Embedding reflective practice in a feedback focused assessment design in a Master of TESOL program

This study reports findings about an assessment design, with embedded reflective practices tied to feedback, and its capacity to enable conceptual change for professional learning in a Master of TESOL in Australia. The assessment design was initially developed to respond to the lack of supervised field experience. The majority of students in the course were future teachers of EFL in Asia. The study used Gunstone and Northfield’s (1994/2007) phases in conceptual change in combination with Farrell’s (2018) stages of reflection and assessment as its theoretical framing. Data gathered from past and current students included interviews, unsolicited student emails and posts, and extracts from a final assessment task. Findings suggest that the embedding of reflective practices develops reflection for professional learning beyond the course. However, the assessment tasks need to be designed with a specific set of design features for real conceptual change to occur. They need to: 1) be centrally focused on interaction; 2) recognize the need for students to gain from the diverse experiences and backgrounds of peers through collaborative, real world assessment tasks that also develop their professional English language; 3) explicitly model and make reflection a criterion for assessment; 4) permit professional and personal


Introduction
Challenges are often presented to teaching staff in Higher Education (HE) that are not of their making. One such challenge was experienced in the Master of TESOL (MATESOL) course at an Australian university. The challenge was how to provide an actual professional experience to students in a program that does not qualify them to teach in Australian schools.
The course can prepare students professionally to teach English as a Foreign Language (EFL) or English to adults overseas or in Australia with additional requirements. It also provides a very small number of locally qualified students the opportunity for professional upskilling in English as an Additional Language (EAL). The course therefore serves disparate needs.
Two main factors contributed to creating this challenge: 1) a course review that resulted in blended teaching being introduced in 2014 where classes were evenly taught in face-to-face and online modes; and 2) an exponential increase in international student enrolments (from 50 in 2015 to 183 students in 2019 with a subsequent drop to 126 in 2020 owing to .
At the same time, discussions about employability and assessment were also taking place.
These issues created a need to review the assessment tasks in the professional practice unit (one of four units in the MATESOL at that time) as they were no longer practical for such large cohorts. The review led to the creation of an assessment design based on Winter's (2003) patchwork model. It aimed to develop students' feedback literacy and support students' reflective practices. At the same time, it provided a limited practical experience.
The purpose of this study is to investigate how pre-service teachers (enrolled in the unit at the time of the study) and former students (who course-completed in the previous 12 months), perceived the impact of practice-based assessment tasks on their understanding of reflective and feedback practices.

Assessment in Higher Education
Historically, assessment in the tertiary sector has been largely designed for summative assessment purposes because of the need to provide accreditation and certification to meet the needs of employers and employment authorities. However, increasingly the changing make-up of societies, globalization and mobility have meant that universities need to take into consideration a wider range of assessment purposes. This is important in order to prepare students with a range of skills for life-long learning (Burgh et al., 2005) that does not simply end with the award of a degree, certificate or diploma. This shift in emphasis has led to the reconceptualization of assessment as a process of skills development and reflection achieved through tasks that are transformative, have real-life application and establish community partnerships (Boud & Molloy, 2013;Hounsell, 2007). At the heart of such tasks lie the principles of reflexivity and interaction through engagement with content, theory and practice, and with a range of people relevant to the discipline. Students are thus provided with richer professional opportunities that can allow them to develop reflexivity and the skills needed for working in increasingly diverse local and transnational communities.
The shift in emphasis in assessment is thus one of calibration in purpose as well as in the establishment of a greater connection between summative and formative assessment so that they are no longer separate from each other. In this process, university staff are encouraged to design assessment tasks that give greater prominence to Assessment as Learning (AaL) and Assessment for Learning (AfL) (formative assessment) alongside Assessment of Learning (AoL) (summative assessment). The greater prominence given to Aal and AfL also provides the potential for students to develop feedback literacy by appreciating feedback, making judgements, managing affective dimensions, acting on them, and creating conditions to enable student uptake of feedback (Boud & Dawson, 2021;Boud & Molloy, 2013).
Pivotal to formative assessment is sustainable feedback (Hounsell, 2007) which involves dialogic student discussions about learning. Sustainable feedback marks a shift away from viewing feedback as being transmitted by tutors, and recognizes and facilitates ways of enabling feedback for self-monitoring. It enhances student metacognitive practices that are useful for lifelong learning. It also involves assessment designs that are varied, take into consideration different modes and sources of feedback (Henderson et al., 2019), and that provide staged feedback to enhance performance (Hounsell, 2007).

Reflective practice and metacognition
There is agreement about the importance of reflective practice in professional learning because it has the potential to change teacher-held beliefs and values that can lead to improvements in their practice. However, there is often a lack of conceptual clarity surrounding the theoretical understandings that underpin how it is understood (Akbari et al., 2010;Farrell, 2018). This makes it particularly difficult to operationalize (van Beveren et al., 2018). As well as a need for clearer definitions of reflective practice, it is also necessary to understand where alignments exist between teacher cognition, metacognition and reflective practice, and the dialogic processes that lead to action or change.
Reflective practice has a long-standing tradition in teacher education beginning with the broader philosophical orientations of Dewey, Freire and Schön. Schön's (1983) concept of the reflective practitioner has impacted teacher education through its focus on reflection-inaction (reflection on experience in the moment) and reflection-on-action (reflection on recollected experience). These have been further elaborated in Killion and Todnem's (1991) concept of reflection-for-action, a cyclic and iterative process (Mann & Walsh, 2017) that identifies what learning can be applied (and how) as a result of reflection.
In English Language Teaching (ELT), Farrell (2018) adds reflection-as-action. This brings attention to a teacher's "inner life" or the self, alongside a professional life/self. It adds a reflexive dimension which enables analysis of how what is done in teaching affects teachers, and aligns with their values. Farrell (2018) draws on these processes to propose a framework that captures teachers' ideas and beliefs about teaching, learning and themselves. It comprises philosophy (linked to identity), principles (comprising assumptions and beliefs), theory (governing teacher choices), practice (focusing on reflections before, while and post-teaching) and beyond practice. Also influential in reflective practice in ELT is the six factor instrument developed by Akbari et al. (2010). The instrument aims to capture teachers' beliefs about practice, the tools they use for reflection, their actions in addressing their professional development, how they focus on their leaners, and what they think about the broader sociopolitical context and moral issues that impact their teaching. The main focus in both these models is on thinking.
In the context of pre-service teacher education, Gunstone andNorthfield (1994/2007) also proposed a framework, one that went beyond the thinking element in reflective practice.
It aimed to capture conceptual change in pre-service science teachers' beliefs and practices.
The critical element here is metacognition. Its inclusion leads to explicit attention being paid to how pre-service teachers make decisions and act on their reflections in a reiterative cycle of reflection, evaluation and action. This explicit focus on action has been missing in ELT.
The framework consists of four phases of conceptual change: 1) recognition and reflect (being presented with a new experience that allows recognition of existing ideas and beliefs); 2) evaluate (judging what has been learned); 3) decision-making and reconstruct (deciding whether, as well as what action to take, to reconstruct ideas and beliefs); and 4) reiterate (for further recognition, evaluation or decision-making/reconstruction).
Gunstone and Northfield also underscored the importance of collegiality in teacher education by emphasizing that ideas cannot be transmitted but must be constructed through opportunities created for reflection, feedback and evaluation through dialogue with other professionals. These dialogic processes enable teacher reflection to be scaffolded (Mann & Walsh, 2017) so that conceptual change emerges through social interaction. This is a constructivist notion (Gunstone & Northfield, 1994/2007, one that sits well with the multiple modes and sources of feedback outlined by Henderson et al. (2019).
Gunstone and Northfield's phases in conceptual change are useful in exploring teachers' reflective practices. In combination with Farrell's (2018) framework and the AaL assessment principle, it will inform the theoretical framework for this study to offer a lens for exploring pre-service/teachers' reflective practice through their shifting beliefs, values, theories, practices and philosophies about pedagogies and about themselves. The unit's practice-based assessment tasks are explicitly designed to include reflection. They enable dialogue with peers, the tutor and EAL learners through team and collaborative teaching, reflective journals, peer assessment and multiple modes of feedback. These are the tools that Farrell (2019) suggests teacher educators need to embed in assignments while at the same time overtly model how they encourage reflection in their own teaching. They are also tools that build systematic pre-service teacher reflection (Jones, 2014).
Building on the previous research on reflective practice, feedback and assessment, the study will be guided by the following research questions (RQs): RQ1. How do pre-service teachers (enrolled in the unit at the time of the study) and former students (alumni), perceive the impact of practice-focused assessment tasks on their understanding of reflective and feedback practices?
RQ2. What areas do the two groups of participants highlight as being especially impactful personally and professionally?

Context for the study: The MATESOL unit
The professional practice unit is a final unit in the MATESOL course. The course has undergone a review with changes implemented from August, 2020 and into 2021 and 2022.
The three assessment tasks discussed in this paper have been retained and offered across two units with the addition of a short practicum component.
The course is planned so that students move from considerations of theory and practice in the initial units to an emphasis on practice in the final unit. The assessment tasks are designed with four principles in mind: 1) Winter's (2003) patchwork model entailing a smaller number of connected tasks that can stand alone; 2) the creation of authentic tasks that align with lifelong and professional learning (Wiggins, 1989); 3) relevance to pre-service/teachers' own contexts; and 4) embedded sustainable feedback practices using multiple modes and sources of feedback (Nguyen & Filipi, 2018) for uptake by students (Boud, 2000;Henderson et al., 2019;Hoo et al., 2021;Hounsell, 2007).
The three tasks that were designed over time (and current in 2021) comprise: Assessment 1 (AT1): Scope and sequence (S&S) chart and rationale Assessment 2 (AT2): Minilesson (part 1) and critical review of the lesson (part 2) Assessment 3 (AT3): Learning intervention project comprising a group task (part 1) and individual reflective task (part 2) In AT1, the focus is on planning. Students develop an overview (S&S chart) of a unit of work comprising six lessons which they then critically discuss. Students also provide each other with feedback about their S&S chart and reflect on how the feedback was used to improve the plan.
The first part of AT2 requires students to microteach a 12-minute segment of one of the lessons included in AT1. On completion of the micro-teaching, three minutes is provided for general discussion/feedback about the microteaching. Students are required to provide written peer feedback about the microteaching, while the tutor provides private feedback during a 10-15 minute discussion after the microteaching. Students are also required to assess their own microteaching. In part 2, students use the feedback to critically review their microteaching, and discuss what they would do differently and why against pedagogical and language learning theories.
AT3 is premised on the principles of project-based learning (Kingston, 2018) and is composed of two parts. In part 1 students work in small groups with an actual EAL learner over five weeks. Students interview the learner, and conduct a needs/linguistic analysis to identify an area for intervention/improvement. They then devise a teaching plan to assist in language improvement, and implement it. In the plan they must explain how learning will be monitored and what feedback approaches will be adopted, and how the intervention will be evaluated. In the final week of the semester, students orally present the completed project to peers.
Throughout, students develop a group document using Google Docs, shared with each other and the tutor, intended to facilitate peer and tutor feedback in weekly consultations. These processes provide formative assessment through feedback that occurs through oral and written discussion so that the students are supported in their project development, and in their understanding of what it means to act on feedback.
The individual component (part 2) requires students to complete an evaluation of the project processes. An online, optional journal is also set up through a Wiki on Moodle (the university's online platform) to facilitate formative reflections.
Through this patchwork assessment approach, the focus is on cycles of planning, teaching and reflection. The intention is to expose students to different practices of assessment through varied models, modes and sources of feedback.
Beyond course completion in 2020, students were offered optional opportunities to copresent at conferences and share their experience of co-mentoring; they were also encouraged to develop a digital professional portfolio for presentation at interviews.
Finally, an additional feature of the unit is that it has been taught in blended format from 2014 to 2019. Students have had the option of conducting these performative elements either online (a/synchronously) or face-to-face. During the COVID-19 pandemic in 2020, all assessment tasks were completed through livestreaming.

Qualitative methodology
This is a qualitative study using a case-study approach. Merriam and Tisdell (2016) use the terms particularistic, descriptive and heuristic to define case-studies. They are particularistic because the focus is on investigating a particular situation, event or phenomenon.
They are descriptive as they allow a rich description of the phenomenon or event under investigation to occur. They are heuristic because of the impacts of the investigation on the reader and/or researcher. These may include extending knowledge or re-conceptualising the phenomenon or event (Merriam, 2009).

The case under investigation
The case-study under investigation is an assessment program in a MATESOL unit in an Australian university with a predominantly international student population. Most students will return to their home countries to teach EFL on completion of their studies. The phenomenon under investigation was how pre-service teachers (PSTs) (i.e., currently enrolled) and former, recently graduated students (i.e., alumni), perceived the impact of the assessment tasks on their understanding of reflective and feedback practices, and whether there were differences between the two groups in perceived benefits.
I drew on two sets of data: interviews with alumni six to 12 months after completion who were teaching or had gone on to further study (Group A), and written documents of students in their final semester (Group B). For Group A, information describing the project and inviting participation in the study was sent to 30 past students. Six students consented to being interviewed and to having their past reflective assignments analysed. Three were local students: two (Rick, from Indonesia and Jane, from Australia) in an adult English program and the other (Iris, from Indonesia) in an Indonesian language program. A fourth (Anita) and fifth (Amy), were teaching EFL in Indonesia and Shanghai respectively. The sixth student (Lucy, from China) was continuing her studies. The interviews focused on the perceptions of alumni about the design of the three assessment tasks, their reactions to feedback practices, and whether they believed the assessment tasks prepared them for future or current teaching. For Group B, discussion posts on Moodle, unsolicited emails, and extracts from the final reflective assignment about the impact of the assessment tasks were drawn on. Their authors were invited to participate. Details are described in table 1 below. All students signed consent forms to satisfy ethical requirements including using the written documents for analysis. In the interests of maintaining the confidentiality of the participants, pseudonyms have been used throughout.

Approach to data analysis
I used an emergent analysis approach driven by the text (Stemler, 2001) in combination with an inductive (bottom-up) and deductive (top-down) analysis (Blair, 2015). The inductive analysis involved looking for themes across the data set in a comparative horizontal analysis (Merriam & Tisdell, 2016), while the deductive analysis proceeded from a priori categories using the three levels in Gunstone & Northfield (1994/2007. Qualitative studies require that procedures and results be derived through transparent and consistent processes in order to inspire confidence in the interpretations and results (Merriam & Tisdell, 2016). In this study it was important to minimize the effects of my position as researcher, as unit developer and as tutor to cloud judgement. Important here are questions of internal and external validity. The internal validity of findings was established through members' checks and the researcher's reflexivity (Merriam & Tisdell, 2016). I consulted participants to gain further background information and to establish clarity in my understanding of the interview and the written data when necessary. With respect to external validity, I sought to minimize the imposition of my beliefs in the interview process. I also took a methodologically self-conscious (Blair, 2015) stance to maintain an awareness of the subjective process in my coding. This entailed interpreting the data according to my positions, and scrutinising the frames used to understand the data. The strategies employed to enhance external validity or transferability include the MATESOL course itself and the participants. The features of the course and the participants themselves (current PSTs and alumni) generate findings to permit comparisons with other contexts with similar variables.

Applying the theoretical framework
For the theoretical framing, three of the four levels in conceptual change proposed by Gunstone andNorthfield (1994/2007) were used: 1) recognition (identified through references to time or a new experience); 2) evaluation (identified through references to thinking and transformation in understanding); and 3) decision-making and/or reconstruction (identified through references to action being taken). The fourth phase (reiteration) was not included because the study did not occur over time.
The analysis was conducted using a search for key words associated with metacognition and action for each piece of data. The 43 terms generated (in table 1 below) were then categorized into recognition, evaluation and decision-making/reconstruction independently by the author and a colleague, an applied linguist. This generated a concordance of 88%. The process for resolving discrepancies entailed going back to the texts to gain a more refined contextual understanding and then jointly deciding on the best category. The categories were then further analyzed discursively. The results of this thematic analysis and approach are presented in table 2 below. Table 2 The key words generated from the thematic analysis Of the texts used in the analysis, the interview was the only instrument that was driven by a research aim. The other data-sets were all naturally occurring as they were unsolicited or part of the assessment tasks. These were chosen because they were the result of reflection about the unit's assessments and therefore unclouded by research concerns.

Summary of findings
The initial key words associated with the three levels in Gunstone andNorthfield's (1994/2007) framework (in Table 2) permitted capturing conceptual change as participants reflected on how they moved from recognition to evaluation to decision-making. In answer to RQ1, this suggests that the framework can be adapted as a tool for evaluating assessment tasks that have reflective or self-assessment practices embedded in them. With respect to Farrell's (2018) framework, four of the five components (philosophy, principles, theory, and theory and practice) emerged as themes in the data while the fifth, beyond practice, did not.
In answer to RQ2, the discursive analysis of the pre-service teacher data (Group B) generated three key findings: 1) the reported value of collaborative group processes and group feedback; 2) a deeper knowledge about feedback principles and practices; and 3) the importance of interaction with a range of "others" (with different levels of expertise and different cultural backgrounds) to make sense of pedagogical and personal learning events. The receiving (for teaching). They also highlighted their shifting view of the teacher as facilitator.
All participant reflections signalled the ways in which the design of the assignments provided opportunities to learn about teaching, to learn about themselves and to better understand the theories they had been exposed to in the course. This suggests a readiness to develop their own theories linked to practice. These findings will now be discussed in more detail. The key words relating to the three levels are bolded in the participant texts and quotations.

Level 1: Being presented with a new or first experience leading to reflection
The first text by Megan is an unsolicited email sent to the tutor at the start of the project about groupwork in AT3.
I have been upset since I started this project because I think it is the first real project that I take part in. I was lost and frustrated because of a lack of experience. A member of our group has a very strong personality too (I have a strong personality as well) and as a result, we appeared to quarrel rather than discuss.
Megan is highlighting the new experience of working in a group in which she felt personally challenged. Being challenged is what Kemmis (2011) refers to as the "trigger" for the reflective cycle to begin. Megan mentions her lack of experience in working as part of a team and having to deal with a peer with a strong personality like her own. She acknowledges this as a cause of tension for the group, resulting in disagreement. She also emphasizes her own feelings of being upset and frustrated. Her reflections unveil her personal attributes and her previous learning experiences (of only having done individual assignments), as well as the dilemma of what to do to overcome the issues.
Group process was an issue raised by an interviewee, Amy, as well, who reported that there was little agreement in the group initially with each member wanting to do things individually:

We spent a lot of time on negotiation because we can't (sic) compromise with others'
opinions.
Students are not given any written instructions about working as a team beyond being asked to state the roles and responsibilities to be assumed by each team member. A specific criterion on group processes is only included in part 2, the individual reflection. This suggests the need for improving the rubric to guide students who may not be familiar with working in groups.
The second email by Alice about AT2 was sent to the tutor at the end of the semester.
The contents of this email also point to differences in cultures of learning through doing tasks that are educationally and culturally unfamiliar. Alice refers to feeling comfortable in a new environment that challenges her previous ways of learning, comments that relate to previously held principles (Farrell, 2018 Boud & Dawson, 2021;Hoo et al., 2021).
Differences in cultures of learning (see Tankard, 2015) is another recurring theme, and illustrates the potential for change in teaching practices as illustrated in the interview with Lucy where she is reflecting on the usefulness of working with a curriculum framework in planning her teaching.
We don't have the curriculum framework in China. At first I get (sic) confused, what is the framework and then I figure (it) (sic) out … and how to use it in the lesson plan so I was very inspired by that and I think it was very useful and um it is about how to validate our teaching into practice (sic). So we find the reasons for why we do that.
The time references indicate a relational before and after, useful in exploring change in attitudes, thinking and/or practices. Lucy makes reference to her initial confusion in understanding the purpose of a curriculum framework, subsequently realizing how it can be leveraged as a useful tool to guide lesson planning.
The sense of being confronted with a new experience provides a strong sense about what students have experienced through the explicit focus on reflection in the unit's assignments. Differences in the education environment in all the above texts (including artefacts (curriculum frameworks) and practices (groupwork, and personal reflection and feedback)) are thus given prominence as providing the participants with a new set of experiences to reflect on, which in turn has the potential for reshaping their beliefs (principles), theories and practices. For Megan and Amy, it is about realising the need to establish effective internal groupwork processes, outlined by Markham et al. (2003)  in Farrell's (2018) framework, and practices that they are not accustomed to (reflecting and focusing on feedback).

Level 2: Evaluation
After recognising their ideas, values and beliefs by articulating a change in awareness and understanding, the next step is for the pre-service/teachers to evaluate what they now know and how valuable that learning is for their practice. Here the most frequent key words were the verbs valuing and transforming (attitudes and thinking) that could lead to change. We return to Megan's email to track how she articulated this step.
After all these mess up (sic), it occurred to me that all the problems may come from the leaderless discussion mode. Actually, compared with the project itself, I learned more than that: how to solve problems, how to communicate and how to cooperate. All these skills cannot be learned through individual tasks… In stating that she is aware that a root cause of the initial dysfunction of the group was a failure in not choosing a leader, it is evident that Megan has evaluated the reasons for the group's dysfunction and how this could be changed. She cites solving problems, effective communication and cooperation as being key. The identification and evaluation process in understanding that teaching and working in a group involve cooperation and collaboration, also enabled Megan to act and apply new knowledge in decision-making, as we will see.
Evaluation of the identified issue that has resulted in learning is clearly a powerful learning experience given Megan's need to share her insights through personal communication.
The learning afforded by this project is also highlighted by Rhianna in the individual reflective assignment (AT3).
This is the first time that I actually realized the power of collaboration … As an international student in a multi-cultural context, most of my thoughts and understanding from the units were not shared and discussed in depth with peers due to … individual assessments. … I got the chance to discuss related knowledge in depth with peers who have different backgrounds ... Lots of ideas, understanding and thoughts were generated tasks that engage PSTs in working with and mentoring their peers is an enriching experience (Nguyen & Bauldauf, 2010). Such opportunities enable them to share perspectives from their diverse contexts, creating the potential for reflection, validation and reconstruction of prior experience, essential to life-long learning (Farrell, 2019). As well, and importantly, students whose L1 is not English are also afforded opportunities for developing their English in professional conversations with peers and mentors. As Rhianna stated:

My English has improved … I have not spoken so much in English like in this unit.
For Alice evaluation is about articulating a personal change in understanding herself and her students.

… changed the way i look into myself as a person, a teacher & the way i (sic) look into my students
Alice makes reference to the immediate impact of learning about the importance of reflection through self-assessment that forced her to look into her personal and professional self in order to understand what it means to be a reflective practitioner. Self-assessment as a practice encourages students to monitor their learning, creating the potential for students to develop their metacognitive skills (Carless et al., 2011). Notably, Alice is also looking forward (reflection-for-action, Killion & Todnem, 1991) to applying this newfound skill in her practice.
Like Alice, Gina's evaluation of the microteaching experience and feedback in a Moodle post, points to the learning that has occurred for her beyond the assignment through the opportunities for reflection and assessment in the assignment design; in particular the value of receiving feedback from different sources, highlighted as being important by Henderson et al. (2019).
From a teaching point of view this has been a very worthwhile exercise … often our own perceptions as a teacher of how a lesson may be received by students is vastly different to the reality, and the feedback allowed a lot of learning to to take place and opportunity to improve lesson delivery in the future.
Here Gina, a local practising teacher who came to the course for professional upskilling, is referring to the multiple-mode, multiple-sourced feedback (self, peer and tutor; orally, in discussion and in writing) (Filipi, 2017;Nguyen & Filipi, 2018) students receive in AT2. The experience of taking part in self-and peer-assessment affords students an opportunity to understand assessment criteria by being required to make sense of them in the very practice of application. Furthermore, tutorials that provide activities that overtly model different feedback practices, in ways suggested by Farrell (2019), provide the necessary unpacking of what feedback means. Importantly, peer-assessment takes place through interaction, providing opportunities for learning, which in Vygotskian terms, emerges through the scaffolding provided by peers and teachers (Mann & Walsh, 2017).
The importance of interaction is underscored by Rhianna in the following extract from her AT3 individual reflection where she makes practical sense of constructivism. The theoretical framing she refers to is emphasized throughout the MATESOL course. It has also been highlighted by other researchers as being important in TESOL courses for the development of self-reflection (e.g., Wyatt, 2010).
This assessment allowed me to apply and view social constructivism from the perspectives of both teacher and student at the same time. While we were designing the project on the base (sic) of social constructivism for facilitating (sic) our target learner, we were somehow improving our understanding about social constructivism. The process of teaching is not just about the delivery of target knowledge to the learner, but a process of interactive learning between teacher and student simultaneously. This new thought has transformed my understanding about teaching and learning also as teacher and learner. Creating or performing a learning process that benefits both teacher and learner at the same time will be one of the principles for my teaching/learning in the future.
Using Farrell's (2018) framework, Rhianna's reflections make visible the changes in her philosophy, principles, theory and practice. Reference is made to improved and transformed understanding about teaching and learning, and the pivotal role of interaction between teacherstudent and student-student, indicating a shift in her assumptions and beliefs. It emphasizes her changed focus of the social nature of feedback practices in ways that have been reported for process writing (e.g., Perez-Amurao, 2014). Her evaluation about being simultaneously a teacher and a learner indicates a conceptual shift in identity relevant to her philosophy. Finally, in pointing to the value of creating pedagogical opportunities in her future teaching that provide learning for both the teacher and students, we see Rhianna's reconceptualization of learning as an active process, and the teacher's work as facilitating rather than transmitting knowledge. 46 This marks a shift in her views about theory and practice.

Level 3: Decision-making and reconstruction
As a consequence of reflection, new understanding is applied to deciding what action ( decision-making (solving) where the action taken to resolve the "leaderless" problems has led to reconstruction of knowledge. The first relates to the learning opportunities provided by group work to develop collaborative skills. Megan's second point is the sense of new learning having taken place to enable decision-making (reflection-on-action) (Schön, 1983), while the third is Megan looking forward to being motivated for further learning (reflection-for-action) (Killion & Todnem, 1991). These reflections also provide an insight into what Farrell (2018Farrell ( , 2021 refers to as a teacher's "inner life" and reflection-as-action. Megan's statements about acting on what she has learned and about looking forward, align with findings that assessments need to provide a process of skills development and reflexivity through tasks that are transformative and have real-life application (Boud & Molloy, 2013;Hounsell, 2007).
In the next extract, Anita reflects on the transferability of the skills gained through the explicit focus on feedback in the assessments to her current teaching in big classes.
…looking from that (AT3) project, and the feedback from the student does matter to the teaching practice, right. So, I did (sic) the same now with my bigger class, so I asked them to write down the feedback, for example, at the end of the session, like at the end of the semester, or at the end of the year ... And then, I also asked them, "What do you think about these materials," … "what's your difficulty, or did I explain it [well]?" and then I know from their perspective that I did the right things, or whether I need to do something else to make it better.
Anita is reporting the reconstruction of her beliefs. This builds on the reported shift in her philosophy, principles, theory and practice where earlier in the interview she mentioned that learning is done in partnership with teaching (hand in hand). Anita intimates that her conceptualization and knowledge of feedback have evolved to include assessment of learners and evaluation of teaching, important to teacher development as it provides feedback on aspects of performance (Clark, 2012). This has been as a result of her experience of the formative feedback practices in the design of the group project (AT3). Her reflections also highlight her practical understanding of the differences and the interconnectedness between evaluation (about teaching) and assessment (about learning) (Ellington et al., 1993). As a feature of AT3, ongoing feedback from the EAL learner to the teaching team enabled adjustments to be made in the groups' planning and subsequent lesson, a practice now applied in Anita's bigger class.

General discussion and conclusion
This study sought to investigate if embedding reflective practices that include multiple sources of feedback in practice-based assessment tasks in a MATESOL course, can build conceptual change in PSTs' and teachers' beliefs and practices. The synergies in the combined framework based on Gunstone & Northfield (1994/2007) and Farrell (2018, and the complementary ways in which they informed each other with reference to AaL, provided a useful lens for interpreting the range of texts produced by both sets of participants. It made it possible to track how the reflective cycle was triggered (Kemmis, 2011) leading to reported change in the cycles of recognising/learning, thinking/evaluating and actioning.
Findings show that in comparing the two groups of participants, PSTs reported valuing collaborative group processes and interaction with a range of others, and their development of a deeper knowledge about feedback and assessment principles and practices. The alumni also highlighted the capacity in the assessment tasks to increase their understanding of the power of feedback. However, this was done in terms of how to apply feedback practices in giving (for learning) and receiving (for teaching) in their current practice. They also highlighted their shifting view of the teacher as facilitator.
Findings also indicate that assessments that focus specifically on interaction with others and that promote reflection can profit from a design that: 1) is sensitive to the social nature of learning by providing tasks where interaction is central; 2) recognizes the need for students to gain from the diverse experiences and backgrounds of peers through peer-feedback and peermentoring, and through collaborative, real world assessment tasks (which aligns with Boud and Dawson's (2020) finding that activities in courses need to enable feedback practices rather than block them.); 3) develops students' (English) competence in professional discourse; 4) explicitly models and makes reflection a deliberate practice and a criterion for assessment; 5) permits professional and personal application of learning beyond the course through optional opportunities such as presenting at professional conferences so that mentoring can continue in a new collegial relationship; and 6) allows students to develop online practices. The findings also suggest the importance of creating opportunities through assessment tasks for students to (co-/re-)construct their beliefs, values and concepts from reflecting on their experiences through genuine peer-mentoring as well as tutor-mentoring. Such practices align with current research on the need for assessment tasks that are practical and that link to professions but that also enable participation and self-regulation in order to develop feedback literacies (Boud & Dawson, 2021;Hoo et al., 2021;Ibarra-Sáiz et al., 2020). In teacher education, reflective practice to practise is thus fundamental.
The assessment tasks can certainly benefit from some further reworking. An issue that has emerged, is the need for explicit attention to be given to the group processes and management, which as noted by the international PSTs, was an unfamiliar experience. One way to address this is to include a more explicit, dedicated criterion about groupwork in the assessment rubric as suggested by Nelson (2011). Thought could also be given to providing more explicit skills on what effective reflection means as urged by Farrell (2019). To this end, the optional journal could be made compulsory as part of a deliberate design to enable explicit reflective practices.
As stated, the MATESOL course has been restructured. The overall student satisfaction with the unit, measured formally by the university, has been consistently high. It indicates that students consider the experiences offered by the assessment program to be a positive experience. These evaluations have been taken into consideration in deciding to retain the three tasks within two separate but related units from 2021. By adding a compulsory short practicum, further possibilities for richer reflections but also deeper possibilities for connecting the assignments in more explicit ways, may result.

Conclusions
The study suggests a need for MATESOL teacher education programs to carefully consider their assessment practices as they rise to the challenge of developing assessment tasks that are authentic and can prepare students for post course life and work. Providing opportunities to explicitly focus on reflection through a need to do so in concert with others, particularly with peers and not just the tutor, is important. The benefits as outlined above are both immediate and long-term. They provide development of dispositions (associated with metacognition) for life-long learning beyond the boundaries of a course.
With respect to limitations, the study has relied on the interpretation of preservice/teachers' own interpretations about their practices, thinking and experiences as a result of the unit's assessment tasks. It also provided a "snapshot" rather than changes in reflective practice over time meaning that the "messiness" of change and its non-linearity (Johnson & Golombeck, 2020) could not be captured. These limitations therefore need to be acknowledged.
Also missing are teacher educators' own reflective practices. Consideration needs to be given to how these might be collected to show how conceptual change is co-constructed through interaction, important when dialogic feedback is a feature of the assessment design. A challenge remains about how to overcome obstacles in gaining permission from ethics committees to study one's own practice in concert with students.