Adapting feedback to the demands of teaching in diverse classrooms. Novel feedback concepts for preservice teachers during practicum

ABSTRACT During their practical training, feedback for preservice teachers should address the demands of diverse classrooms. Using the critical incident technique within a participatory research framework, this study investigated whether and how mentors adjust their feedback in this regard. Based on a sample of 33 preservice teachers and 46 mentor teachers in German primary and lower secondary education, the results indicate that although pupils’ differing prerequisites clearly presented challenges for preservice teachers, only part of the mentors referred in their feedback to these demands. Mentors who adapted their feedback referred to well-known methods for addressing differences in pupil performance and background and suggested specific novel strategies by specifying what is possible and expedient (or not) when pupils’ prerequisites and backgrounds differ. These methods were found to enhance preservice teachers’ self-efficacy and competencies in teaching diverse classes, and mutual diversity-oriented feedback also benefited mentors. In conclusion, the study identifies implications for feedback development.


Introduction
There is clear evidence that targeted feedback is a key element in the training and professionalisation of preservice teachers (e.g.Cohen, Hoz, and Kaplan 2013;Ellis, Alonzo, and Nguyen 2020;Lawson et al. 2015).However, existing research on feedback during practical training assumes an initial situation that no longer reflects the realities of everyday school life.In particular, migration and issues of inclusion have transformed individual schools and the wider school system.As the student population becomes increasingly diverse (Dixon et al. 2014), the demands on teachers also change, and this clearly has implications for the education of future teachers (Jensen and Rasmussen 2011).
In Germany, the location of the present study, most future teachers encounter diversity for the first time during their first field experiences as both a characteristic of the classroom and a demand on their teaching and classroom management.However, feedback doesn't automatically take account of diversity in classrooms despite the fact that it is the norm.Existing research has not investigated whether or how mentor teachers have incorporated the feedback they provide to meet this need.To bridge this gap, the present study explored feedback during preservice teachers' practical training in diverse classrooms in more depth.
To explore these concerns, we used Flanagan's critical incident technique within a participatory research framework.Sometimes referred to as collaborative action research (Bergold and Thomas 2012;Vaughn and Jacquez 2020), this approach captures the knowledge and perceptions of agents of change -in this case, mentor teachers and their preservice mentees.In particular, we sought to investigate whether mentor teachers provided feedback addressing the demands and characteristics of diverse classrooms.We also explored how mentors and mentees contributed to the development of novel feedback concepts and how this might guide the further development of approaches to feedback.

The changing classroom
Current migration necessitates greater inclusion and means pressure for action in education systems and individual schools.As of January 2021, the EU population included 23.7 million third-country nationals, as well as a further 37.5 million residents who were born outside the EU (Eurostat 2021a).In Germany 23.8 million of the country's inhabitants were migrants; as defined by the Federal Statistical Office (2022) of Germany, this means that they themselves or at least one of their parents were not born with German citizenship.The most common countries of origin are Turkey, Poland, Russian Federation, Romania, Kazakhstan and Syria.Migrant groups differ widely in terms of language skills depending above all on the lengths of time living in Germany and also on the country of origin and gender (for details see the data report by Metzing 2021).
Migration has had far-reaching effects at the local and national level.In contrast to an EU average of about 30%, the number of minors migrating to Germany is much higher at 54% (Eurostat 2021b) and, on average, 31% of pupils in a German classroom come from a migrant background.Differences in language of origin and competence in the language of instruction contribute significantly to classroom diversity (Busse et al. 2019); in primary schools in large cities pupils speak more than two hundred different language(s) of origin (Mehmedbegović et al. 2015).Overall, the highly diverse pupil population differs in performance, academic readiness, learning preferences and styles, motivation as well as social, methodological, and self-regulatory competencies (Hardy, Decristan, and Klieme 2019).Pupils with additional needs as well as with different social background further increase the diversity of learning prerequisites.
Although the issue of classroom diversity is not new, research has not yet addressed the implications for in the context of practical training.In particular, researchers have only explored the feedback provided to preservice teachers in very general terms on lessons and classroom management attempts.Research has stated some changes with respect to mentoring and feedback, however, changes tend to relate to the role patterns that underly the feedback process.

Changes in mentoring during the school practicum
In Germany, some teacher training programmes are divided into a bachelor's and master's degree programme at the university level, some other programmes finish with a state examination (KMK 2020).All programmes in Germany extent over two phases: the first one is university-based with a focus on a theoretical education, the second phase (18-24 months, depending on federal state regulations) starts after graduation and takes place at schools and teacher training institutes with a focus on practice.During the first phase, future teachers also have to complete a practicum.In Bavaria, the German state where the present study is located, this practicum has to be completed either in full-time, over several months, or over one year with block phases and fixed days in the week.
During school practicum, mentoring can be defined as 'a strategy to assist pre-service teachers in learning to teach' (Ambrosetti, Knight, and Dekkers 2014, 224).There is consensus that preservice teachers should be supported and supervised during their practical training by mentors.However, mentoring concepts have changed over time.In earlier apprenticeship-oriented models of mentoring, feedback was conceived as onesided guidance or advice, and the mentee learned through modelling and observation (Wang and Odell 2002;Zeichner 1987).Observational learning must not be negative per se.Preservice teachers value the (technical) guidance of their mentor teachers and can use the observed practice to critically-reflectively deal with it.Nevertheless, this could carry the risk that the observation of traditional patterns of teaching will be uncritically adopted (Hobson and Malderez 2013).
Constructivist mentoring styles that focus on cooperation and co-construction between mentor and mentee show promise as a means of achieving these goals (e.g.Canipe and Gunckel 2020;Richter et al. 2013).By partly dissolving asymmetric roles between preservice teachers and mentor teachers, constructivist approaches support more egalitarian interactions -not in the sense of equality of status but in creating opportunities for mentor and mentee to learn from and with each other and to construct shared understandings while working together on teaching-related activities (Canipe and Gunckel 2020;Gallo-Fox and Scantlebury 2016).

Feedback as a key element of professionalisation
How feedback is shaped depends on the framework of teacher education.Different approaches in teacher education, for example, a master-apprentice approach, a teacher identity approach, a competence approach and a reflective approach result in different implications for curriculum design (Robinson and Mogliacci 2019).As described above with a view to the master-apprentice approach, feedback is conceived as modellingoriented and one-sided guidance or advice by the mentors; mentees have few opportunities for reflection or experimentation.In contrast, reflective teacher education approaches create leeway for teachers to construct personal theories in and from practice (Robinson and Mogliacci 2019) which effects on feedback; for example, mentor teachers can inspire preservice teachers to question their own concepts of learning and teaching; this is further described below.
Feedback is generally regarded as an important tool for mentor teachers (e.g.Bjørndal 2020; Clarke, Triggs, and Nielsen 2014;Orland-Barak 2016).Hattie and Timperley's wellestablished model defines feedback as 'information provided by an agent [. ..] regarding aspects of one's performance or understanding' (Hattie and Timperley 2007, 81), which is generally considered 'necessary to improve on a task' (Wisniewski, Zierer, and Hattie 2020, 12).By reducing the gap between actual and target states of performance ('feed-up'), feedback links existing achievements to independent criteria ('feed-back') and provides information about next steps for achieving specified targets ('feed-forward').Effective feedback answers the questions 'Where am I going?How am I going?Where to next?' (Hattie and Timperley 2007, 87) at four distinct levels: task, process, self-regulation and self.Ellis and Loughland (2017) found that mentors are more likely to focus on process, asking questions like 'How am I going?' Focusing on task-rather than self-related issues can help to improve learning and mentees' professional development (Wisniewski, Zierer, and Hattie 2020).Constructivist approaches to feedback address the mentee's scope of action and reflection capability (Hobson et al. 2009).By using feedback to encourage mentees to reflect on their behaviour and monitor their goals (Crasborn et al. 2008), mentor teachers can inspire preservice teachers to question their own concepts of learning and teaching and to identify inhibitions that hinder learning (Hobson et al. 2009).Offering mentees leeway for experimentation and taking risks encourages trial and error (Izadinia 2016;Wisniewski, Zierer, and Hattie 2020).Feedback supports mentees to develop (effective) solutions and to test them in different contexts.This more egalitarian form of feedback can help to achieve the desired balance between support and challenge (Orland-Barak 2016); when feedback is shaped through two-way interaction, 'the mentor and mentee would work together as a team in order to develop the needs of both' (Ambrosetti, Knight, and Dekkers 2014, 233).Mentors can also benefit, for example, by further developing their practice in accordance with the knowledge and methods preservice teachers bring from the university (Talbot, Denny, and Henderson 2018) and their feedback can include how mentors use such new impulses in their practice (Hobson et al. 2009).
To summarise, the existing body of research confirms the advantages of constructivist mentoring and mutual feedback.However, it is not yet known whether and how mentors' feedback addresses the issues of lesson content and classroom management in an increasingly diverse education environment.Therefore, the present study will this gap in the research.

Research questions
To investigate whether and how mentor teachers address the demands of diverse classrooms, we used the critical incidents technique (CIT, Flanagan 1954; see also Viergever 2019;Watkins et al. 2022) within a participatory or collaborative action research framework (Bergold and Thomas 2012;Vaughn and Jacquez 2020).The revealed examples reflected significant patterns in how preservice teachers and mentor teachers handle situations on feedback and the consequences that arise from them.The participatory approach ensured that mentors and mentees were equally involved as 'subject-matter experts' with expertise in the respective field.The study addressed the following research questions.
(1) Which situations reported by mentor teachers and preservice teachers serve as points of departure for feedback in diverse classrooms?(2) What feedback approaches do mentor teachers use in diverse classrooms, and which demands are addressed?(3) Is feedback implemented as one-sided or a mutual process?(4) What are the consequences of feedback for the preservice teachers and mentors involved?

Research context
The present study forms part of the project Mentor Teacher-Mentee Relationship in Practical Training in Heterogeneous Classes (Memphis plus), which is run by the Department of Education and Rehabilitation at Ludwig-Maximilians University in Munich in Germany.That project is itself part of the Qualitätsoffensive Lehrerbildung, a joint initiative of the Federal Government and the Länder to improve the quality of teacher training.The project is funded by the Federal Ministry of Education and Research which is the highest authority of all areas and institutions of education in Germany (located in Berlin).The project's higher-level goal is to provide an evidence-based account of how mentor teachers collaborate with preservice teachers during their practical training.For the purpose of this research, the chosen research setting was the school practicum, focusing in particular on feedback in diverse classroom settings.

Research procedure: participatory research, critical incident technique, focus group discussions
To co-construct knowledge and strategies to address professional and everyday challenges, we employed a participatory research framework to capture the perspectives of the researchers and practitioners (preservice teachers and their mentors) (Bergold and Thomas 2012;Vaughn and Jacquez 2020).As that knowledge is generated through collective decision-making, we used CIT (Flanagan 1954) to identify important feedback situations as experienced by preservice teachers and their mentors during practical training in diverse settings.CIT is an established method for collecting and analysing empirical data on behaviour in real-world professional situations characterised by high demands, high levels of responsibility, complex problem solving requirements and high levels of social interaction (Watkins et al. 2022).Detailed analysis of critical incidents enables researchers to identify similarities, differences and patterns in a given context and to gain an insight into how and why the involved actors behave as they do.The overall goal is to identify behaviours and mechanisms that contribute to success.As used here, the term critical does not mean 'negative' or 'difficult' but refers to key incidents that determine outcomes in a given context.A critical incident report includes a detailed description of the incident and relevant causes, antecedents and outcomes.
Based on Flanagan's specification, focus group discussions were conducted to identify critical incidents in the present context as the basis for an empirical analysis of relevant social subsystems, collective phenomena and supra-individual behaviours (Barbour 2018).These discussions are not the source of collective patterns but facilitate the articulation of existing patterns; statements that identified a collective pattern were deemed valid for inclusion.The mutually stimulating conversation during group discussions typically generates more opinions than purely quantitative forms of inquiry, as participants engage with each other's responses and ideas within a developing dynamic (Krueger and Casey 2020).

Sampling
CIT assumes that participants are 'subject-matter experts' (Heine et al. 2016) with experience in their respective fields.Our sample included 46 mentor teachers (TOR) (42 females, 4 males); of these, 31 worked in primary education (with children from 6 to 10 in Germany), and 15 worked in lower secondary education (with children/youth from 10 to 16).The sample also included 33 preservice teachers (TEE) (29 females, 4 males) who had recently completed a one-year practicum involving block phases and fixed days across one school year.Of these, 23 worked in primary education, and 10 worked in lower secondary education.The prerequisite for participation was that the practicum took place in classrooms which were markedly diverse.Such classrooms differ in performance, academic readiness, learning preferences, motivation and also in linguistic, social, methodological, and self-regulatory competencies (Hardy, Decristan, and Klieme 2019).
To initiate recruitment, the project was publicised by an announcement to the department responsible for organising the practicum; that announcement was then shared with mentor teachers and preservice teachers.Both groups were invited to take part in a focus group discussion at the Ludwig-Maximilians University Munich during the semester.During the selection process, care was taken to include participants from different schools and geographical areas to avoid regional agglomeration (Krueger and Casey 2020).The participants were divided into 21 small discussion groups (typically four per group): twelve with mentors and nine with preservice teachers (Barbour 2018;Krueger and Casey 2020).Mentor teachers' and mentor teachers' discussions took place in different locations to avoid that mentees fear the exposure to the mentors.Each group was led by an experienced moderator who was currently active in teacher education, and moderators were trained using a guideline.The guideline contained methodical suggestions on how to structure the focus group discussion and included support on developing critical incidents (e.g. which elements are comprised in a critical incident).The focus group discussions lasted two hours.The course of discussion was structured thematically around the collection of participants' experiences of critical incidents, and these were documented.

Analysis and validation
The group discussions yielded a total of 117 critical incidents with 404 codings (245 by mentor teachers [C TOR ] and 159 by mentees [C TEE ]) of relevance to the present study.According to Flanagan (1954), critical incidents should be analysed by creating categories and subcategories.The goal is to reduce the material to its fundamental content, using abstraction to produce a straightforward corpus that adequately captures the original meaning.Using the MAXQDA programme (MAXQDA 2018), we employed qualitative content analysis (Elo et al. 2014;Mayring 2022) to create relevant categories.For the purposes of analysis, each category was listed and described, using explanations and examples and identifying overlaps and distinctions between categories.More precisely, the procedure involved the following four steps.
(1) Through intensive reading, incidents were assigned to the main categories.
(2) These main categories were further refined by creating subcategories.
(3) All categories were thematically restructured, summarised or renamed, and incidents were assigned accordingly.(4) Using the MAXQDA summary grid, thematic summaries of coded passages and critical incidents were created to build systematic case overviews for each group (Rädiker and Kuckartz 2020).To compare patterns and relations across categories, frequencies and distributions were processed in tabular form using the code-matrix browser.
The procedure could be explained by the following example in Figure 1.There needs to be an exact analysis to distinguish when 'reflection' was a one-sided or a shared feedback procedure.
The category system was reviewed by a second researcher; in cases of disagreement between raters, the relevant critical incidents were revisited for communicative validation (Kvale 1995).Revisiting the categories ensured greater selectivity and higher inter-rater reliability, and the achieved coefficient of .93;percentage coefficient of .80 are considered satisfactory (see Belur et al. 2021).Additionally, the categories achieved a Cohen's kappa coefficient κ (Brennan and Prediger 1981) of .94;according to Landis and Koch (1977), a κ value greater than .80 is considered 'almost perfect agreement'.

Feedback model
As shown in Figure 2, the observed approaches to feedback during the practicum in diverse classrooms followed two distinct paths or courses.The two courses differ in terms of being one-sided or mutual, method, content, reference to the context and benefits for mentors and mentees.
The following sections trace commonalities and differences in mentors' and mentees' perspectives on feedback, supported by selected excerpts from critical incidents to illustrate feedback processes.

Path 1: one-sided feedback: partial references to the diverse classroom
Figure 3 shows the first path, the first approach to feedback, in more detail.All categories on preservice teachers' and mentor teachers' activities, gains and the number of codings of each category are included.

Preservice teachers' and mentor teachers' activities in diverse classrooms
Within the one-sided approach to feedback, the situation that most frequently initiated feedback was preservice teachers' Attempts at Differentiated Teaching (39 C TOR , 22 C TEE ) as illustrated in the following examples.All 61 incidents on attempts at teaching differed in terms of whether and to what extent preservice teachers noticed pupils' different prerequisites and were able to adapt their lesson and language level to meet pupils' needs: 'My mentee had great difficulty in adapting her language level for a first class with diverging performance levels and some cases of reduced vocabulary' (PE_TOR_8_8).Similar issues arose in relation to Support of Pupils with Different Prerequisites Individually or in Small Groups (7 C TOR , 7 C TEE ).
A fellow student and I took over a small group of pupils with different learning levels.We had developed learning tasks for this situation.Within a short time, half of the children had finished all the tasks while some had scarcely begun.We were overstretched.My fellow student tried to support the children who were still working, and I tried desperately to come up with new tasks.(PE_TEE_4_5) These situations were mostly reported in the framework of differentiated lessons.Within these small groups of pupils, preservice teachers had to address different performance levels, prerequisites and, in particular, paces of learning which was an unknown demand for them.
A further frequent starting point for feedback was preservice teachers' Appearance in the Classroom (17 C TOR , 4 C TEE ).Among reported situations, seven mentor teachers referred to mentees' language, gestures or facial expressions as 'flippant' (PE_TOR_13_2), 'patronising' (PE_TOR_12_9) or 'impudent' (PE_TOR_12_7): 'She rolled her eyes, blew her nose and clicked her tongue when she thought weaker pupils' solutions were not good enough' (PE_TOR_12_8).In 14 situations, preservice teachers and mentors referred to difficulties in classroom management.
I taught a lesson on 'weather' that required pupils to match weather icons to pictures on the blackboard.One pupil with emotional problems came to the front without being instructed to do so.I told him that he must wait for his turn, and I asked him to sit down again.At that point, he tore down all the pictures from the blackboard and roared very loud.I had no idea how I should respond.(PE_TEE_5_2) Such difficulties occurred mostly when mentees showed an inappropriate posture, appeared insecure or when pupils with emotional needs engaged in unknown, unexpected or disruptive behaviours.

Feedback process
The analysis of subsequent feedback indicated in 44 situations (30 C TOR , 14 C TEE ) that mentors did not address the issue of diversity in any meaningful way.Feedback included well-known methods such as Debriefing the Preservice Teachers' Lesson Attempts (31 C TOR , 11 C TEE ) and Suggestions on subject matter, appearance and teaching (7 C TOR , 5 C TEE ); these were linked to Deriving New Goals, which were reviewed in subsequent attempts (11 C TOR , 5 C TEE ).However, the feedback provided was quite general and made no reference to the particular demands of the diverse classroom.Although the preservice teachers reported their inability to deal with issues like differentiation and diverse prerequisites, they did not receive relevant feedback.
In contrast, in 52 situations (28 C TOR , 44 C TEE ), mentor teachers and preservice teachers reported that mentors have adjusted their feedback to fit the context -for example, by adapting and further differentiating the above-mentioned methods of Debriefing the Preservice Teachers' Lesson Attempts (24 C TOR , 11 C TEE ) and Deriving New Goals (10 C TOR , 6 C TEE ).These mentor teachers referred to the setting by explicitly reviewing mentees' activities against the background of diversity.
My first lesson on a history topic in a class with diverging language skills fell below my expectations.I had underestimated how precisely I needed to consider pupils' individual language skills.My mentor discussed the lesson with me and helped me to formulate specific goals for my next lesson.She advised me to reduce my speaking time and to visualise parts of my subject matter, using cue cards for difficult words.After my second lesson, we again discussed whether these goals had been reached.(SE_TEE_4_1) My mentee used the lunch break to talk about healthy food, and she took a pupil's lunch box without asking him.The pupil is a refugee child living in refugee accommodation.The boy freaked out and threw his drinking bottle at her.She was totally shocked and froze with fear.I stepped in with a stabilising exercise.During the debrief, I offered an insight into trauma.We discussed how she could find a balance for herself and the children.This helped her to understand why the boy reacted as he did and how she could deal with similar situations.(PE_TOR_10_1) In this way, the preservice teachers received precise feedback on how they managed individual prerequisites and needs, differentiated teaching, appropriate language level and behaviours.Developing new feedback strategies and content based on the demands of the diverse classroom offered a more specific way of tailoring feedback to the setting.One key strategy was Introducing in Balancing the (Im)Possibilities of Diversity (11 C TOR , 11 C TEE ) to clarify the effects and backgrounds of diversity, as well as the opportunities in a transparent and comprehensible manner and for negotiating teaching options.
We discussed my lesson, and my mentor teacher explained how I could differentiate even more precisely.She emphasised that there is no catch-all differentiation solution to reach every child; it has to be weighted individually in each class according to what is possible and what is not.Her explicit statement that one could never reach the same level for all pupils in diverse classes came as a great relief to me.I still need to work on accepting this, but much of the burden has been lifted.(PE_TEE_1_3) In the aftermath of her lesson, my mentee expressed her dissatisfaction regarding differentiation.She did not like the idea of assigning different tasks to the weaker pupils, as she had put a lot of thought into preparing her lesson.We discussed the impossibility of good differentiation for every task.(PE_TOR_6_4) Feedback helped to explain that diverse classrooms invite negotiation of suitable teaching methods, classroom management strategies and support measures that reflect the individual situation and classroom conditions that sometimes change rapidly.This approach was augmented and enriched by further feedback strategies as illustrated in the following example.
In an English lesson, I planned a scenic play to explain a short story.For this purpose, I divided the class into five groups.I did not really think about the grouping, and I did not take account of the fact that one group comprised all the pupils with language acquisition difficulties.This group was unable to perform the tasks, and conflicts ensued.Two pupils started to cry, and one was pushing the others.My mentor remained in an observer role and did not intervene.I was able to resolve the dispute, but the interruption meant that the scenic play did not finish until the end of the lesson.I felt it was a very good thing that she did not interfere, which meant that I had to deal with the situation myself, although my attempts were only partly successful.(SE_TEE_8_7) Following this example, both preservice teachers and mentors reported situationoriented feedback processes that started with a very special feedback strategy: Not to Intervene (5 C TOR , 7 C TEE ).This meant 'postponing' feedback to allow mentees to try out their own solutions for the diverse classroom.In subsequent feedback tailored to these solutions, a few mentors used the strategy Contrasting Preservice Teachers' and Mentor Teachers' Acting (5 C TOR , 7 C TEE ).
During my English lesson, I lost the thread several times because some pupils with learning difficulties were otherwise occupied.My attempts to rebuke these pupils failed.My mentor teacher encouraged me to continue the lesson.Immediately after, she completed a lesson sequence using only body language to re-engage these pupils.For me, this was very impressive feedback.(PE_TEE_8_9) Following a preservice teacher's lesson attempt, the mentor performed a short teaching sequence, which prompted immediate reflection on the two lesson sequences.It should be noted that the aim was not to encourage imitation of the mentor's approach but to reflect on the different approaches -in this case, the use of a verbal rebuke and engaging body language.

Gains
When preservice teachers did not receive feedback against the background of diversity, they noted the limited added value of feedback that offered only superficial insights into the teacher's role and said they had learned nothing about managing the diverse classroom.Therefore, preservice teachers welcomed tailored feedback that afforded a Deeper Understanding of Effects, Backgrounds and Opportunities Related to Diversity and the Need to Negotiate Options and Deal with Uncertainty as described by 30 mentors and 26 mentees.
Children in diverse classrooms can learn so much through teamwork.In most cases, however, only the weaker children received individual support, but the stronger pupils were also assigned challenging tasks.Individual support must strike a balance between weaker and stronger pupils; my mentor helped me to understand that.(PE_TEE_5_3) The excerpt is a further illustration of those benefits because mentees developed an understanding that teaching and support need to consider a broad spectrum of learning levels ranging from pupils with additional needs to pupils with special gifts.
Additionally, ten preservice teachers reported a sense of Social Participation in Class Life and nine an Increase in Self-efficacy and Self-confidence which was also stated by seven mentor teachers: In her next lesson attempt, she offered two different approaches to the learning task.Using a differentiated worksheet with detailed instructions, she demonstrated the action steps slowly and comprehensibly.She gained in self-confidence and became more secure and authentic in her dealings with the pupils (PE_TOR_13_1).

Path 2: mutual feedback: changing feedback relations consistently to manage the diverse classroom
Figure 4 shows the second approach to feedback in more detail including all categories on mentees' and mentors' activities, gains and the number of codings of each category.

Preservice teachers' and mentor teachers' shared activities in diverse classrooms
This second path was reported in only a small proportion of incidents (22 in total; 15 C TOR , 7 C TEE ).Based on mutual feedback, it followed activities performed jointly by mentors and mentees, notably Shared Teaching (15 C TOR , 7 C TEE ) in the form of differentiation: 'Collaboratively, we prepared a learning circle.I arranged the materials, and my mentee developed differentiated learning tasks'.(PE_TOR_2_12) In my class, two pupils exhibited very disturbing and sometimes aggressive behaviour.There were eight children who had recently arrived in Germany and did not speak German, as well as three children with learning disabilities.My mentor and I both noticed that we were paying too little attention to the other seven children in the class because the lessons were always adapted for children with specific needs.(PE_TEE_5_3)

Feedback process
Mutual feedback, in which mentor and mentee switch roles as feedback giver and receiver, was tailored to the diverse setting in all of the reported situations.The most frequently used feedback methods were reflection-based; both groups reported using Shared Reflection on Differentiated Teaching and Diversity (10 C TOR , 7 C TEE ) to arrive at a shared understanding of pupils' different prerequisites and negotiation of options.
In group work and differentiated tasks, my mentor and I observed the children together.It was very informative to observe how differently pupils dealt with the materials and options provided.We exchanged observations and gained a new understanding of the children, their needs and possible supports.(PE_TEE_5_8) In a lengthy exchange after a math lesson on integers, we reflected together on each individual pupil.Because their capabilities diverged so widely, we agreed that each of them needed their own worksheet, with differentiated assistance according to the individual's learning level.(PE_TOR_1_2) Developing a shared concept of teaching in diverse classrooms involved Mutual Criticism (4 C TOR , 1 C TEE ).
Over two nights and days, we co-developed a concept and taught pupils in four parallel courses.The collaboration was very successful, and the mentees introduced valuable ideas. . . .The framework was always very respectful, and everyone was free to criticise each other's parts of the lesson.(PE_TOR_5_7) Mutual criticism allowed also preservice teachers to give their views on their mentor teachers' lessons and handling of the pupils.

Gains
Mutual diversity-oriented feedback delivered positive outcomes for the preservice teachers in terms of their understanding of diversity-related effects, backgrounds and options.Additionally, mutual feedback proved beneficial for mentor teachers (6 C TOR ); for example, they can reflect upon and further develop their practice in accordance with the knowledge and methods preservice teachers bring from the university.
Two mentees working with me suggested the creation of lap books for the topic 'Teeth'.I was not familiar with lap books, but the children were enthusiastic about the end product, and since then, I have created lap books in my class twice a year.(PE_TOR_8_10) My mentee came up with different exercises and introduced the long jump.On observing the children, she realised how differently they performed these exercises.She was touched by the different movement patterns, which she analysed and documented very accurately for each child.Her emotion and her analysis both proved very beneficial for me as the basis for my physical education lesson plan, which was tailored to the various pupils' needs.(SE_TOR_10_3) The pupils also benefited, reported in nine incidents (6 C TOR , 3 C TEE ), mostly because the lessons could be further oriented to support their individual needs.

Novel feedback for preservice teachers in the diverse classroom
On entering their chosen career, teachers should be able to take account of pupils' individual prerequisites and behaviours, as well as their language and background.In their first attempts at teaching and classroom management in diverse conditions, novel forms of feedback are needed to guide them through and to initiate learning and professionalisation.Feedback is thereby a key element (Bjørndal 2020;Clarke, Triggs, and Nielsen 2014;Orland-Barak 2016), and the present findings confirm that adapting feedback to the demands of the diverse classroom offers benefits for mentors and mentees involved.
However, despite the obvious demand, only part of the participating mentors adjusted their feedback.Both mentors and mentees identified teaching (attempts) and classroom management as points of departure for feedback.Faced with diverse language skills, performance levels and social backgrounds, preservice teachers often had difficulties to complete lessons, manage pupils or match tasks to individual prerequisites.Paradoxically, all participants agreed about mentee difficulties in the diverse classroom and their need for support.However, in a little less than half of the reported situations (30 of 71) mentors have made no reference to this when providing feedback.Consequently, preservice teachers had relatively few opportunities to learn or to gain the relevant competencies.
In contrast, about half of the mentors involved introduced novel feedback strategies and content for the diverse classroom.These mentors specified well-known feedback techniques such as debriefing, making suggestions about subject matter, appearance and classroom management and differentiating tasks according to individual performance and background.In addition to differentiation, individualised teaching and plurilingual approaches, other strategies addressed the diverse classroom in a more targeted way.In particular, it was considered important to help preservice teachers to balance the difficulties and opportunities of diversity and to make these transparent and comprehensible.
Displaying an awareness of the need to negotiate options, mentors responded to concerns about 'equal treatment' by specifying what is possible and expedient (or not) when pupil performance differs.Beyond improving differentiation skills and individualised teaching, participants noted that teaching and classroom management options must be negotiated according to the individual situation in diverse groups (Tomlinson and Imbeau 2010).At the level of self-regulation, preservice teachers acknowledged the importance of clear feedback on what is possible and what is expected of them.Mentors' feedback can be further characterised by its emphasis on mentees' self-responsibility.In particular, mentors stressed the value of 'postponing' feedback rather than intervening when mentees encountered challenging situations, even when things got out of hand.The underlying aim was to afford mentees an opportunity to develop their own solutions against the background of the diverse classroom.
Many of the reported feedback processes were one-sided, where mentors gave feedback and mentees received it.However, preservice teachers gained more from feedback that emphasised reflection, self-responsibility and scope for decision-making (Canipe and Gunckel 2020; Gallo-Fox and Scantlebury 2016), which in turn increased mentees' selfefficacy, self-confidence (Fives, Hamman, and Olivarez 2007) and competencies for teaching and managing diverse classes.Constructivist mentoring emphasising co-construction of differentiated teaching through reflection and mutual feedback (e.g.Canipe and Gunckel 2020;Richter et al. 2013) incorporates role switching that broadened reported benefits and extended those benefits to all of the groups involved.When linked to preservice teachers' existing skills and knowledge, joint reflection on pupils' individual prerequisites and classroom structure enriched lessons and teaching practices in diverse classrooms (Hagger and McIntyre 2006;Lopez-Real and Kwan 2005) and enhanced the response to additional pupil needs.

Implications
Both preservice teachers and mentor teachers highlighted the demands faced in diverse classrooms, but many mentors made no reference to this issue in the feedback they provided.It seems important to investigate the reasons for this oversight.Even if the study did not explicitly point to such findings, negative attitudes and beliefs about diversity could influence perceptions and interpretations of educational practice (Pajares 1992) and may contribute to some mentors' failure to address this issue (Kumar and Hamer 2012).This could be related to the fact that mentors themselves have a lack of knowledge and strategies for managing diverse classrooms and cannot provide appropriate guidance for their mentees.
Therefore, on the policy level, there is a need for implementing or expanding (mandatory) further education programmes that intertwine the issues of attitudes to diversity, knowledge and strategies for managing diverse classrooms and being an effective mentor (Talbot, Denny, and Henderson 2018).Such education programmes could be located in school extern teacher training institutions but could be also subject to school development processes at the individual school.Beyond gaining knowledge on managing diverse classrooms, core values that are the basis for attitudes can be addressed by creating a school's mission statement that fixes these values of the whole school community (Ainscow et al. 2012).Bringing attitudes and knowledge together with mentoring practices lays the foundation for effective diversity-oriented and mutual feedback.Changing both attitudes and role concepts, for example, in order to accept feedback by the mentees, could be a lengthy and mentally challenging process for mentors (Canipe and Gunckel 2020;Gallo-Fox and Scantlebury 2016).Thus, further education programmes could also include supervision or peer consulting with other mentors.Additionally, such concepts need to be assessed with the aim of improving their effectiveness as described in the further research section below.
Additionally, higher education and practical mentoring in school settings should be more closely integrated (Talbot, Denny, and Henderson 2018), with an emphasis on developing the professional skills required to manage diverse classrooms.Moreover, future teachers need to be prepared for a novel feedback culture from the beginning of their education.Collaborative learning activities which include the aspects of mutual feedback, appreciation and a negotiation of goals and work distribution during the university-based part of training can establish a basis for an effective feedback culture which targets also the demands of diverse classrooms.

Limitations and further research
The use of CIT raises questions about the extent to which reported situations can be generalised (Gremler 2015), as critical incidents are individual appraisals.Additionally, well know effects of group discussions could interact or coincide with well-known biases.Some of the participants may have been less open about events that were unusual or easily misunderstood.Additionally, some may have may have feared the exposure that can result in socially desirable descriptions or in omitting details that they do not dare report (Krueger and Casey 2020).The latter could also occur when participants orientate themselves on other mentors and mentees they experience as 'prestigious members of their group', for example, because they reported successful acting during practicum.This carries the risk of the Prestige Bias (see Brand, Mesoudi, and Morgan 2021).For that reason, we deliberately assembled groups from different schools and separated mentors' and meentes' discussion groups to counteract any fears of exposure to work colleagues or mentors.Moreover, the validity of the data used and the analysis in the present study was ensured by thematically structured group discussions and the provision of guidelines for training moderators (Heine et al. 2016).Additionally, mean IRR and, in particular, communicative validation was used (Kvale 1995) to make sure that the statements of all participants were equally considered and weighted (e.g.see the meta-analysis by Dierdorff and Wilson 2003).
To further mitigate such issues, in-depth interviews teachers could help to validate and refine these results.Individual interviews might serve to clarify why some mentors' feedback fails to address the diverse classroom.For example, the described issue of attitudes on diversity should be investigated in more detail.Moreover, with respect to the project's higher-level goals and the mentioned above policy level, novel feedback concepts need

Figure 1 .
Figure 1.Example on the analysis.

I
Figure 2. Model of feedback in diverse classrooms.

Figure 3 .
Figure 3. One-sided approach to feedback in diverse classrooms.

Figure 4 .
Figure 4. Mutual approach to feedback in diverse classrooms.