, ''dialogic reflection'' (Mann & Walsh, 2013, p. 297) was encouraged in the focus group, in which participants took account of the perspectives of their peers and maintained a continuous questioning and critical attitude to their practice.
During the discussion, the author acted as a moderator and facilitator.The following measures were taken to reduce the potential negative impacts on the participants.First, the goal of the discussion was made explicit at the beginning stage (to evaluate the effectiveness of the workshops and discuss potential adjustments), stressing that critical thinking was welcomed and there were no right or wrong answers.Second, to prevent some participants from dominating the discourse, the author consciously ensured that the discussion rule was not infringed upon and the conversation did not stray unduly from the parameters.Last, since two out of four participant teachers were working colleagues, a ''rule of engagement'' was made explicit to all participants through initial contact that all participants should respect the confidentiality of each other by not disclosing the identity of specific individuals when linking to their experiences to contribute to the discussion.

The Researcher's Self-Reflective Journals (SR)
To supplement the focus group, the researcher wrote reflective journals (Supplemental Appendix G) to evaluate the impact of the workshops and reflect on the process of being a teacher educator.Borg (2006, p. 249) mentions that reflective writing can be used for teachers to ''express their thoughts, beliefs and attitudes typically in relation to a particular experience.''The reflection can be categorized as a ''retrospective method'' (Borg, 2006, p. 263) to make explicit the thinking of the teacher educator (i.e., the author).Apart from that, despite its potential limitation as a subjective data collection, it can facilitate data triangulation when combining multiple data resources to bear on a single point (Marshall & Rossman, 2011;Patton, 2015).Thus, linking a reflective journal with a focus group and the additional written feedback will likely provide various insights for the workshop evaluation.

Data Analysis
A qualitative thematic analysis (Braun & Clarke, 2006) was employed to analyze three main sources of the data collection: (1) in-depth interviews with participants (Supplemental Appendices B and C); (2) workshop documents including online videos, training materials (Supplemental Appendices D and E) and participants' written feedback (Supplemental Appendix H); (3) focus group discussion (Supplemental Appendix F) and the researcher's self-reflective journals (Supplemental Appendix G).
Initially, the first interview data was fully transcribed to identify key themes.The analysis adopted a deductive process.The transcript was color-coded based on the topics in the interview schedule (Supplemental Appendix B).However, I found it easier to work with my written interview notes than the transcription for two reasons: 1) the interviewees sometimes made random comments relevant to two research questions.2) In my interview notes, I could identify similarities and differences among comments through concise note-taking.My notes could work as an organizer to make these comments relevant to the research questions.Hence, the rest of the interview was not transcribed and was analyzed through ''tape analysis'' (Do¨rnyei, 2007).If a critical comment was identified in the notes, the relevant recordings were listened to and analyzed repeatedly to ensure that my interpretation was reliable regarding the actual utterances and how they were uttered.After that, the initial interview notes, interpretations and transcripts were analyzed and reorganized inductively into thematic categories that reflected similarities and differences among them (Table 2) (Supplemental Appendix C).For example, the first theme was labelled as ''participants' understanding of LA,'' which included six subthemes, such as ''students have clear goals and take initiative to learn'' and ''selfcontrol and self-discipline.''It could be noted the tentative categories were separately sent to two of the four participants (who have prior research experience) to evaluate the appropriateness of the coding and categorization of the data.The consensus was further reached through multiple discussions.
A similar analytical approach was adopted in the focus group, written feedback and reflective journals to answer the second and the third research questions.The data was transcribed with line numbers and read multiple times with an analytic memo aside.By colour-coding the common themes and patterns with the same colour across the whole data, it was easier for me to assign small chunks of data to common themes.This further helped me identify broader themes through a recursive data reduction process.After identifying all initial thematic codes, a cross-analysis was conducted recursively through constant comparing and contrasting themes that emerged across these data resources.The concurrent and iterative nature of data analysis offered chances to verify tentative interpretations and to allow new understandings and unexpected themes to emerge.I also returned to the literature to make more links to the emerging themes.Such theorizing helped me to formalize and systematize my interpretations.To ensure trustworthiness, memberchecking was conducted by sending the data analysis results to the participants for potential comments and revisions.

The Design of Online Workshops
Purpose.The workshops were designed to encourage teachers' in-depth reflection and develop pedagogical skills (Wright, 2010).As Borg (2006) claims, teachers' knowledge can be enhanced through effective training, further modifying their classroom practices that may ultimately improve student learning.According to Johnson and Golombek (2020), language teacher education (LTE) should make explicit the goals and intentions when designing and enacting pedagogy.Therefore, the following learning outcomes of each workshop were clearly presented in workshop slides, trainer's notes and participant manuals (Supplemental Appendices D and E): By the end of the first workshop, participants will: 1. Gain an understanding of various perspectives of LA.
2. Clarify some common misunderstandings of LA.
3. Raise awareness of feasible strategies to promote LA in their contexts.4. Critically evaluate the effectiveness of LA-focused pedagogical practices in their contexts.
By the end of the second workshop, participants will: 1. Raise participant teachers' awareness of the significance of classroom interaction competence (CIC) in promoting LA. 2. Understand how the SETT (self-evaluation of teacher talk) framework works in helping teachers effectively evaluate and reflect on their classroom interaction.3. Understand the significance of action research (AR) in contextualizing individual classroom practices.4. Understand how to conduct AR in their own classroom contexts.

Selection of Workshop Topics
Adapted from Borg and Al-Busaidi's (2012) baseline research, the following workshop topics were guided both by the literature and the preliminary interview results: First, understanding and contextualizing LA.The literature reveals a need to raise teachers' awareness of the concept of LA.Borg and Al-Busaidi (2012) highlight that professional development should emphasize feasible strategies to foster LA, especially allowing participants to reflect on and refine their current beliefs and practices.The first workshop was therefore focused on participants' understanding of LA and its application in the particular context.Activities such as evaluating the effectiveness of LA-focused pedagogical practice in participants' contexts and discussing feasible strategies to promote LA were thus included.Apart from that, some necessary definitions of LA in the literature as well as clarification of LArelated misconceptions were also provided as ''theoretical knowledge'' for participants to link with their ''experiential knowledge'' as they reframe and reorganize their understanding and interpretation of themselves and their classroom practices (Johnson, 2009).
Second, Maximizing Learners' Learning Opportunities Through Effective Classroom Interaction.LA does not mean ''freedom from constraints'' (D.G. Little, 1991).Instead, more realistically, teachers should explore ''spaces of freedom'' in their classroom practices.According to D. G. Little et al. (2017), autonomous classrooms should focus on dialogic and co-constructive teaching and learning.Informed by Walsh and Li's (2013) statement that teachers can create ''spaces for learning'' through effective classroom talk, the SETT (self-evaluation of teacher talk) framework (Walsh, 2011) (Supplemental Appendices D and E) was included to help participants analyze their classroom interaction.To accompany the SETT framework, some interactional features (Supplemental Appendix D-2: 4.1) were added to assist participants in interpreting and evaluating their classroom talk.Besides, the transcribing system (Supplemental Appendix D-2: 4.3) was introduced for participants to research their classroom discourse.To encourage participants to reflect on their classroom utterances and evaluate the effectiveness of their interaction competence in their classroom contexts, authentic classroom data (i.e., video extracts) was used as analyzed to familiarise them with the SETT framework and stimulate their self-reflection of their classroom exchanges.
Third, Understanding Action Research in Teachers' Exploration of LA in Their Contexts.Since Nunan (1990) indicates that one obstacle for teachers to become researchers is their lack of appropriate training in collecting and interpreting classroom data.A set of procedures for conducting AR was introduced by Kemmis et al. (2014).Additionally, its spiral process (Burns, 2010) was incorporated to highlight its' inquiry-based and problem-oriented nature (Nunan, 1990;Wallace, 1998).Also, a classic AR process was exemplified collectively to facilitate a participantcentred approach to research, including generating research interests and questions, evaluating questions, collecting data, justifying instruments, and analyzing data.Additionally, as collaborative AR is recommended by many researchers (Burns, 1999(Burns, , 2010;;Edwards & Burns, 2016;McDonough, 2006;Q. Wang & Zhang, 2014) , I finalized this part with an open discussion of the potential challenges and feasible strategies to conduct collaborative AR in their teaching contexts.
Selection of Instructional Practices.Drawing from the literature on second language teacher education (Burns & Richards, 2009;Richards, 2008), the workshop tasks and activities were mainly underpinned by the following three design principles: 1. Heighten teachers' self-awareness of past experiences, current beliefs, practice, and knowledge (Borg, 2009).2. Facilitate interactive negotiation with fellow teachers, addressing their own practice, beliefs, and the social constraints affecting work (Borg, 2009;Johnson, 2009).3. Promote critical and collaborative reflection (Farrell, 2015;Vieira, 2020).
First, Borg (2009) states that neglecting teachers' prior cognition might impede their capability to internalize new input.The inductive approach was thus adopted to make participants' previous LA-related beliefs and experiences explicit and to link their past experiences with the new input.In addition, a case-based approach was used to enrich teachers' understanding of key issues and critical questions by representing them in their full complexity (Jime´nez Raya & Vieira, 2018).
Second, to encourage participant involvement, a dialogic approach was adopted to enhance the co-construction of ideas within a community of practice (Wenger, 1998).In particular, participants were constantly paired with different partners to enhance group dynamics and mutual trustful rapport.Additionally, online interactive tools such as Poll Everywhere and Word Cloud were used to add fluidity and dynamics to the communication.As Ellis (1990) points out, such activities can help participants examine their beliefs about LA and develop their conscious understanding of the principles underlying EFL teaching and the practical strategies teachers can adopt in their contexts.
Third, as facilitating in-depth reflection was one of the purposes of the workshops, it is more likely to be sustained when conducted collaboratively with other teachers and guided by facilitators who can structure the learning process (Johnson, 2009).Therefore, most reflective activities were conducted in pairs or within the whole group under the author's conscious guidance.Additionally, open-ended discussion activities were included at the end of each workshop to assist participants in reflecting on their practices.For instance, questions and prompts such as ''How does the input so far resonate with your experience?''and ''What steps towards developing LA would you like to take?'' were provided to help develop their reflection on their classroom practices.

Results
This section reports the research findings in relation to the three research questions.A detailed and descriptive account of teachers' views of the desirability and feasibility of promoting LA and their reported roles and practices in promoting LA (RQ1) are first presented.After conducting professional development workshops, the evaluation and future improvement are presented and discussed (RQ2 and RQ3).
What Are Secondary School EFL Teachers' Beliefs and Understandings About LA and Its Impact on Language Learning?
Teachers' Conceptions of LA.The first recurrent sub-theme was active learning.All five participants related autonomy to active learning and acknowledged the importance of motivation and personal willingness.They emphasized acknowledging the students' interests and encouraged their autonomy by designing tasks with strong personal relevance to students' lives.For instance, Zoe compared autonomous learners to crying babies, as there is a Chinese saying: a crying baby can get breast milk.This strongly highlighted a psychological orientation.Similarly, Susan defined that LA comes from learners' intrinsic motivation.The statement conception indicated a close interrelationship between autonomy and motivation (Spratt et al., 2002;Ushioda, 2011).
The second sub-theme was related to self-controlling various aspects of learning, including setting goals, monitoring the learning process and assessing the learning outcomes.This idea was in line with Holec's (1981, p. 3) description that ''autonomous learners are able to have and to hold the responsibility for all decisions concerning all aspects of learning.''Additionally, the ''learning to learn'' orientation was also supported as most participants believed that acquiring effective learning strategies can help them self-control their language learning process.For instance, Susan and Queena mentioned the importance of acquiring sound learning strategies in different learning stages.Alex also exemplified some timemanagement skills that autonomous learners in her classroom already had.An assumption can be made that teachers might view LA as a set of skills or abilities that learners need to acquire to be independent.
The third sub-theme featuring an individualistic view of LA was detected in the expression ''du zi xue xi'' (literal translation: learn independently), which was employed repeatedly by all participants when explaining their understanding of the concept of LA.For example, Susan defined LA as ''self-directed learning.''Likewise, Zoe highlighted that LA means ''learners are independent in making decisions without being affected by others.'' The fourth sub-theme from the interview was the teachers' roles in fostering LA.All participants agreed that fostering LA requires teachers to act as a facilitator and a guide, instructing students to study independently and creating the conditions for possible cooperation between teachers and students and among students.However, further analysis revealed that their teaching beliefs were inconsistent with these responses.For instance, all teachers except Helen mentioned that teachers should reduce their talk time and encourage students to express themselves more in the target language.But they all agreed that teachers should lead and control the class.The potential inconsistency might be due to their negative experiences or a lack of confidence regarding involving students in decision-making (Borg, 2006).Therefore, the fact that the teachers showed an understanding of LA that partially matches the academic definitions does not mean that they find it contextually suitable.
Teachers' Attitudes Toward LA.Participants' attitudes toward LA were classified into two major sub-themes: desirability and feasibility of LA and perceived challenges in fostering LA.The striking feedback noted in the interviews was that all teachers found it quite desirable to involve students in making various learning decisions in their classroom practices.Many stated benefits were highlighted by teachers, including acquiring extra learning resources (Helen), improving learning efficacy (Helen), facilitating goal setting (Susan), building group dynamics (Susan), enhancing independent thinking (Zoe), and so on.However, regarding the feasibility of student involvement in making decisions, teachers seemed tentative and cautious.For instance, Alex expressed her uncertainty in engaging students in making decisions because ''not all learners are able to provide constructive ideas.''She added that it was impossible to hear every student's voice in decision-making since the learners' abilities varied greatly in big classes like hers.Therefore, she would consider middle-level students' needs in making teaching plans, deciding on teaching methods, and choosing tasks and activities.It indicated that although they believed that their students' ability and freedom in decision-making was an important attribute to LA, teachers showed doubt and uncertainty in bringing what they believed into their teaching practices.Likewise, such views were also prevalent among other teachers as they showed a less positive attitude toward the feasibility of student engagement in decision-making in their classrooms.
The findings also identified various challenges the teachers encountered when fostering LA in their contexts.First, at the macro level, all the participants expressed that the prescriptive curriculum with examination-oriented context causes a negative washback effect in both the teaching and learning processes: students learn for the sake of achieving high grades and teachers teach under the pressure of the appraisal system.For example, Susan expressed, ''It's kind of like industrial mass production'' when asked to elaborate on her working environment.
At the meso level, the schools and administrators did not understand LA or LA-related policies.Therefore, all the participants expressed that promoting LA was not prioritized.They felt frustrated being forced to shoulder extra work outside of office hours, which meant finding additional time for learning LA-related principles and practices was challenging.For instance, some teachers (e.g., Zoe and Susan) reported a need to scaffold goal-setting, self-evaluation and reflection of the learning process before engaging students in decision-making activities.
At the micro classroom level, factors related to teachers and students that challenged the development of LA were discussed.First, one major hindrance reported was students' learning pressure.For example, Helen explained many senior students spent twelve hours a day at school studying for the exam, and they felt demotivated, thus preferring to adopt a ''safe'' approach to completing what was assigned.This situation also concerned their learning habits established from their prior experience.Most participants acknowledged that their students were used to being spoon-fed, indicating that students were influenced by the concept of a teacher being an authoritative person who requires obedience and respect.
Furthermore, students' low language proficiency further complicated the implementation of LA.Alex viewed autonomy as something associated with a higher level of language proficiency.She stated that low English proficiency might demotivate student engagement in language learning and further prevent them from becoming autonomous.Teachers had a similar experience regarding their long-held teaching beliefs and relatively low professional competence.For instance, Queena said, ''Teachers are used to taking charge of everything.''The rest of the participants echoed such an authoritative mindset without the awareness and proficiency to foster LA.
Teachers' Reported Practices.Although they claimed that fostering LA was not the key focus in their classroom practices, teachers who felt they did promote LA were asked to give examples of their strategies.The analysis of their teaching experience suggested two broad LArelated sub-themes: relatedness to students' personal interests and continuous communication and negotiation with students.Participants mentioned introducing interesting materials, such as English movies, stories, and songs (Helen, Queena), or sharing personal learning stories through small talk inside and outside classrooms (Helen, Susan).From their perspectives, making classroom activities and materials relevant to students' personal lives could improve their learning engagement and autonomy.Participants also mentioned the importance of continuous communication and negotiation with students about various aspects of learning, such as goal setting (Zoe), effective learning strategies (Alex)and so on.For instance, Alex particularly mentioned using logbooks among her senior students to help them keep track of their learning progress.These descriptions echo D. G. Little's (1991) statement that the development of LA is likely to be non-linear and incremental, that can hardly be achieved through single andshort-term manoeuvres.
Participants also admitted that they lacked confidence and effective training in coping with the complexity of implementing these practices.Although all participants claimed that their teaching practices were LA-related, the extent to which they reported diminishing control of students in these practices remained unclear.Furthermore, their doubt about the necessity to foster LA in their contexts was also identified.Overall, some stereotyping was detected among these teachers who believed LA was hard to promote in the Chinese secondary school context.
Teachers' Opinions Toward Professional Development.The last theme explored during the semi-structured interviews was teachers' opinions about professional development.As for the current teacher development program, almost all teachers felt disappointed about the effectiveness of the input in developing teachers' beliefs and practices.For instance, Alex said the big class size (over 100 teachers) and trainer-centered teaching mode resulted in teacher burnout and resistance.Furthermore, both Helen and Queena felt ''disconnected'' from the knowledge delivered in the program, a point articulated by other teachers as well.When asked to elaborate, Alex expressed her concerns: It's too theoretical.Even though they invited some teachers to give a presentation.Most of them are from top schools in the district.Students are not the same.(Alex, Interview) Queena showed the same worry, stressing that most training content was theoretical and couldn't be related to her teaching context.Other issues are a lack of systematic teaching method, inconsistent formative assessment (Queena), inadequate peer cooperation among trainee teachers (Alex, Zoe), a lack of systematic training for practitioner research (Helen, Queena), and a lack of ongoing reflection about their teaching beliefs and practices (Alex).
Regarding their preference for ideal professional development, some aspects, such as integrative activities, in-depth reflection, personal relevance, and systematic training on practitioner research, need to be considered.The responses indicated that all teachers regarded professional development as a prerequisite for enacting LA in their teaching practices.
To sum up, the findings of the interview mostly aligned with those prevalent in the existing LA literature (e.g., Barnard & Li, 2016;Borg & Al-Busaidi, 2012;Borg & Alshumaimeri, 2019;Lin & Reinders, 2019) in which teachers theoretically viewed LA as a desirable quality for their learners but holding a negative view on its feasibility in their practices.Although teachers did adopt some strategies to promote LA, on the whole, they generally lacked a profound understanding of the concept.They needed professional teacher development training with a focus on the promotion of LA.
How Effective Are the Professional Development Workshops in Developing in-Service Teachers' Understanding, Beliefs, and Practices Regarding the Development of LA? Theme 1: Training Outcome.First, in terms of participants' understanding of LA, all participant teachers agreed that they had a clearer understanding of the notion of LA.Helen explicitly stated that she was ''inspired to reflect on her teaching process and would pay more attention to develop LA'' (FG: 192), and Queena made a similar statement: Although our context is totally exam-driven, it is still positive to involve more opportunities for students to make choices during tutoring or class.(FG: 220) This is further supported in their written feedback (Supplemental Appendix H), as they all explicitly linked LA with individual differences (WF: 2, 15) and learning motivation (WF: 21), emphasizing teachers' crucial roles in guiding learners to boost their learning interest and self-confidence in language learning.However, a certain degree of uncertainty toward the definition of LA and its feasibility in their contexts still existed.For instance, when discussing the feasibility of fostering LA, a paradox was identified in Helen's comment that although she changed from viewing LA as an ''unrealistic'' and ''impossible'' dream to regarding it as a ''realistic' ' and ''doable'' matter (FG: 209), a negative attitude emerged later as she indicated that fostering LA might prevent teachers from helping learners get higher scores.There seems to be an interpretation that developing English proficiency and becoming autonomous are separate (sometimes contradictory) processes.According to D. G. Little et al. (2017), this opinion may mislead teachers to view LA as something ''added'' to learners, resulting in them being resistant and passive toward fostering LA in their teaching practices.
Hence, even though all teachers claimed that they had gained a better understanding of the notion of LA and its practical implementation in their classrooms, more space might be needed for teachers to reflect on practices they had already experienced and make their beliefs and implementation more explicit.
Second, concerning participants' classroom interaction competence, all participants revealed that they became more aware of the effect of their classroom talk.Zoe stressed that CIC helped her to ''make sense of the classroom practice'' (FG: 178).Likewise, Queena shared the same idea and viewed CIC as a reflective tool for teachers to self-evaluate the effectiveness and appropriateness of their classroom instruction (FG: 245).In this sense, the workshop provided them with opportunities to enhance their understanding of effective classroom talk and, to some extent, reinforce what they had already implemented in their classroom.For instance, Helen came to realize that there was a ''theoretical base'' for what she did and labelled the learning process as ''conceptualizing'' (FG: 186).
As for participants' competence to analyze classroom discourse through SETT framework (i.e., four classroom modes), the training outcome was partially accomplished.For example, although Helen stated that categorizing classroom talk into four modes helped them evaluate their classroom interactions (FG: 178), participants were not very clear about the meaning of different modes (SR: 32).They further pointed out that the improper layout and order of the activities seemed to contribute to this ambiguity (FG: 78; SR: 33).Hence, despite the perceived usefulness of the mode analysis (SR: 83), participants' unfamiliarity of the discourse analysis skill was also detected.This implies that a reasonable amount of time should be allocated to address this topic.
Third, as for participants' understanding of AR, all participants recognized that AR was a useful tool in conducting practitioner research.For instance, when discussing the benefits of conducting collaborative AR, the participants commonly considered many attributes, such as sharing ideas about problems, saving time for undertaking research activities, reducing workload, developing critical reflection, and having chances to learn from each other (SF: 93).In terms of the perceived challenges, some group variability factors were discussed, including different interests and research backgrounds as well as different disciplinary knowledge among group members.Participants also claimed that disagreement and misunderstanding might occur when jointly conducting classroom research (SE: 94).To cope with the challenges, the participants proposed that the research team should be built based on common interests and shared objectives and responsibilities (SE: 103).Additionally, regular conferences should be held to keep track of the research activities undertaken by each member (SF: 104).
Notably, although persistent obstacles were perceived concerning the implementation of LA-oriented activities, all the workshop participants stated that they would pay more attention to the development of LA, including classroom interaction and action research.In this sense, their ideas on the above aspects have been either reaffirmed or, at least, re-introduced.
Theme 2: Online Interaction and Case Discussion.When participants remarked on their favourite activity in the workshops, they all commented positively on the interactive tools employed in the two workshops, including breakout room discussion, forming Word Cloud, and using Poll Everywhere.For instance, Queena claimed that sharing ideas with different partners in breakout rooms was impressive, stressing the open-mindedness and creativity that the discussion could breed (FG: 8).This view was also supported by Helen, adding that the sharing of ideas during the workshops was particularly productive (FG: 32).
Additionally, to meet the participants' requirement that any LA-focused practice be closely linked with classroom reality (SR: 65), participants also discussed several cases of pedagogy for autonomy and the teacher educator.This arrangement aligns with Jime´nez Raya and Vieira's (2015) comment that a case-based approach intends to enrich teachers' analyses of critical issues by disclosing constraints to autonomy and opening up spaces for manoeuvre.This was also supported by Zoe and Helen in the FG, as they highlighted the effectiveness of cases in linking their prior knowledge with experience (FG: 77, 400).
However, some problems related to the online discussion occurred.First, some discussion questions were ineffective; some were skipped, and some were repetitive (SR: 124).In the focus group, Queena noted that the discussion activities in task 5 (True and False) in the first workshop were ''a bit time-consuming'' (FG: 62).Similarly, Zoe expressed the same concern about the final part of the second workshop, stating that the final discussion of AR implementation ended ''a little in a rush' ' (FG: 371).Notably, all teachers pointed out a lack of interactive opportunities in the second workshop and agreed that it was due to the lack of time (FG: 321,327).
Theme 3: Organization and Cohesion.All the participants recognized the internal cohesion between the two workshops.As noted in the FG, a few references existed to ''link'' with participants' former experiences (FG: 312,355).The design corresponded with the interview's findings that the workshops' content should be practical and related to classroom reality.It was also evident in the SR journal as the researcher used real classroom videos and photos to add authenticity to training materials to boost participant engagement (SR: 80).
However, despite the design of the workshops seemed to help participants make sense of the theory, Zoe stated that there were some issues related to the layout of the tasks (FG: 83).This opinion was also shared by Helen (FG: 102), claiming that participants misunderstood the matching task in Activity 3.3 (Workshop 2) and suggesting that the instructions and the design be improved.To enhance cohesion, Activity 3.1 should be put after Activity 3.3 (Workshop 2), supported by all participants (FG: 140-142).Furthermore, there seemed to be a poor balance between theory and practical application in the second workshop regarding classroom interaction and action research.Participants complained about excessive teacher talk delivering these two themes (FG: 132).
What Improvements Can Be Made to the Design and Delivery of These Workshops?
Participants had two major suggestions for the workshops regarding case illustration and the discussion format.First, the case-based approach should be illustrated with problem-focused narratives to promote teachers' continuous professional learning and development (Jime´nez Raya & Vieira, 2015).Through writing and sharing their LA-related teaching stories, teachers can better investigate and modify their deeply-held beliefs and scrutinize their practices, facilitating more humanistic and democratic actions in the future.
Second, regarding the discussion format, clear and concise questions and feedback should be designed to ensure the natural flow of communication.For instance, Queena stated that cutting down repetitive questions could help make discussions more ''targeted'' on the topic (FG: 71).When reviewing how to do action research, Zoe suggested multiple coding examples with different qualities be used for further comparison.
Furthermore, both Zoe and Helen recognized the usefulness of practical examples in supporting them to make sense of abstract terms (FG: 26, 47-50).Apart from that, participants should be given more opportunities to practice the AR process collaboratively.Therefore, the teacher educator should fully explore and use online interactive tools to boost participant involvement.

Discussion and Implications
The present study aims to investigate and develop language teachers' beliefs and practices in fostering LA in the context of Chinese curriculum reform.The first research question focused on what LA means to the participant teachers.The findings mostly aligned with those prevalent in the existing LA literature (e.g., Barnard & Li, 2016;Borg & Al-Busaidi, 2012;Borg & Alshumaimeri, 2019;Lin & Reinders, 2019).Most teachers theoretically viewed LA as a desirable quality for EFL learners but held a negative view of its feasibility in their contexts.All the teachers showed a consistent concern for independence and freedom.The former was considered students' capability and willingness to be selfregulated , and the latter was linked to how students could be offered chances in the decision-making process.
As for students' capacity to be self-directed, although teachers admitted their important role in cultivating students with adequate learning strategies and skills, they highlighted that LA is more than providing learning techniques.Viewing the promotion of LA as giving a first aid kit to students runs the risk of seeing teaching merely as knowledge transmission (Voller, 1997).Rather than merely being resource providers, teachers need to involve learners in negotiation with each other and the context within which explicit strategy instruction can be presented.Regarding students' willingness to be selfdirected, teachers argued that the key lay in teachers to build a ''safe'' classroom that highlights the acceptance of diversity and a redefinition of errors as potential learning opportunities.In line with this, Ushioda (2011) mentions that informative and motivational instructions should be given to students to enhance their self-determination, boosting their capacity to take responsibility for their learning.This emphasizes the need to foster the socio-psychological attributes associated with LA or D. G. Littleat's (1991) concept of interdependence.Such considerations indicated that collaborating with learners in a non-threatening circumstance should be favoured to facilitate autonomous language learning (Cotterall, 2017).
With regard to students' spaces for decision-making, the finding illustrated that most teachers showed less positive about students taking full control of the learning process.As LA is highly contextually sensitive (Benson, 2013), the opinion might be linked to the role teachers were expected to play in their classroom contexts.This echoes Littlewood's (1999) research that social hierarchy in the traditional EFL classrooms in Asian countries might be one of the hindrances in promoting LA.However, as D. G. Little (1991) argues that freedom and control are not mutually exclusive, this justifies the legitimacy of conditional freedom in foreign language teaching and learning.For this reason, teachers need to work as mediators between students' autonomous learning and educational constraints.However, the findings suggested that teachers lacked this kind of understanding and commitment to the principles of LA, which might be the decisive factor in their inadequate engagement in pedagogies for autonomy.As R. C. Smith (2000) uses the term ''teacher-learner autonomy'' to compare teachers as learners who are also in the process of learning to be autonomous to foster LA, it hence implies that professional development that focuses on teachers' agency to go against the broader systems and to create spaces for students to take charge of their learning can be a feasible target.
The second and third research questions focused on LA-related teachers' professional development.Overall, the evaluation revealed that the workshops, to some extent, mediated between research and practice as they facilitated participants' critical reflection on their LArelated teaching practices and their teaching contexts.Participants were willing to incorporate LA-focused professional development workshops into teacher professional development.Despite participants' proposals targeting the workshops' time limit, organization and cohesion, and the online discussion format, the workshops generally did what they were meant to achieve: they encouraged participants to explore their beliefs and understandings about LA and make sense of LA-related teaching practices.With the proposed improvements, the workshops can potentially help in-service teachers develop their LA-focused beliefs and practices.
Therefore, some implications regarding teacher professional development toward autonomy are proposed.First, as Crabbe (1993) argues, the significant factor in promoting LA depends on the pedagogical dialogue.Therefore, professional development should provide teachers with pedagogic discourse knowledge and strategies to explore their interactional features that either obstruct or optimize students' learning potential.In this sense, teachers should involve students in a dynamic negotiation process by which learners gradually tend to shoulder more responsibility for their learning.One of the ways negotiation can be enriched is through careful scaffolding and monitoring unfolding exchanges as it is being fabricated to stimulate negotiation of meaning and learner engagement (Walsh, 2011).
Second, as the findings suggested, teachers' pedagogical knowledge regarding the implications of fostering LA needs to be further enhanced.This implies a role of AR that involves teachers inquiring about their practices through self-monitoring.According to Burns (2010), the introduction of AR in professional development will likely motivate teachers to conduct reflection through inquiry-based research in their teaching practices.Particularly, Queena claimed that in-service training should help teachers identify various practical issues and equip them with sufficient tools to help them make informed pedagogical decisions.Therefore, AR can be seen as a Sine qua non for teachers to revisit their pedagogical assumptions about LA and eventually operationalize it more systematically in their classrooms.
However, there is a tendency for teachers to fall into an inflexible and confined AR cycle (i.e., planning, acting, observing and reflection), which might merely promote Mackenzie's (2018) concept of display and formulaic reflection that isolates their practical knowledge from the language teaching discourse community.Hence, opportunities to reflect on reflections can be encouraged in teacher professional development (Akbari, 2007) .
Last but not least, the findings suggested that dual foci on theory and practice in teacher professional development should be well managed.Thus, it is recommended that teacher educators create a problem-solving environment where pedagogical theories illustrated with authentic cases should be incorporated to give teachers greater opportunities to reflect on and amend their current beliefs and practices.

Conclusion
This research illustrates the process of developing and evaluating the professional development workshops about developing in-service EFL teachers' beliefs and practices about LA.In terms of the findings of the workshops, it revealed that the evaluation provided valuable insights into the process of teacher learning, with opportunities for participants to contextualize the theory in the workshop activities.Theoretically, the study contributes to the existing literature on language teachers' beliefs and reported practices regarding LA.As a teacher professional development-based inquiry, the study offers practical pedagogical and research implications on developing teacher beliefs and practices about LA incrementally and systematically.However, the study is not without limitations.First, the workshops were conducted at the end of the summer term when the participants had no time to validate their claims and could not immediately implement refined LA-focused pedagogy.It would be better to carry out workshops lasting a longer period with real classroom observation.Besides, longitudinal research might be needed to explore the long-term development of teachers' LA-related beliefs and practices.Furthermore, research tools such as narrative frames, think-aloud protocol, and focus group can be adopted to enhance the understanding of the long-term impact of professional development workshops on teacher learning.Lastly, considering the limited number of participants involved in the study, future research can thus be conducted with a wider sample in relatable English language teaching contexts elsewhere in China.

Table 2 .
Teachers' Conceptions and Reported Practices.