Teacher and student teacher views of agency in feedback

ABSTRACT The present study compares the feedback themes that groups of teachers (n = 5) and student teachers (n = 15) discussed in a professional development programme concerning teachers’ classroom interaction and formative assessment, and the agents they assigned the feedback to. The results of the thematic analysis show little variation with the conversation themes between the groups, but they did show more with the appointed agents of feedback themes. The teachers assigned feedback themes to the teacher and the students, as student teachers assigned them to the teacher and the learning environment. By referring to feedback as the shared responsibility of the teacher and students, the teachers demonstrated more modern views of feedback than the student teachers. Ignoring students’ part limits the understanding of feedback as a shared process between teacher and student. Teacher education should ensure that attention is also paid to students’ role in the feedback process.


Introduction
Feedback is highly influential for learning, for both good and bad (Hattie and Timperley 2007).The current understanding of feedback has broadened from feedback as the provider's process to feedback as the receiver's process.The rationale behind the change is that no matter how good the feedback might be, it is not efficient if it is not accepted and used by the receiver.Recent studies have considered the recipient side by, for example, exploring students' responses to feedback (Winstone et al. 2017), students' ability to seek and utilise feedback (Carless and Boud 2018), teachers' perceptions of responsibility-sharing in the feedback process between teachers and students (Winstone, Pitt, and Nash 2020), and teachers' ability to facilitate the development of students' feedback seeking and utilisation (Carless and Winstone 2020).What is common to these studies is that they were conducted in the context of higher education.In primary and secondary education, discussion about student agency in feedback processes has not yet gained momentum, even though the 'assessment for learning' movement has advocated for the use of peer and self-assessment (Black and Wiliam 2009).Little is known about whom teachers consider to be agents of feedback in primary and secondary school and how their views might differ based on their experience.The present study explored primary and secondary teachers' and student teachers' discussions of feedback in professional development (PD) programmes concerning teachers' classroom interaction and formative assessment to determine who they believe possesses the agency of feedback and to compare their understanding of feedback.

Student agency
The notion of students' agency challenges the traditional views of learning and traditional educational settings in which teachers are expected to lead and students to follow (Lipponen and Kumpulainen 2011).Learning has become more student-centred, and students are encouraged to be agents in their own learning.Teachers have a significant role in promoting and constraining student agency (Lipponen and Kumpulainen 2011), as they are ultimately responsible for classroom practices.
Agency is the ability and possibility to influence one's life (Juutilainen, Metsäpelto, and Poikkeus 2018).However, the concept has various definitions and emphases.In psychology, agency is conceptualised primarily as an individual characteristic (Eteläpelto et al. 2013;Rajala 2016) that is closely connected to self-efficacy (Bandura 2001), while sociocultural approaches see agency, in different degrees, as influenced by sociocultural context.Earlier sociocultural approaches emphasise the degree of domination of society over the individual, while more recent ones see a two-way relationship (Eteläpelto et al. 2013;Rajala 2016).In this study, we take the latter stand and consider agency a dialectic interplay between the individual and the environment (Rajala 2016), meaning that the environment guides individuals' agency, but individuals also transform the environment.In groups, individuals develop patterns of agency that create expectations for the quality of their participation (Gresalfi et al. 2009).Hence, agency is not possessed but achieved through engagement (Biesta and Tedder 2007).
In schools, agency is co-produced by students and teachers in the school classroom environment (Charteris and Smardon 2018).The teacher, the students and the environment together create a setting that both supports and constrains participants' agency.The teachers' role is significant in the creation of agency.They can support students' agency by giving them choice and authority and by crediting their contributions, but they can also constrain it with an authoritarian teaching style (Lipponen and Kumpulainen 2011).Even minor choices in teachers' interactions influence students' agency (Clarke et al. 2016).Theoretically, students can cross the boundaries of the traditional teacherstudent relationship in any environment, but they are more capable of doing it when they are encouraged by the teacher (Lipponen and Kumpulainen 2011).

Feedback and agency
Feedback follows the general transformation of teaching and learning, as the understanding has broadened from feedback being a teacher's process to feedback becoming a student's process (Dawson et al. 2019;Molloy, Boud, and Henderson 2019;Winstone, Pitt, and Nash 2020).Traditionally, teachers were considered the sources of feedback, and students were left with the passive role of absorbers.The research focused on the qualities of effective feedback.The meta-analysis of Hattie and Timperley (2007) defined feedback as 'information provided by an agent (e.g.teacher, peer, book, parent, self, experience) regarding aspects of one's performance or understanding' (p.81).Hattie and Timperley's main ideas of effective feedback were that (1) effective feedback should answer three questions -where to go, how to go and where to go next -and (2) feedback can focus on four levels: task, process, self-regulation and person, which have diverse effects on learning.They claimed that teachers should consider the timing, form and focus of feedback to ensure the efficiency of feedback.Hence, they implied that feedback is mainly the teachers' responsibility.Sadler (1989) was ahead of his time, considering students as active participants in the feedback process and claiming that information could be called feedback only if learners used it.Teachers should therefore not only provide feedback but also nurture its reception (Boud and Molloy 2013).Lack of agency is one of the reasons for students' disregard of feedback, but their agency can be strengthened by sharing the responsibility of feedback between teacher and students (Winstone et al., 2017).To emphasise students' agency, feedback can be defined as the student process of seeking and using various forms of feedback to enhance their work or learning strategies (Carless and Boud 2018).The focus is on the receiver's capabilities, and teachers only facilitate the feedback process.Both tasks of feedback, providing and receiving, are the two sides of a coin and are equally valid.Ideally, the feedback is purposeful and offered in an appropriate way, and the receiver is willing and capable of using it.
Besides theorising who the agents of feedback are, it is important to know to whom the teachers, in different phases of their careers, attribute agency.The research on students as the agents of feedback has been conducted predominantly in higher education (Kleij 2019), and the research on primary and secondary education is scarce.A survey from secondary school showed that teachers and students were more positive about the quality of teacher feedback than about its facilitation (Kleij 2019), which implies that teachers were more proficient in their traditional role of providing feedback than in supporting students to participate in the feedback process.More positively, Brown, Harris, and Harnett (2012) claimed that primary and secondary school teachers were inclined to involve students in the provision and use of feedback.Nevertheless, the recipience of feedback was not their study's central concern.Despite the inclusion of agentic elements, the focus was on the function that teachers attribute to feedback -whether it was for learning or grading.
Even less is known about student teachers' views on the agents of feedback.Xu and He (2019) demonstrated that student teachers' conceptions about assessment develop during teaching practice, but, regarding feedback, the researchers considered feedback provided by student teachers.However, one of the student teachers in their study developed an understanding of the importance of student engagement in assessment.Despite this orientation, with feedback, his focus was on teacher ability to share quality feedback.Lutovac and Flores (2021) considered agency in assessment, as they introduced student teacher conceptions indirectly by examining their narratives of failure.The student teachers considered that teachers have the majority of responsibility for a student's failure in assessment, even though they acknowledged that students must do their part, especially in the form of making effort.
Teachers' conceptions make a difference (Lutovac and Flores 2021).First, conceptions influence their assessment practices (Xu and Brown 2016).Feedback is an essential part of teachers' work and classroom interaction (Hamre et al. 2013), and teachers' conceptions of feedback are reflected in how they carry out that interaction.Second, teachers' conceptions are related to practice and learning in two ways: conceptions can filter teachers' learning and prevent them from putting new knowledge into practice, and practical experiences can also induce changes in teachers' conceptions (Lutovac & Flores, 2022;Xu and Brown 2016).Teachers' and student teachers' conceptions of feedback and agency inform adjustments in teacher training.A comparison of these conceptions can uncover whether such conceptions are deep-seated, that is, whether years of practice change them.

The aim of the study
It has been shown that the important life skills of utilising feedback can be practiced long before beginning higher education (Ketonen, Nieminen, and Hähkiöniemi 2020), and teachers can nurture these skills by sharing responsibility for the feedback process with students (Winstone et al., 2020).However, little is known about whom teachers and student teachers consider agents within the feedback process.As they can support or constrain student agency, their views require attention.Therefore, the aim of this study was to learn about teachers' and student teachers' understandings of feedback.The research questions were as follows: RQ1 Which themes do teachers and student teachers discuss while reflecting on the issue of feedback?
RQ2 Whom do the teachers and student teachers consider to be the agents of feedback processes?
RQ3 How do teachers' and student teachers' views differ?

Participants
This study explored a group of teacher education students (n = 15, age range 20-27 years) and a group of in-service teachers (n = 5, age range 40-50 years) participating in a professional development (PD) programme on teachers' classroom interaction and formative assessment.After being informed about the research, the participants gave their written consent to participate in the study.The teacher education students studied in a Finnish university.The PD programme was a part of student teachers' pre-service programme as a module of preschool and early primary school education (25 credits) in their second, third, fourth or fifth year of optional studies in education.During the time when the theme of the feedback was targeted in PD, they had studied together intensively for the previous four months and knew each other well.The in-service teachers' group of two elementary teachers and three subject teachers was from one urban school in Central Finland.Two subject teachers worked with lowersecondary students, and one with both primary and lower-secondary students.The teachers had over 10 years of work experience as a teacher (avg.17 years).Because they worked as colleagues for several years, the teachers knew each other, and the atmosphere in the group during the programme was relaxed and straightforward.Given that the teachers voluntarily joined the PD, which took place after school hours and was not rewarded, the teachers can be described as particularly motivated.

PD programme of classroom interaction and formative assessment
The professional development programme was a part of a project that focused on developing teachers' classroom interaction and formative assessment during one school year.The PD programme was built on the Teaching Through Interactions (TTI) framework (Hafen et al. 2015;Hamre et al. 2013), which focuses on teacher-student interaction in the classroom.The Classroom Assessment Scoring System -Secondary (CLASS-S; Pianta, Hamre, and Mintz 2012) observation manual is operationalised within the TTI framework.The CLASS-S focuses on effective teacher-student interactions in three main domains: emotional support, classroom organisation and instructional support.Instructional support focuses on teacher-student interactions that foster students' learning and participation through versatile strategies and well-targeted feedback (Hafen et al. 2015).Feedback includes such dimensions as scaffolding (guiding learning, prompts), feedback loops (feedback chains, follow-up questions), and building on students' responses (clarification, extension, focused feedback).To strengthen the connection with previous feedback research, supplementary elements were brought into the feedback theme, though the focus on interaction was maintained.The new dimensions were focusing feedback on participation, effort, process and learning progression (Hattie and Timperley 2007;Mueller and Dweck 1998) and supporting students' self-regulation (Clark 2012).
In the PD programme that was inspired by the MyTeachingPartner intervention (Pianta et al. 2008), both participant followed the same structure: theory, implementation and reflection.Each cycle focused on specific dimensions of classroom interaction: teacher sensitivity, feedback, dialogicity and classroom management.The teachers' programme comprised three cycles led by the first researcher, and student teachers' programme comprised four cycles led by the third researcher (Figure 1).
During the first meetings (90 minutes each), new themes were introduced and discussed with a criteria-based rubric of the theme.Before the second meetings (90 minutes each), the participants videotaped their own lesson.For videotaping, student teachers chose a lesson they taught as part of their teaching practice and in-service teachers chose any lesson they wished.For in-service teachers, the researcher chose examples from videotapes of teaching practices relating to the theme.The student teachers chose the video examples themselves.In the second meetings, the chosen video clips were watched together, reflected on and discussed in the group.The earlier presented rubric guided the video observations and provision of peer feedback.The video extracts were chosen to present good examples of teaching practise, and correspondingly, the participants were guided to notice the best moments and examples concerning the dimension under discussion.However, in both groups, the conversations entailed multifaceted conversations of the dimension, including aspects that the participants experienced as challenging.

Data collection and analysis
Teachers' meetings were audio-recorded.Student teachers' meetings were video recorded; since the number of participants was large, individual speakers could be more easily identified on video compared to audio alone.
The data were analysed using a data-driven thematic analysis (Braun and Clarke 2006), which is a way to describe and interpret the patterns in the data -both the qualitative differences and similarities.The analysis began by becoming familiarised with the data and continued with transcriptions of the conversations.Topics of conversation were examined in the transcripts and codes describing the conversation themes were attached to data units.Subsequent comments on the same topic were considered a unit of analysis, which reached from one comment to several participants' commenting on the topic.In some cases, units contained more than one theme.Next, the codes were explored and sorted into categories.The data were recoded with category names, and the categorisation was then further explored and adjusted.In this phase, the codes relating to organisational issues, such as placement of the camera in the classroom, were left out.During the next step, creating the thematic mind-map (Braun and Clarke 2006), the categories were related to three meta-themes: the teacher, the students and the environment.The metathemes emerged from the data, but they stem from the notion of agency as being coproduced by students and teachers in the school environment (Charteris and Smardon 2018).The division of categories under meta-themes was not unequivocal, since some categories were appointed to different meta-themes in different data units.Therefore, the data units were coded with the categories and meta-themes.The coding continued until it did not produce any changes.At the end of the analysis, there were 24 conversation themes that each related to one or more meta-themes.
The reliability of the analysis was examined in two parts.Near the end of the category creation, the first author tested categories in peer negotiation with the second author, who had not participated in data collection and was therefore more objective in judgment.After the second author tested coding the data, commented on the categories and discussed divergent views with the first author, categories were clarified.Later, peer coding was used to test the coding of meta-themes.The second author coded 20 data extracts relating to meta-themes (teacher, student, environment, unclear) with 75% agreement with the first author.A number of the disagreements appeared to be due to insufficient background information and were solved by telling the third author more about the video clips that the participants referred to.The rest of the differences were discussed until the researchers reached a sufficient consensus.As the last step of analysis, the results were presented to the in-service teacher group participating in the PD programme and discussed.The rationale for member checking (Onwuegbuzie and Leech 2007) was to receive feedback and hear supplementary ideas about the interpretations and conclusions.

Results
First, we studied which themes teachers and student teachers discuss while reflecting on the issue of feedback.Second, we identified to whom the participants assigned agency within these themes.Third, we examined the differences between teachers' and student teachers' views regarding feedback.The findings are introduced in the following subsections.

Discussion themes
The feedback cycle of the PD programme engendered conversation about 24 themes that describe the scope of discussion (see Table 1).Most of the themes, 19, were common between the in-service teachers and student teachers.In addition, the student teachers raised one theme and the teachers raised four.

Discussion themes' attributions to meta-themes
Themes' attributions to meta-themes are presented in Figure 2 (student teachers) and Figure 3 (teachers).When participants discussed the themes, they explicitly or implicitly referred to them as deriving from the teacher, student or environment.For example, the teacher's emotional expression was in both groups referred to as the teacher's territory and not related to the students or the environment.In a few cases, the references remained unclear -for example, when a participant made a single comment and talked in passive.'Good climate' was a theme that student teachers noticed but did not attribute to teacher, student or environment; hence, it is not attributed to any meta-theme (Figure 2).

Differences in teachers' and student teachers' views
Both groups connected several themes to the teacher, such as provision of corrective feedback, differentiation of learning and clear communication, but there was also some variance, especially with theme's connections to the meta-themes.The common themes with teachers and student teachers are drawn with a solid line and the groups' own themes are drawn with dashes.One must not make conclusions about single themes and their attributions -concluding, for example, that The teacher elicits information from students for summative assessment or to adapt teaching.

Emotional expression
Teacher's emotional expression and teacher's calmness Good climate The atmosphere in the class is (or should be) good.The atmosphere is more permanent than a 'peaceful moment'.

Learning from mistakes
The potential of mistakes on learning Misbehaviour Student behaviour is something other than the teacher would hope it to be.teachers consider the teacher's concentration in the classroom as a student responsibility (Figure 3).Instead, the pictures as a whole illustrate the discussions regarding feedback.Two divergent trends are evident in teachers' and student teachers' theme maps.First, the students referred only one theme to the students (one case), while teachers referred multiple themes to students (altogether 16 cases).Second, compared to the in-service teachers, the student teachers referred more themes to circumstances.Next, examples introducing these trends are presented with two themes.The first concerns agents of a positive phenomenon with an example of dialogue, and the second concerns agents of a negative phenomenon, misbehaviour.
Student teachers attributed dialogue not to the students but to the teacher and the environment.The teachers saw dialogue as teachers' and students' common territory.In the following excerpt, the teacher described the dynamics in one of her student groups, where the majority of students refrained from participating in classroom discussion: Teacher 2: They just stay silent.Like we discussed the chemistry grades one-on-one and with everyone, I tried to encourage . . .It's like you cannot punish for being shy if it feels awkward to talk in front of the class, but I mentioned that one could try.Especially in the moments that I know they understand and no one raises a hand and no one saves the teacher, even though we could proceed if someone would raise a hand and say it.So, in those moments they could, if it needs only one word or so to answer.(Teacher: dialogicity, student) In this piece, the teacher considered herself responsible for classroom discussion, but also considered students to be active participants.She even expressed disappointment in students when no one bothered to help her with the dialogue.The next piece concerned the same teacher in another video extract in which a student held a dialogue with the teacher.Two other teachers gave feedback on that piece: Teacher 1: Well, with that guy you get far.Like . . .you two bene-benefit all the other group.They follow you two to . . .Teacher 3: . . . to discuss.
Teacher 1: It probably gives much, but let's say that I know the guy and he is capable of many types of language, but in that moment there, [he is] infinitely important helper for the teacher.(Teacher: supporting learning, teacher & students) If teachers had considered students responsible for dialogue failure, they also credited them for successful dialogue.The teacher's and student's discussion was seen to benefit the learning of the whole group.The student was considered a crucial element to maintain the dialogue.
The student teachers (ST) related learning and dialogue to the teacher and circumstances.In the following piece, a student teacher discusses a video excerpt of another student teacher: ST 11: And just like through interaction, as it arises there [on the video], maybe through it . . .with it you can insert some feedback.So it would be like 'very good', or it would be direct, but without the student noticing, through the dialogue, the message will be understood.(Student teacher: dialogue, teacher) When talking about dialogue, the student teacher referred to the teacher as responsible for transmitting the message.She mentioned that the message may be transmitted 'without the student noticing', which implies that the student is not the agent, but more of an object of the dialogue.The student teachers also considered dialogue to be affected by the environment: The student teacher's reflected the contradictory demands of student-centred teaching and carrying out the lesson according to the pre-made plan in their teaching practice.In student teacher's video extract, students and the student teachers joined the conversation about foxes.The student teacher stated that the environment kept her from continuing the dialogue the way she wanted.
Both participant groups discussed the challenges of the feedback process.Similar to dialogue, student teachers attributed students' misbehaviour to the teacher and the environment.The following comment relates to the episode on videotape of student teacher 9, where she stops the activity on the music gymnastic lesson and gives further instructions.While giving instruction, one student performs a cartwheel, and the student teacher does not react to it.I must have been so concentrated in giving good feedback and instructions that I couldn't interrupt her that 'hey, you, stop now'.(Student teacher: misbehaviour, teacher) The student teacher criticised herself for not reacting to student's cartwheels and stopping her.She did not blame the student for being restless or disobedient.Hence, she considered herself, not the student, responsible for the student's behaviour.Student teachers also attributed the misbehaviour to the environment: ST 12: Or when they [students] should tell about themselves.I think that is hard for them, and maybe I was alarmed by their reaction, like the intensity of the reaction.Like almost every one of these children can do and concentrate and like this, but in an insuperable situation like that, so how come they react so strongly or get all disorganised.They can't even listen to the instructions they are given.(Student teacher: misbehaviour, environment) This piece of conversation concerned child-initiated learning that student teachers had experienced as challenging to implement, since the students were not accustomed to such a policy.The student teacher did not blame the students for misbehaviour, but considered the situation too difficult and unfamiliar to them.Hence, they attributed the misbehaviour to the environment.Conversely, the teachers attributed the misbehaviour only to the students: Teacher 4: That is quite common for some [students], I also have a handful of students that do that [destroy the classroom conversation].I don't know how far it is intentional destruction or don't they just . . .are they so much in their own bubble that . . .often a student sees his prerogative to ask in the middle of other students' sentence, 'Hey, what have we for lunch today'?(Teacher: misbehaviour, student) Here, the teacher analysed and tried to understand the reasons behind students' behaviour, but still attributed the destruction of discussion to students.

Discussion
This study explored in-service teachers' and student teachers' discussions during reflection on video extracts in a PD programme in order to find out who they considered to be the active agents of the feedback process.The majority of the identified discussion themes of the two groups overlapped, meaning that the teachers and the student teachers considered the same aspects of feedback central or interesting.The main difference between the groups' discussions was in the metathemes that they attributed the themes to: the teacher, the student and the environment.With both groups, the themes were most often seen as teacher's territory, which was natural, since the PD programme focused on the teachers' classroom interaction and practices.The first difference between the groups was that the student teachers referred more often to environment as a factor that affected feedback processes.The predominant difference was that while the teachers considered students responsible or partly responsible for several feedback themes -for example, supporting learning and working, dialogue, or the culture of the classroom -the student teachers only once attributed a theme to students.Thus, on the discourse level, the teachers saw students as agents of feedback, which implies willingness to share agency with students.Feedback is a twofold topic entailing the perspectives of providing and receiving feedback.Teachers should consider their own part by ensuring the provision of quality feedback in the classroom (Black and Wiliam 2009) and the students' part by facilitating the reception of feedback, that is, advancing the attitudes and capabilities that enable utilisation of feedback (Carless and Boud 2018).
There are several reasons that may explain the difference between teachers' and student teachers' orientations.First, teachers early in their careers tend to concentrate more on themselves and gradually turn their focus towards students' problems (Conway and Clark 2003).During their studies, they are encouraged to reflect on their own performance.In this intervention, the focus on the teacher's performance was natural, since the observation rubric concentrated on the teacher's actions.However, the rubric proved not to be restrictive; the teachers referred frequently to the students.
Second, the teachers were more familiar with their students and therefore more inclined to consider their perspective.In field training, student teachers did not have knowledge equal to that of the teachers and did not have equal responsibility for the students.Given that teachers, especially class teachers, spend considerable time in their daily lives with their students, the relationships become close.The teachers referred to the students almost as colleagues and expressed disappointment with their misbehaviour and lack of interaction.For a developing student teacher, assigning feedback agency to students requires balancing their own and students' responsibilities.Sharing agency with students may appear almost oxymoronic: How to perform the responsibility of handing out responsibility -especially where one is not ultimately responsible.Similar conceptions about student teachers' responsibility-sharing, where teachers' responsibility is emphasised, have been reported with assessment in general (Lutovac & Flores, 2022).However, feedback is not efficient without students' active participation (Nash and Winstone 2017).If in the reflections on feedback students are not considered coparticipants, the receiver's side and students' potential as providers of feedback are missed.
Third, it is possible that the culture of the student teacher group did not support criticism of the students but was considered defensive.A similar tendency was seen in student teachers' written reflections (Lutovac & Flores, 2022).Defensiveness can become a barrier to teacher reflection (Jaeger 2013), and since the purpose of this study's programme -and of teacher education as a whole -was to develop as a teacher, attributing the outcomes to students might seem to hinder professional development.Avoiding defensiveness is reasonable (Carless and Boud 2018), but it should not preclude the view of students as active participants or acknowledgement that not everything is in the teacher's hands.The teachers of this study continuously brought up the capricious nature of their work.As sharing concerns with colleagues helps teachers to deal with stress (Lewis 1999), such discussion should already be practised in teacher education.
Student teachers' various references to the environment derived from the different positions of teacher and teacher trainee.In field training, student teachers did not have as independent a position as teachers had.They could not, for example, abandon their plans and continue with them the next day, since the next lesson would be given by someone else.Student teachers also had less influence on classroom culture and policy than the teachers, who worked daily with the class.
The results have implications for teacher training.The fact that experienced teachers had more developed views of students as agents of feedback suggests that traditional, teacher-centred conceptions of feedback are not necessarily deep-seated.When planning and implementing assessment, teachers make compromises between the tensions they face between practice, conceptions, and knowledge (Xu and Brown 2016).These tensions can induce changes in any of these elements, which can result in positive development.Traditional conceptions of feedback not being deep-seated means that they develop with practice rather than prevent teachers from developing their practice.In such cases, the topic of students' feedback agency could be left mostly for in-service training.A related topic, student agency in assessment through the use of peer and self-assessment, could be easier for student teachers to process (Xu and He 2019).Such prioritisation is important for pre-service training, since teachers' professional skills are abundant and there is no time to cover everything.Nonetheless, in the planning of in-service training, it is valuable to identify themes that are not well digested during pre-service training and need further attention.
This study acknowledges its limitations.First, the order of dimensions of teachers' and student teachers' PD programmes was different (Figure 1).Even though both groups had discussions relating to previous and future dimensions, the previous ones may have influenced the conversations about feedback.Second, the group sizes were different, potentially making student teachers' discussions more formal due to the larger group size.Third, the in-service teachers were not a random selection but were experienced and especially motivated.Hence, their discussions do not necessarily represent the general situation of all teachers.Finally, the results must not be interpreted to signal that all in-service teachers see students as agents in feedback processes.Rather, they imply that development-oriented teachers do not have insuperable obstacles to reaching such a view.

Conclusions
This study examined teachers' and student teachers' views of student agency in feedback processes.The findings show that experienced primary and secondary teachers discussed feedback as a shared responsibility between student and teacher, but student teachers referred to feedback as only the teacher's responsibility.The findings have two main implications.First, when student teachers reflect on their performance in teaching practice, they should be encouraged to acknowledge students' roles and responsibilities.Such discourse lays a basis for understanding student agency and may relieve stress by messaging that not everything in the classroom is teacher's fault or achievement.Second, the results show that students' feedback agency is a challenging topic for student teachers.This does not necessarily mean that they should spend more time reflecting on it, however, because a comparison of teachers' and student teachers' views suggests that teacher-centred feedback conceptions are not deep-seated; they can change with experience.Therefore, during pre-service training, it might be more efficient to place emphasis on other qualities of feedback, such as its objective of supporting learning.Agency in feedback processes could be an advanced topic that is targeted later in in-service training.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Figure 1 .
Figure 1.The themes and timing of professional development programme.
Motivation Students' motivation to learn Peaceful moment A certain moment is (or should be) peaceful Reacting to wrong answer Teachers' reactions to wrong answers Rules of participation The limits and requirements of participation -what is allowed and expected in the classroom Scarce feedback The lack or insufficiency of feedback Specific feedback* The specificity of the provided feedback Specific student Providing feedback for a particular student Supporting learning* Supporting and guiding learning: observations, experiences, tools, challenges, possibilities, etc. Supporting working* Supporting working with observations, experiences, tools, challenges, possibilities, etc. Teacher's concentration Teacher's concentration in the classroom Teacher's plans Keeping, adapting or changing of teacher's plans * The themes that were explicitly mentioned in the observation rubric.

Figure 2 .
Figure 2. Theme map of student teachers' discussion themes and references.

Figure 3 .
Figure 3. Theme map of teachers' discussion themes and references.

ST 12 :
If we have the lesson planned and limited time, and we cannot carry it on to the next lesson . . .Versus if you'd be a class teacher there.Like how long can we stay pondering foxes' colour and size, or placement of clefs, or the multiplication sign?It is hard, hard to add or to lead to such [open discussion], if there's no more than a short moment.(Student teacher: dialogue, environment) ST 9: [While watching my video,] I wondered what has gone in my mind.

Table 1 .
The discussion themes and their explanations.