3. Contextual detail

Feedback is highly contextualised, and while the most obvious influencing factor is the preceding observation and its participants, purpose, and outcomes, the setting, both institutional and wider cultural context, will also impact feedback. As well as shaping feedback, these contextual details can help us understand it. In terms of the observation, Copland and Donaghue (Reference Donaghue2021) distinguish between evaluative and developmental approaches, with evaluative observations focused on assessing teachers’ performance, often to inform high stakes decisions such as employment, remuneration, promotion, or passing/failing teaching practice on a course. A developmental approach aims to support teachers in understanding their practice, helping them identify their strengths and weaknesses, and engage them in reflection. Table 2 below lists the studies in this review in terms of setting and observation purpose, and Appendix 2 gives contextual detail about each study. Perhaps the most striking information is the number of studies which do not clarify the purpose of observations, especially within university course settings. We may guess that these observations are evaluative because they are part of students’ teaching practicum, but it would be helpful to state this explicitly, especially as this may have some bearing on feedback and participants’ perceptions.

Table 2.

Contextual information

In terms of setting and participants, from the 52 research papers, 38 were carried out in pre-service contexts, 13 in in-service contexts, and one involved both pre and in-service teachers (Gakonga & Mann, Reference Gakonga and Mann2022). In the pre-service contexts, while the majority of studies (n = 31) were situated in university language teacher education courses (e.g. MA TESOL) with students doing a teaching practicum, only seven studies explicitly state the observation to be evaluative, while 21 give no detail about the purpose or outcome of the observation. Only three of this group are described as developmental, two of which were carried out during microteaching (Ekşi̇, Reference Ekşi̇2012; Karakaş & Yükselir, Reference Karakaş and Yükselir2025) and the third involving an experienced teacher on a MA TESOL course mentoring a novice teacher on a different version of the course (Gakonga & Mann, Reference Gakonga and Mann2022). The remaining pre-service studies were with trainee teachers on short courses such as the Cambridge English CELTA (Certificate in Teaching English to Speakers of Other Languages) course or Trinity London’s CertTESOL (Trinity Certificate in Teaching English to Speakers of Other Languages), all of which are classed as evaluative as observations are assessed and contribute towards course participants’ final grades.

From the in-service contexts, three studies give no indication about the nature of observations because all three focus on feedback in general, rather than feedback after a specific observation. Agheshteh and Mehrpour (Reference Agheshteh and Mehrpour2021) elicit general perceptions of feedback though a questionnaire issued across different institutions in Iran, and while Mehrpour and Agheshteh (Reference Mehrpour and Agheshteh2017) interview teachers and supervisors within one institution, they again ask about feedback in general. Similarly generic, Clark-Gareca and Warkentin (Reference Clark-Gareca and Warkentin2024) interview in-service teachers, asking them to remember past feedback that resonated with them and that they incorporated into their practice. Three of the in-service studies clearly indicate a developmental approach. Yaşar and Gürbüz (Reference Yaşar and Gürbüz2023) look at feedback following a one-year training course for teachers new to an institution, in which observations described as ‘non-directive’ were carried out by a professional development unit, and Inwang (Reference Inwang2024) investigates feedback following observations which were part of an e-coaching intervention to support teachers dealing with challenges. Both these studies involve expert observers, which means that the only in-service study (and in fact the only study in the whole data set) involving peer observation is Batlle and Seedhouse (Reference Batlle and Seedhouse2022) who focus on observations with participants of equal status and similar teaching experience. The rest of the in-service studies are evaluative, some with high stakes like Donaghue’s (Reference Donaghue2020a, Reference Donaghue2020b, Reference Donaghue2020c, Reference Donaghue2021) context in which observation reports contributed towards decisions about teachers passing a probationary year or having their three-year employment contracts renewed. Although Topal and Yiğitoğlu Aptoula (Reference Topal and Yiğitoğlu Aptoula2022) claim to study feedback following peer observations, the peer observers were accompanied by a ‘level coordinator’ manager, the observation purpose was to improve practice and reach ‘a quality standard amongst school’s practitioners’ (p. 3), and written feedback forms in which observers scored the teachers on a scale of 1 to 5 were submitted to the administration, making these observations more evaluative than the developmental focus expected from peer observation schemes. In Shah and Harthi’s (Reference Shah and Harthi2014) study, there is no information given about the purpose of observations or who the observers were. Mention of power imbalances and the fact that observers scored teachers’ performance suggest the observers were senior colleagues but given teachers’ negative attitudes towards observation and feedback, this study needs more contextual detail to understand and strengthen the researchers’ claims. In contrast, Louw and Billsborrow (Reference Louw and Billsborrow2016) provide much contextual detail in their study situated in four large private English-medium schools in Cambodia. Observations were found to be evaluative, bureaucratically imposed, and carried out by the head teacher, and feedback was directive, concerned with identifying and improving teachers’ weaknesses, and serving as a quality control mechanism. This contributed to both teachers’ anxiety about observation and the lack of reflection in feedback.

Returning to the themes identified, this review starts with studies which examine POF participants’ perceptions of feedback, elicited through surveys, interviews, and focus groups. Next I look at what actually happens during feedback talk, with studies analysing feedback interaction through audio and video recordings and transcripts. I look at two important aspects of feedback talk: teacher reflection, which underpins the developmental goal of feedback, and relationships, in particular the identity work that participants do in feedback and the ways feedback participants navigate through often difficult and delicate talk (facework). The next section focuses on observer training, and in the last section I offer recommendations for further POF research.

4. Positive perceptions

Many participants reported positive feedback experiences. Using similar methodologies, Martínez Agudo (Reference Martínez Agudo2016), surveying and interviewing student teachers in Spain, Tapia and Evison (Reference Tapia and Evison2024), surveying and interviewing student teachers and supervisors in Chile, and Yaşar and Gürbüz (Reference Yaşar and Gürbüz2023) interviewing in-service teachers on a training program in Turkey, all reported teacher satisfaction with feedback. The positive aspects highlighted included observers’ constructive and supportive feedback (Martínez Agudo, Reference Martínez Agudo2016; Yaşar & Gürbüz, Reference Yaşar and Gürbüz2023), the reflective and interactive nature of feedback (Yaşar & Gürbüz, Reference Yaşar and Gürbüz2023), and the value of feedback for teacher development (Tapia & Evison, Reference Tapia and Evison2024). The in-service teachers in Yaşar and Gürbüz’s (Reference Yaşar and Gürbüz2023) study found post observation feedback to be realistic, to the point, detailed, practical, and objective. They perceived neither hierarchical order nor (somewhat oddly) an obligation to action observers’ suggestions, perhaps because they were experienced in-service teachers. Also positive were student teachers’ evaluation of online and face-to-face feedback in Farr and Riordan’s (Reference Farr and Riordan2024) study. While observers and students had broadly positive views of both modes, students’ perceptions of face-to-face feedback included the opinion that interactions were rich and personal, while the online mode made for a calm, less face-threatening environment.

Tapia and Evison (Reference Tapia and Evison2024) elicited participants’ opinions on what they considered good feedback. Participants thought that observers should be open-minded, empathic, flexible, and good listeners, highlighting the importance of establishing a comfortable environment for the observed teacher. Observers’ pedagogical knowledge was considered important with observers valuing theoretical and methodological mastery while teachers valued teaching experience in schools. Mehrpour and Agheshteh (Reference Mehrpour and Agheshteh2017), interviewing Iranian in-service teachers and supervisors, drew up advice based on participants’ negative perceptions of feedback which included the recommendations that observers adopt a creative approach, be socioculturally sensitive, and elicit teachers’ beliefs and attitudes. Both studies point to an ideal view of feedback as being teacher centred.

Importantly, some studies asked teachers whether feedback had any impact on their practice. The teachers in Yaşar and Gürbüz’s (Reference Yaşar and Gürbüz2023) study responded positively, reporting a positive impact on their teaching and their adaptation to the institution, and expressed willingness to action feedback. In Inwang’s (Reference Inwang2024) study with in-service teachers in Nigeria on an e-coaching programme, teachers also reported learning and enhancements to their practice but these teachers made a more convincing case by giving specific details about what they had learned and implemented into their teaching. They reported thinking about and changing seating arrangements and student grouping, and challenging their previous teacher-centred approach. Teachers in this study also reported increased student participation, indicating the impact of feedback enactment. With a slightly different focus, Clark-Gareca and Warkentin (Reference Clark-Gareca and Warkentin2024) interviewed in-service ESOL teachers in the United States to find out what they had retained of their supervisors’ feedback from their student teaching days, and what ‘nuggets of supervisory wisdom’ continued to resonate with them. Teachers detailed aspects of feedback they continued to enact in their practice, including using a louder voice, preparing lessons and materials to anticipate student needs, and finding ways of connecting with students and making them feel welcome. A significant limitation to these studies is that they are all self-reports of feedback enactment, rather than empirical investigations into the changes teachers implemented in their practice and the impact of this on student learning. This point will be picked up in the recommendations section of this review.

5. Negative perceptions

In contrast to the research detailed above, a number of studies report dissatisfaction with feedback.

Mehrpour and Agheshteh’s (Reference Mehrpour and Agheshteh2017) interviews with Iranian in-service teachers and supervisors showed that teachers were generally dissatisfied with post observation feedback. In another study (Agheshteh & Mehrpour, Reference Agheshteh and Mehrpour2021), the authors again surveyed and interviewed in-service teachers in Iran and reported that teachers felt a lack of power to make their own decisions in the face of observers’ prescriptive approaches to feedback. The authors use strong language such as ‘power abuse’ and ‘struggle’ when describing teachers’ supervision experiences. Teachers felt that supervisors relied more on their institutional position and status than on their expertise when giving directives which led them to resist feedback. Teachers said they would prefer, but did not experience, a more reflective and collaborative approaches and a more equal power balance. Shah and Harthi (Reference Shah and Harthi2014), interviewing in-service teachers in Saudi Arabia, also revealed that teachers felt a lack of autonomy in an environment of mistrust, frustration, and insecurity. Teachers believed observers’ assessment and feedback to be subjective, biased, and unreliable, and observers’ lack of training contributed to teachers’ anxiety and stress. Shah and Harthi (Reference Shah and Harthi2014) refer to the ‘psychological plight’ (p. 1600) of teachers who considered observation a means to control teachers. Power imbalance reduced teachers’ participation in feedback sessions as teachers were reluctant to disagree with or challenge observers for fear of risking their jobs, thus undermining the developmental potential of the whole process. It is interesting that all these studies involve in-service teachers, indicating the need to further explore in-service teachers’ experiences of feedback.

While most of the studies discussed so far were limited to questionnaires and interviews in a specific location, Brandt (Reference Brandt2008) conducted participant interviews and questionnaire, collected participant journals, and shadowed trainees on CELTA and Trinity pre-service courses over a four-year period, involving 95 participants in nine countries. Results showed that feedback was ‘contentious and problematic’ (p. 37). The first reason for this was the nature of feedback. While teachers expressed the desire for authentic feedback, maintaining that sometimes feedback was overly lenient or critical, they equated real feedback with criticism (which they accept more readily from tutors than peers) and believed feedback became increasingly harsh as the course progressed. Participants believed this was because tutors were under increasing pressure to ensure that trainees understood what they had to do to pass the course and increasing familiarity with each other. Like the teachers in Shah and Harthi’s (Reference Shah and Harthi2014) study, teachers experienced inconsistency among observers in interpreting course objectives and assessment criteria. Some teachers found that the value of listening to peer feedback, while useful in the early stages, diminished as the course progressed. Teachers preferred to avoid the confrontation that could result from giving negative feedback to peers and became increasingly focused on their own survival and performance. This finding chimes with Ekşi̇’s (Reference Ekşi̇2012) study which focused on trainee teachers giving feedback to each other after microteaching lessons. While these teachers acknowledged the benefits of giving and receiving peer feedback, their concern for each other’s feelings, their worry over potential face threat, and cultural constraints made them reluctant to provide overt oral feedback. The teachers were more comfortable providing feedback confidentially via a written form. The form developed teachers’ own observation and reflection skills and teachers reported that they learned more from one another’s performance through the structured feedback form. The form increased feedback dramatically and the observed teachers valued the detailed written feedback from their peers. Finally, like those detailed in Tapia and Evison’s (Reference Tapia and Evison2024) paper, the teachers in Brandt’s (Reference Brandt2008) study believed time constraints limited their opportunities to explain and justify their actions and choices in feedback sessions.

Louw and Billsborrow (Reference Louw and Billsborrow2016), surveying and interviewing teachers and academic managers in Cambodia, investigated participants’ views about the purpose of feedback. Both parties viewed feedback as wholly evaluative. Observers viewed feedback as a bureaucratically imposed means of quality control rather than a means of teacher development and rejected the notion of eliciting or discussing teachers’ ideas during feedback. Teachers viewed observation and feedback as a job requirement rather than as a means of improving their classroom practice or helping them develop.

6. Observers’ perceptions versus reality

One of the most interesting aspects of research into participants’ perceptions of POF is a mismatch between observers avowed feedback approach and the reality of their practice. Louw et al. (Reference Louw, Watson Todd and Jimarkon2016), through interviews and feedback transcript analysis, found that while observers valued a collaborative, reflective approach to feedback, in practice they adopted a directive approach and were interactionally dominant. Copland (Reference Copland2012) discovered a similar disjuncture where CELTA observers professed to value developmental, dialogic feedback but actually pushed their own strong opinions on language teaching pedagogy, giving their views on best practice and privileging these through self-selections, interruptions, and long turns. Teachers were encouraged to listen to and take on board these views, resulting in teachers’ learning agendas being rarely heard. Copland brings our attention to ‘legitimate talk,’ that is, ‘who is allowed to speak, to whom, about what, and whose knowledge counts’ (p. 16) and the importance of being aware of how this is established and maintained in POF.

In the next section, we take up the idea of empirical investigation and look at studies which go beyond feedback participants’ perceptions of feedback to look at what actually happens in POF meetings. About a decade ago, Copland (Reference Copland2012) and Farr (Reference Farr2011) lamented the fact that research on feedback talk was scarce. However, since then research into feedback talk has increased steadily, due in part, perhaps, to technological advances which makes it increasingly easier to record and transcribe feedback talk, but also in response to the growing interest in the role of talk in learning (Sert, Reference Sert, Dippold and Heron2021). In the following section, we also circle back to the goal of feedback by examining an important aspect of developmental feedback: reflection.

7. Reflection

There is a considerable body of research to support the view that reflection is important in language teacher learning and development (Farrell Reference Farrell, Walsh and Mann2019). Although reflection is a vague concept (Copland et al., Reference Copland, Ma and Mann2009), a central skill is the ability to analyse experience to improve practice. Reflection often starts by identifying a problem – a ‘puzzling or troubling or interesting phenomenon’ (Schön, Reference Schön1983, p. 50) – followed by description, analysis, and action (Dewey, Reference Dewey1933). During this process, teachers build knowledge of educational practices and outcomes, consider context, integrate knowledge, theory, and practice, and examine personal beliefs, experiences, emotions, and attitudes. As reflection influences and changes practice, it becomes reflective practice. However, reflection can be hard for teachers to do alone as they are not always able to identify, analyse, and solve classroom problems without help (Akbari, Reference Akbari2007). This has led scholars to recommend more collaborative, evidence-based, and data-led approaches to reflection (Brandt, Reference Brandt2008; Mann & Walsh, Reference Mann and Walsh2013, Reference Mann, Walsh, Howard and Donaghue2014; Sert, Reference Sert, Dippold and Heron2021). Theoretically, post observation feedback conferences, as ‘designed interactional spaces for teachers’ professional development through dialogic reflection’ (Topal & Yiğitoğlu Aptoula, Reference Topal and Yiğitoğlu Aptoula2022, p. 101,016), represent ideal opportunities for teachers to engage in reflection (Kobayashi & Kobayashi, Reference Kobayashi and Kobayashi2024; Wagner & Lewis, Reference Wagner and Lewis2021). Observers, as interlocutors, are well placed to help teachers explain, conceptualise, justify, and critically evaluate their teaching (Engin, Reference Engin2013; Louw & Billsborrow, Reference Louw and Billsborrow2016). However in practice, reflection in feedback varies and is often scarce (Copland et al., Reference Copland, Ma and Mann2009; Engin, Reference Engin2013; Waring, Reference Waring2013).

Two earlier studies (Copland et al., Reference Copland, Ma and Mann2009; Hyland & Lo, Reference Hyland and Lo2006) reveal limited reflection in POF. Copland et al. (Reference Copland, Ma and Mann2009), examining group POF sessions with novice teachers on training courses in South Africa and the UK, found that while reflective talk was present in these feedback meetings, it was less than the authors expected, as observers viewed reflection as important and also asked many questions during feedback. Hyland and Lo (Reference Hyland and Lo2006) studied feedback with pre-service teachers in Hong Kong and also discovered limited teacher reflection due to a dominance of observer talk. Observers spoke much more than teachers, were much more likely to initiate topics, and expressed mostly directives. Teachers were largely passive and most of their responses were either giving information or agreeing with and accepting the observer’s comments. While observers invited teachers to express their views at the start of feedback meetings, teachers subsequently had few chances to express their feelings, negotiate meaning, or initiate topics that were not on the observer’s agenda. Hyland and Lo also ascribed the lack of reflection to an imbalance in the power relationship between student teacher and observer.

Watson and Williams (Reference Watson and Williams2004) evaluated delayed feedback which involved teachers keeping a structured reflection journal, comparing this with immediate feedback. Their analyses offered some evidence of a higher level of teacher reflection in the delayed feedback through increased teacher topic initiation and types of reasoning talk. However, reflective journals have also been critiqued for eliciting mechanical, inauthentic reflection, and superficial engagement (Hobbs, Reference Hobbs2007; Mann & Walsh, Reference Mann and Walsh2013). Hobbs (Reference Hobbs2007) raises and explicates the problematic nature of forced/assessed reflective practice, describing how this provoked strategic and inauthentic responses, and in some cases hostility, from the teachers in her study.

Copland et al. (Reference Copland, Ma and Mann2009) identified four main reasons for the many missed opportunities for reflection. First, because observers believed that their primary role was to develop trainees’ teaching skills, most of the POF talk centred around the pedagogy of teaching – what to do and how to do it. This meant that although trainees were often invited to comment on their own and others’ teaching, these promising beginnings rarely led to reflective talk because observers’ motivation for eliciting trainees’ views was usually to orient talk to their own agenda of teaching pedagogy. Second, the assessment criteria related to teaching performance dominated feedback, squeezing out the focus on reflection. Third, the shortness of time for each trainee in the group feedback meetings compromised opportunities for them to reflect on practice. Fourth, trainees either did not know how to reflect or did not want to reflect.

8. Resource-enhanced reflection

Since these earlier studies, there has been a growing body of research examining whether resources such as written artefacts and technology (especially video) can help create space for reflection in POF. Engin (Reference Engin2015b) examined the affordances of a written artefact – the running commentary written by an observer while watching the lesson – as a tool for thinking, reflection, and discussion. The running commentary acted as a prompt for observer and teacher to share an understanding of the lesson and helped teachers recall, and thereby reflect on, the lesson. It also acted as a catalyst for discussion and guided, structured, and focused the discussion. However, the shared understanding prompted by the running commentary also limited discussion as the lack of justification or elaboration needed stifled reflection. The commentary also created a power asymmetry and bias in favour of the observer who chose which events to record which then structured the feedback conversation. Donaghue (Reference Donaghue2020b) uncovered similar constraints with a different artefact – an institutional observation form with assessment criteria which the observers used to score teachers’ performance during the observed lesson. The observer in her study structured the subsequent POF meetings around the criteria on the form which resulted in him asking few questions and making mostly evaluative statements. The form also gave prominence to his own perception of the lesson, limiting opportunities for teacher interaction and reflection.

Three studies (Ahmad, Reference Ahmad2024; Kanat Mutluoğlu & Balaman, Reference Kanat Mutluoğlu and Balaman2023; Sert, Reference Sert2024) evaluated the use of a video tagging tool (Video Enhanced Observation – VEO) to promote reflection in POF. VEO allows viewers to tag and annotate parts of lesson recordings. Tags can be both pre-established or newly created. For example, observers in Sert’s (Reference Sert2024) study used a framework of pre-established tags aligned with practicum learning outcomes which included tags for classroom interaction, classroom management, use of language, and use of gestures. Teachers and observers can then watch and discuss the tagged episodes together. All three studies investigated VEO through participant evaluation. Participants reported enriched, evidence-based reflection, collective engagement and dialogue, and future-oriented transferability of learning. The studies reported only minor challenges which included technical issues and the need to become familiar with the software. In a similar study, Turan and Yiğitoğlu Aptoula (Reference Turan and Yiğitoğlu Aptoula2023) examined how participants made use of the screen-sharing function in video-mediated feedback sessions but rather than elicit participants’ perceptions, they analysed video recordings of feedback sessions. Analysis showed that screen sharing helped repair misunderstanding, balance knowledge asymmetry, mitigate potential resistance to advice, and promote dialogic reflection. Gakonga and Mann’s (Reference Gakonga and Mann2022) analysis of peer observers’ video-stimulated recall showed that this allowed for a more teacher-centred approach and although there was some discomfort for the teacher in watching their classroom performance, it reduced face threat and allowed reflection because of the evidence-based nature of the feedback.

Three studies (Baecher & McCormack, Reference Baecher and McCormack2015; Karakaş & Yükselir, Reference Karakaş and Yükselir2025; Kobayashi & Kobayashi, Reference Kobayashi and Kobayashi2024) compared video-mediated feedback with ‘traditional’ feedback to find out if video enhanced pre-service teacher reflection. The smaller-scale studies (Karakaş & Yükselir, Reference Karakaş and Yükselir2025; Kobayashi & Kobayashi, Reference Kobayashi and Kobayashi2024) reported some enhancement in reflection: video prompted teachers to notice students (Kobayashi & Kobayashi, Reference Kobayashi and Kobayashi2024), become more evaluative and more aware of areas for improvement in their teaching, contemplate future actions to improve their teaching and solve problems (Karakaş & Yükselir, Reference Karakaş and Yükselir2025). Baecher and McCormack’s (Reference Baecher and McCormack2015) detailed analysis involved examining word count, turns taken, and turn length. Analysis revealed that the video-based feedback involved more teacher talk than observer talk and teachers evaluated, suggested, and described more, and participated more actively and collaboratively. Conversations were more teacher than supervisor centred. Interestingly, evaluation and suggestion, usually so prevalent in POF talk, occurred less in the video-mediated POF sessions. Like Gakonga and Mann’s (Reference Gakonga and Mann2022) participants, the use of hedging to soften potentially face-threatening statements decreased for all participants in the video-based feedback. As observers referred to the video record, their use of questioning was also much higher. Thus in all three studies, using the evidence of video data enhanced teacher reflection.

Bates and Donaghue (Reference Donaghue2021) examined the use of synchronous computer-mediated communication (SCMC) as a peer feedback tool with teacher trainees. As a group of trainee teachers and their tutor observed a peer teaching, they chatted via an instant messaging app, commenting on the lesson in real time. The written chat feed was later made available to the observed teacher. Through analysing chat records, Bates and Donaghue show that SCMC supported the development of reflective practice as it promoted collaboration and dialogue, enabling trainees to identify and solve problems, share ideas and opinions, give constructive feedback, and reflect on their own and their peers’ teaching. Tutor questions helped develop dialogue, in particular questions that asked for further explanation and clarification, and SCMC helped build rapport and democratised feedback as trainees initiated discussions, chose topics, and often interacted with each other in stretches of dialogue without the tutor. This contrasts with Copland’s (Reference Copland2012) findings that in face-to-face post-observation group feedback, interaction between trainees and multiparty discussions were rare as trainees talked directly to the trainer while the other trainees listened, and trainers typically controlled turns, nominating trainees to speak. The SCMC chat feed, as a tool to aid memory and gain a different perspective, also prompted post-lesson reflection for the observed teacher.

9. Observer reflection

While the studies above focus on teacher refection, Topal and Yiğitoğlu Aptoula (Reference Topal and Yiğitoğlu Aptoula2022) focus on observers and show that in a peer observation scheme with in-service teachers, POF also prompted observers to reflect as they questioned, clarified, and considered future changes to their own practice. Thus the observer was a subject as well as agent of change. The fact that this paper alone features observer reflection points to this as a possible avenue of further research.

10. Suggestion on how to promote reflection in POF

Copland et al. (Reference Copland, Ma and Mann2009) offer suggestions on how to increase opportunities for reflection in feedback. Their primary recommendation is that observers move from a directive to a dialogic approach (Brandt, Reference Brandt2008), suggesting that observers reduce their talk, make space for reflection, and use and value trainees’ experience and knowledge of classrooms and education. However, Waring (Reference Waring2013) maintains that there is also room for reflection in directive approaches. According to Waring, directive moves such as assessment or advice can engender teacher reflection. Waring’s (Reference Waring2013) analysis of feedback talk shows how teachers engaged in reflective talk as they accepted or rejected assessments. Teachers offered detailed reasoning for an alternative course of action, critically assessed materials, reconsidered teaching practice, and narrated the means and difficulties in reaching a desired goal. The fact that the ‘much valued commodity of reflection’ (Waring, Reference Waring2013: 115) can emerge in response to advice and assessment is especially interesting as these speech acts persist as integral components of observer talk (Nguyen, Reference Nguyen2023). Turan and Yiğitoğlu Aptoula (Reference Turan and Yiğitoğlu Aptoula2023) suggest that reflection and assessment agendas can coexist by showing how observers moved between and managed both reflection and evaluation-oriented sequences in the same meeting. When evaluating, observers in their study asserted their knowledge and authority (or in Turan and Yiğitoğlu Aptoula’s terms ‘upgraded epistemic asymmetry’) in order to ensure teachers’ acceptance of feedback. In these evaluative sequences, observers’ turns were maximised and they used deontic modality to express certainty and obligation (e.g. must, should). In contrast, teachers’ turns were minimised – often reduced to acknowledgement tokens such as nodding or ‘mmhm’ – with little or no response or resistance to suggestions or advice. However, these POF meetings also had reflection-oriented sequences in which observers downgraded their knowledge and authority while upgrading teachers’, decreased their turn-taking, and used epistemic modality that marked guarded commitment (e.g. may, might, perhaps) while also listening more to teachers. In these sequences teachers’ turns and reflective contributions were increased. These studies are important in helping us move away from a mutually exclusive dual purpose view of feedback to one in which participants navigate both evaluation and development.

Copland, Ma, and Mann’s other recommendation to promote reflection is to better prepare teachers and observers for feedback. For teachers, training on the importance of reflection, how to reflect, and how to participate in observation feedback could be built into a course. Observer training could include explicit focus on strategies for eliciting (and restricting) reflective talk as well as awareness raising analysis of feedback transcripts, through which observers may surface their impact on teacher reflection. Observer training is a recommendation made by many of the authors in this review and I return to this in the recommendation section below.

Brandt (Reference Brandt2008) makes a number of suggestions to promote reflection but some seem unfeasible due to administration, staffing, and resourcing, for example including a non-evaluating facilitator and language learner representatives in the POF meetings and making peer attendance optional or by invitation of the observed teacher (which would be difficult to organise in group feedback). In addition, the requirement for assessment on CELTA and Trinity courses would make difficult Brandt’s suggestion that course participants have the option not to attend some meetings. Her suggestion that tutors resist telling trainees what they did right and wrong and offer no unsolicited feedback is at odds with pre-service teachers’ expectations and desire for an expert’s assessment of their teaching (Farr, Reference Farr2011; Vásquez, Reference Vásquez2004). Suggestions which seem more feasible are making peer evaluation optional rather than obligatory or routine, and positioning reflection and feedback as goal-related, with teachers being encouraged to ‘identify their individual goals, reflect upon their achievement of these goals and seek feedback related to them where desired’ (Brandt, Reference Brandt2008: 44).

Engin (Reference Engin2013) looked at the relationship between reflection and scaffolding, examining levels of scaffolding talk, that is, observer talk that supports and guides the construction of teaching knowledge by prompting teachers to reflect on, justify, explain, and consider alternatives in their teaching. Engin presents a data-generated framework (see Table 3 below) of different levels of scaffolding talk, explaining that the quality and specificity of scaffolding questions and prompts at particular moments in POF interaction can support the construction of knowledge.

Table 3.

Levels of scaffolding talk (adapted from Engin, Reference Engin2013)

However, she includes the important caveat that effective scaffolding is contingent on the learner and the context – interaction depends on variables such as observer and teacher experience, confidence, the topic of the discussion, expectations, and emotional states. Observers therefore have to try out various strategies and techniques to elicit reflection (and judge whether teachers are ready for reflection (Bąk-Średnicka, Reference Bąk-Średnicka2024)). Engin concedes that a hierarchical conceptualisation of scaffolding talk may at first seem an overgeneralisation, but presents these levels as a point of reference for observers interested in eliciting teacher reflection. In a further paper, Engin’s (Reference Engin2015a) detailed analysis of feedback talk confirms both the potential for observers to scaffold teachers’ reflection and learning in POF and the influence of context on the success of this process.

11. Reflection and context

POF, like all social interaction (and indeed all learning) is highly situated and contextualised (Donaghue, Reference Donaghue2021; Engin, Reference Engin2013; Gakonga & Mann, Reference Gakonga and Mann2022) so it is important to guard against advocating ‘best practice’ (Gakonga & Mann, Reference Gakonga and Mann2022) for eliciting teacher reflection in POF. However, the studies in this section offer feedback participants points of reference (Engin, Reference Engin2013) and empirically based strategies that observers can consider using in their feedback talk. In some contexts, reflection may not be possible (or even desirable). For example, Louw and Billsborrow (Reference Louw and Billsborrow2016) contend that a reflective practice paradigm is inconsistent with the context of their study – Cambodia – due (in part) to the incompatibility of reflection with the fast-paced school culture in which the data was collected and with the high power distance of Cambodian national culture. The authors explain high power cultures to be those in which followers endorse unequal power distribution and which conform to a status hierarchy. They conclude that:

A teacher, as ‘inferior interlocutor’ (p. 50), may feel it inappropriate to question or challenge the ‘superior’ observer. Similarly, Nguyen (Reference Nguyen2023) questions whether the dialogic, reflective approaches advocated in Western contexts are appropriate for her Vietnamese context in which teachers are used to occupying a listener role and observers are positioned as ‘knowledge givers.’ Thus, cultural aspects, such embedded social power, can influence feedback considerably, giving weight to the argument for studying feedback in context.

12. Relationships

We move now to review research which examines how feedback participants negotiate relationships. In the first part, we look at the identities teachers and observers claim, contest, and verify in feedback and the linguistic means by which this is done. In the second part we look at how teachers and observers manage to maintain social, collegial relationships while giving and receiving critical feedback, that is, how feedback participants do facework.

13. Identity

From the POF literature, identity emerges as an important aspect of feedback. Teacher identity has become a central focus in language teaching research, prompted by the growing realisation that teacher education involves the development of a teacher identity as much as acquiring teaching knowledge and skills (Varghese et al., Reference Varghese, Motha, Park, Reeves and Trent2016). Teachers construct and negotiate ever-evolving identities throughout their career (Kanno & Stuart, Reference Kanno and Stuart2011), and identity guides teachers’ beliefs, assumptions and values, as well as influencing what teachers do (Farrell, Reference Farrell2011). Identity work also helps teachers exercise professional agency, development, and growth (Clarke, Reference Clarke2009). Talk is perhaps the most important resource for constructing identities (Varghese et al., Reference Varghese, Morgan, Johnston and Johnston2005). As one of the few opportunities teachers have to talk about their teaching with another person, POF represents a discursive space for teachers to explore and construct professional identities (Urzúa & Vásquez, Reference Urzúa and Vásquez2008).

One way in which identities are achieved in POF is through language – particular utterance types, speech acts, or word choices. For example, Urzúa and Vásquez (Reference Urzúa and Vásquez2008) looked at the future forms will and going to, and Vásquez and Urzúa (Reference Vásquez and Urzúa2009) analysed reported speech and reported mental states. In both studies student teachers simultaneously communicated identities of confident/knowledgeable/assertive teachers and hesitant/inexperienced/unskilled teachers. For example, Vásquez and Urzúa (Reference Vásquez and Urzúa2009) showed how teachers used direct reported speech to foreground accomplishments and developing expertise, enabling them to present themselves as efficient, skilful, and confident, but when reporting mental states, the novice teachers highlighted uncertainty, gaps in knowledge, or negative feelings and emotions, thereby indexing an insecure, unskilled novice identity. Donaghue (Reference Donaghue2019), studying feedback with experienced, in-service teachers, focused on how display questions and answers were used to claim identities. Teachers claimed positive identities by voicing their knowledge and expertise in response to display questions and an observer used display questions to claim a practising teacher identity in order to reassure teachers of his worth as an observer and leader. Display questions also helped the observer perform a manager identity which allowed him to control topics and evaluate teachers. While some teachers verified this identity by complying and providing the required answer, one teacher resisted, replying reluctantly and with irritation, thereby contesting the observer’s manager identity. This shows the importance of the interlocutor in maintaining a claimed identity.

Feedback participants also construct identities through telling stories. Vásquez (Reference Vásquez2007) explored reflective narratives (when teachers explain or make sense of their thoughts and actions during the observed lesson) and relational narratives (stories about interactions with problematic students beyond the observed lesson) produced by novice language teachers, revealing the unstable, uncertain, and inconsistent identities which were built in these narratives. These unstable identities were constructed by either the teacher narrator or by the observer responder who destabilised, reconstructed, interpreted, and evaluated teachers’ narratives, again pointing to the interlocutor’s role in constructing, maintaining, and contesting the identities of their conversational partner. Wagner and Lewis (Reference Wagner and Lewis2021), while not researching identity explicitly, present sequences which are similar to the relational narratives in Vásquez’s study. They analysed sequences in which a novice teacher complained about non-present third parties and, like Vásquez, also examine the observer’s response. The observer displayed a lack of affiliation and disattended to the teacher’s complaints, either by shifting focus to another aspect of the complaint story or by offering a more positive reinterpretation. Wagner and Lewis ascribe this lack of affiliation to the observer’s institutional and professional role or identity. However, they argue that observers could better support novice teachers’ developing reflective practice, provide them with guidance, and help them to understand and manage their emotional responses if they did not downplay their complaint stories. Complaints enable important identity work and bring up important professional development issues (e.g. teaching beliefs, relationship conflicts, reactions to negative feedback) so novice teachers may benefit from a more affiliative response and more empathy and understanding of their emotions, interpretations, and responses to events.

Feedback participants also claim identities through various power moves. For example, an observer in Donaghue’s (Reference Donaghue2020a) study manifested an identity of experience, power, assessment, and control at the beginning of his feedback meetings by opening meetings, claiming expert knowledge, exercising the right to evaluate a conversational partner (the observed teacher), taking and keeping the floor, and initiating new topics. However, this powerful identity shifted as the in-service teachers he had observed also indexed powerful identities by initiating new topics, taking/keeping control of the floor with extended turns, assuming the more powerful identity of questioner by challenging the observer’s scores, and positioning the observer as the less powerful explainer/justifier. At key points in the meeting, this reversal of institutional power hierarchy rendered the observer unable to commit to critical feedback he had intended to deliver. The dominance of identities of experience and power from both parties also resulted in the observer making mostly evaluative statements and asking teachers very few questions, leaving little room for participants to engage in collaborative, reflective, or developmental discussion. Thus identity work compromised the developmental goal of feedback.

Context also influences identity work in POF. For example, Donaghue (Reference Donaghue2020b) showed how one particular identity was co-constructed, ratified, and prioritised by teachers and observers during feedback: a teacher proficient in and enthusiastic about technology. Observers ascribed this positive identity to teachers through praise and favourable comparisons. Teachers claimed the valued identity for themselves by demonstrating knowledge and expertise. However, other teachers were less successful in their inability to inhabit the valued identity and found themselves marginalised, with one even leaving the institution. The technologically proficient identity was related to a broader context of government and institutional initiatives encouraging and often mandating the use of technology in teaching. Thus, feedback became a mechanism of conformity as identity work during feedback acted as a means by which participants absorbed and reified institutional priorities.

14. Facework

Feedback is complex (Farr, Reference Farr2011), and while POF meetings can be opportunities for teachers to reflect and critically assess their teaching to mediate change (Farr, Reference Farr2011), they can also be ‘difficult discoursal events’ (Copland, Reference Copland2008: 67). POF is often evaluative and observations can carry high stakes for teachers (e.g. a course or probationary assessment). Additionally, observers are frequently negotiating tensions between maintaining social, collegiate relationships and giving/receiving critical feedback aimed at improving teaching practice and enabling teacher development (Brandt, Reference Brandt2008; Louw et al., Reference Louw, Watson Todd and Jimarkon2016). Teachers, too, are trying to balance requirements to demonstrate reflection and self-evaluation with the need to be positively evaluated. Feedback therefore often features difficult and delicate talk such as criticising, questioning, defending, advising, and suggesting. This section examines how feedback participants balance these tensions, focusing on the work participants do to negotiate relationships with others, that is, relational work or facework (Locher & Watts, Reference Locher and Watts2005).

Although outside the time frame of this review, Ruth Wajnryb’s (Reference Wajnryb1994) seminal work deserves mention as she introduced the idea of face in feedback, showing how supervisors soften critical messages via different linguistic devices to mitigate face threat, that is, a challenge to a person’s social identity or self-esteem. Similarly, Vásquez’s (Reference Vásquez2004) study also focused on observers who, when criticising, suggesting, and advising, employed politeness strategies such as lexical hedges (e.g. maybe, just), modals, mental verbs, indirectness, positive evaluation, and self-denigration to avoid face threat. Vásquez found, however, that these face protecting strategies often softened the critical message so much that teachers missed the fact that their teaching needed improvement, compromising the purpose of feedback. While Vásquez’s (Reference Vásquez2004) study examined a variety of strategies, two later studies (Batlle & Seedhouse, Reference Batlle and Seedhouse2022; Waring, Reference Waring2017) focused on particular facework moves. In Waring’s study, observers employed facework by ‘going general’ or depersonalising advice and invoking larger disciplinary or pedagogical principles which served to take the spotlight off the teacher and present a problem as common. Batlle and Seedhouse looked at a third turn which followed a negative assessment from an observer and a response from the teacher. In the third ‘affiliative’ turn, observers made use of the teaching community, presenting themselves as fellow teachers and construing the problem as common to all teachers, in a similar move to Waring’s ‘going general’. These facework moves allowed observers in both studies to accomplish the developmental work of ensuring the teacher understood and accepted a problem, and socialising teachers into important disciplinary and pedagogical practices and ideas.

These studies focus on the strategies supervisors use to avoid or manage face threat. In contrast, Copland (Reference Copland2011) and Donaghue (Reference Donaghue2018, Reference Donaghue2021) take a wider view of face in feedback and maintain that context influences participants’ orientation (or not) to face threat: cultural norms of interaction, personal relationships, and speech genre all influence what is said and how this talk is interpreted by interactants. Donaghue (Reference Donaghue2021) illustrated the influence of context by revealing how account requests (i.e. a supervisor asking the teacher to explain their actions) were interpreted as face threatening in some feedback meetings but not in others. Copland (Reference Copland2011) showed that the feedback participants in her study oriented to speech acts such as giving advice and criticising as normal, leading her to claim that the feedback genre can allow, even expect, acts that might be considered face threatening in other contexts. However, even though certain kinds of behaviour can be considered generically acceptable, feedback participants can overstep the generic norms and move into face threatening behaviour, often deliberately. In Donaghue’s (Reference Donaghue2018, Reference Donaghue2021) studies, both observers and in-service teachers were frequently willing to risk face threat. In-service teachers risked face threat to justify and explain their actions while observers risked face threat to fulfil feedback goals – they challenged decisions and behaviour they felt could be improved and attended to their institutional duty of evaluating teachers as they advised, suggested, questioned, and criticised. However, while this risky behaviour often paid off and their interactional partners accepted these moves, sometimes the risk backfired and the interactional partner oriented to face threat. For example, in Donaghue’s (Reference Donaghue2018) study, one observer aggressively delivered bald, unmitigated criticism to ensure the teacher improved his practice. However, the teacher oriented to this criticism as face threatening and resisted the observer’s feedback, choosing to believe that his teaching required no attention or improvement. In another example (Donaghue, Reference Donaghue2021), an observer oriented to face threat after a teacher challenged his suggestion that she ask students to read aloud. He closed down the topic, ending the teacher’s attempts at discussion and reflection.

At the other end of the spectrum, research reveals that POF talk can be face-supporting, that is, interactants recognising each other’s merit and position. At an interactional level, Donaghue (Reference Donaghue2021) showed how in-service feedback participants engage in face support. Teachers do this by conceding to and accepting observers’ advice and asking for and agreeing with suggestions, confirming the observer’s position of expert, authority, and advice giver. Observers engage in face support by recognising teachers’ ability and independence and expressing confidence in teachers’ ability to incorporate suggestions into their teaching. This facework promotes alignment and allows observers to perform their role and fulfil the goal of the meeting. Lewis and Wagner (Reference Lewis and Wagner2023) focused on empathy in POF, looking at observers’ responses to teachers’ problems in a move the authors term ‘relate to self’, that is, sharing a similar experience. This empathetic move normalised teachers’ problems by framing them as typical or an expected part of becoming a teacher. This provided emotional and professional support for teachers, helped socialise teachers into the profession, and reassured teachers that their problems are surmountable. Other linguistic means of providing emotional support include providing space for teachers to develop self-confidence and identity (Bąk-Średnicka, Reference Bąk-Średnicka2024), delivering compliments before criticisms or specific suggestions, and helping teachers to pinpoint their own problems (Le & Vásquez, Reference Le and Vásquez2011).

15. Observer training

In this final review section, we look at studies focused on observer training. Three POF studies focus on observer development and training. Mutlu–Gülbak and Akcan (Reference Mutlu–Gülbak and Akcan2024) explored how observers’ feedback skills improved after they did an online mentor training program. Similarly, in a study focusing on feedback interaction, Vásquez and Reppen (Reference Vásquez and Reppen2007), adopting a longitudinal researcher-as-participant design, analysed their own feedback interactions over two semesters. Analysis of the first semester transcripts raised awareness of their linguistic and interactional practices which then prompted them to implement changes in semester two, aiming for a more teacher (rather than observer) centred approach. Baecher et al. (Reference Baecher, Graves, Ghailan and College2018) also used feedback recordings in an initiative where observers in a self-development group analysed videos of their post-observation conversations with a peer.

All three studies reported raised observer awareness of practice. Observers in Mutlu–Gülbak and Akcan’s (Reference Mutlu–Gülbak and Akcan2024) project reported recognising the importance of a positive attitude and the need to prompt reflection to foster teacher development. In Vásquez and Reppen’s (Reference Vásquez and Reppen2007) study, analysis of the first semester interaction revealed that observers produced most of the talk with teachers being more passive. Baecher et al.’s observers found watching the video records of their conversations to be ‘intriguing … affirming, enlightening, and motivating in terms of a desire to further grow in their supervisory process’ (Baecher et al., Reference Baecher, Graves, Ghailan and College2018: 567). However, individual refection and peer discussion also uncovered a gap between their stated positionality and their actual behaviour, a recognition that POF did not always achieve observers’ objectives, and realisation that feedback talk was often formulaic. Observers recognised that they tended to focus on the observed lesson rather than considering a larger developmental plan. Observers also recognised how constraints such as limited time impacted their feedback.

While Baecher et al.’s (Reference Baecher, Graves, Ghailan and College2018) study focused solely on raising observers’ awareness, the observers did also consider ways of changing their practice such as how to draw teachers into conversation and, more broadly, how teachers could be better prepared to take part in POF conversations. Observers in Vásquez and Reppen’s (Reference Vásquez and Reppen2007) study took a further step by implementing changes in their second semester. In an effort to prompt dialogue and reflection, observers consciously tried to ask more questions to prompt teachers to talk more, a strategy which was also adopted by the observers in Mutlu–Gülbak and Akcan’s (Reference Mutlu–Gülbak and Akcan2024) study. Observers in Vásquez and Reppen’s study also considered the metadiscursive positioning that takes place at the beginning of POF meetings which sets the stage and establishes interactional roles and responsibilities for each participant. Observers consciously changed from positioning themselves as primary speakers and teachers as primary listeners to positioning teachers as active contributors to the conversation. As a result of these changes, teachers talked more in the POF meetings in the second semester and observers were less dominant and less directional.

16. Research recommendations

Based on the research reviewed above, I offer four research recommendations. The first comes from many of the authors in this review (e.g. Baecher et al., Reference Baecher, Graves, Ghailan and College2018; Baecher & McCormack, Reference Baecher and McCormack2015; Copland, Reference Copland2011, Reference Copland2012; Copland et al., Reference Copland, Ma and Mann2009; Donaghue, Reference Donaghue2019, Reference Donaghue2020a, Reference Donaghue2020b; Engin, Reference Engin2013; Hyland & Lo, Reference Hyland and Lo2006; Kobayashi & Kobayashi, Reference Kobayashi and Kobayashi2024; Le & Vásquez, Reference Le and Vásquez2011; Louw et al., Reference Louw, Watson Todd and Jimarkon2016; Mutlu–Gülbak & Akcan, Reference Mutlu–Gülbak and Akcan2024; Shah & Harthi, Reference Shah and Harthi2014; Vásquez & Reppen, Reference Vásquez and Reppen2007) who, although perhaps not individually aware of this, join a loud and insistent call for observer training. As the reviews above show, feedback is complex and often difficult, and observers’ behaviours, discourse, and judgements can cause teacher stress and anxiety (Agheshteh & Mehrpour, Reference Agheshteh and Mehrpour2021; Shah & Harthi, Reference Shah and Harthi2014). This indicates that observers would benefit from training and support. In the field of language teaching, observers have little (or no) training in managing feedback. Observers are rarely required or encouraged to examine and reflect on their own practice, and are given few professional development opportunities, unlike the teachers they observe, despite the important role they play in teacher development. To help observers become more critically aware of their professional talk, they could be guided in analysis of extracts from their own feedback talk, like the observers in Baecher et al. (Reference Baecher, Graves, Ghailan and College2018), Mutlu–Gülbak and Akcan (Reference Mutlu–Gülbak and Akcan2024), and Vásquez and Reppen’s (Reference Vásquez and Reppen2007) studies. Suggested foci for observers’ analysis include noting the degree to which teachers are given a chance to make their own points and explore their own agendas (Hyland & Lo, Reference Hyland and Lo2006), examining the distribution (i.e. who talks most) and type (monologic or dialogic) of feedback talk (Copland & Donaghue, Reference Donaghue2021), noticing how observer talk impacts on the response, and thus the reflection and learning opportunities, of teachers (Engin, Reference Engin2013), and noticing the identity and face work of both teachers and observers and the impact this has on teacher’s acceptance of feedback. While the studies in the observer training section of this review offer promising results, more is needed in both practice and research to support observers in maximising the learning affordance of POF.

My second recommendation is for POF research and practice to include a focus on feedback enactment. Some of the papers in this review (Clark-Gareca & Warkentin, Reference Clark-Gareca and Warkentin2024; Inwang, Reference Inwang2024; Yaşar & Gürbüz, Reference Yaşar and Gürbüz2023) touch on feedback enactment. It is interesting to note that these are all recent papers, perhaps influenced by the burgeoning literature over the last decade on student feedback and the recognition within this field that effective feedback requires the receiver to understand and enact the feedback (Pitt & Quinlan, Reference Pitt and Quinlan2022). Considering the attention that feedback enactment has been given in the student feedback literature, and the fact that feedback is useless unless acted on, it is surprising that there is so little research looking at post observation feedback enactment (Donaghue & Heron, Reference Donaghue and Heron2025; Heron et al., Reference Heron, Donaghue and Balloo2024) and at how feedback changes teachers’ behaviour or leads to improved student learning (Baecher et al., Reference Baecher, Copland and Mann2024). A limitation of the studies in this review is that all rely on teachers’ self-reports of uptake and have no empirical data detailing how feedback was carried out in teachers’ practice. It is important that research looks beyond teachers’ claims and includes more longitudinal research that investigates teaching after POF to look for evidence of feedback enactment and impact (Donaghue, Reference Donaghue2021; Donaghue & Heron, Reference Donaghue and Heron2025; Kanat Mutluoğlu & Balaman, Reference Kanat Mutluoğlu and Balaman2023). Student feedback might assist this research focus. Teachers and institutions often gather student evaluation of teaching, but it would be interesting to see how student input might help evaluate feedback enactment. Also connected to feedback enactment is the idea of feedback literacy which has garnered much attention in the student feedback and higher education literature. Heron et al. (Reference Clark-Gareca and Warkentin2024) propose a framework of post observation feedback literacy which would be interesting to test out empirically.

My third recommendation is for feedback researchers, scholars, and practitioners to embrace different modes of feedback. The studies in this review show clear affordances of video-mediated feedback, both for POF and for observer training. The study by Bates and Donaghue (Reference Donaghue2021) indicates that synchronous chat is also worth further investigation. As technology becomes central to how we do education in the post COVID-19 world, the profession is presented with an opportunity to change and enhance its POF practices. An inevitable change will be the use of AI in POF, an area yet to be investigated. An additional gap in the literature is the scarcity of studies looking at written feedback, which is surprising considering that this is commonplace, especially in pre-service training contexts.

My final recommendation is to extend a recently emerging strand of POF research which focuses on how support and empathy is (or can be) realised in feedback. The teachers surveyed and interviewed in the perceptions studies valued observers that were open-minded, empathic and flexible, and appreciated a teacher-centred approach to feedback. These views underscore the importance of relationships and creating a supportive environment for the observed teacher. The POF research on identity and face is relevant to this endeavour but much of this research focuses on face threat rather than face support. The research conducted by Bąk-Średnicka (Reference Bąk-Średnicka2024), Donaghue (2022), Lewis and Wagner (Reference Lewis and Wagner2023), and Wagner and Lewis (Reference Wagner and Lewis2021) provide a useful starting point for future research focusing on support and empathy in feedback.

Overall, it is interesting to see how feedback is practised and perceived across the world, and encouraging to see a rich body of research within the field of language education, indicating an interest in POF and a commitment to enhancing its educational affordances. These studies show that POF is highly situated and contextualised, influenced by the preceding observation, the purpose and outcome(s) of the observation process, the institutional and cultural setting, and the participants – their relationship, experience, confidence, expectations, and emotional states. This makes it impossible to offer ‘best practice’ for POF. Research instead presents points of reference and empirically based advice and strategies for readers to consider for their own particular context. There is, however, one recommendation which is relevant to all POF contexts and which transcends geographical, cultural, and contextual boundaries, and this message is repeated by POF researchers across the world: observer training is crucial. Helping observers become more aware of their practice and its consequences, and helping them recognise their feedback motives, intentions, goals, and expectations may result in maximising the learning potential of POF and enhancing this important but often difficult and complex route to language teacher learning.