A qualitative study of primary teachers’ classroom feedback rationales

Abstract

Background

As part of teachers’ everyday classroom assessment practice, feedback can be seen as connected to the formative function of assessment, with the aim of helping students in their learning processes. Much research on teacher feedback focuses precisely on the feedback’s formative quality. However, in order to strengthen our understanding about the nature of teacher feedback, we also need to understand more about teachers’ rationales for giving feedback to their students, especially in primary school settings.

Purpose

The present study aimed to explore and conceptualise primary school teachers’ rationales for giving students feedback.

Sample

Thirteen Swedish primary school teachers ( 10 women and 3 men) with 4 to 40 years of teaching experience working with students aged 7–9 years-old (grades 1–3), participated in the study. An open sampling procedure was adopted to recruit the teachers.

Design and methods

Data were collected using a semi-structured interview approach. We employed a constructivist grounded theory design for the coding and analysis of the transcribed data.

Results

Analysis indicated that two main concerns emerged as regulating teachers’ assessment practices. These addressed what the teachers perceived as (1) students’ academic needs and (2) students’ behavioural and emotional needs. According to the findings, the teachers’ rationales for giving students feedback were based on those needs, and dependent on factors such as situation, relationships, time and effort. This resulted in a constant comparison and weighing of different needs by the teachers. Some needs were described as prioritised before others, which caused some rationales to be identified as taking precedence over others.

Discussion and conclusions

Based on a systematic analysis of – and thus grounded in – interview data from primary teachers, the current qualitative study offers a framework for surveying, understanding and discussing teacher feedback. Overall, the study showed how everyday practices of classroom assessment and classroom management overlapped, thus underlining the importance in teacher education of understanding classroom assessment, classroom management and the relationships between the two.

Keywords:

• Classroom assessment
• feedback rationales
• grounded theory
• interview study
• primary school

Introduction

Teachers’ actions in social situations are, according to a symbolic interactionist approach (Blumer 1969; Charon 2007), affected by the teachers’ perspectives on the situation and the meaning that they derive from them. ‘Humans act in a world they define, and although there may actually be a reality out there, their definition is far more important for what they do’ (Charon 2007, 136). With reference to Charon (2007), it is therefore imperative to understand teacher feedback from the perspective of the teacher. How teachers define the situation and rationalise what is needed are central to how they give feedback. It is, therefore, important to investigate this in order to understand more clearly the everyday teacher feedback situation in the classroom. When included in teachers’ dialogic repertoire and everyday classroom assessment practices, feedback can be seen as an important part of formative assessment. As such, feedback can then be discussed in terms of effectiveness, depending on its level of formative function (e.g. Gamlem and Munthe 2014; Jonsson, Lundahl, and Holmgren 2015; See, Gorard, and Siddiqui 2016). It is evident that, no matter how teachers carry out classroom assessment, it will affect students’ learning (Popham 2009) and self-esteem (Gipps 1995). Hargreaves (2011) found that teachers perceived feedback as most effective when there is trust in the teacher-student relationship, the feedback clearly relates to progress and criteria, and students fully understand the feedback. Research on feedback related to formative assessment has often drawn attention to how teacher feedback is associated with learning outcomes and academic achievement (Brookhart 2012; Hattie and Timperley 2007). There have been some large studies within the assessment for learning field, in the UK (Black et al. 2003), in Scotland (Hayward 2015) and in Norway (Hopfenbeck et al. 2013). It has been noted that balancing a formative approach with accountability demands was something teachers found challenging (Hayward 2015), and some teachers felt that developing their formative practices required a change in view on teaching (Black et al. 2003).

Studies based on classroom observation offer the opportunity to analyse, in detail, different dimensions of teacher feedback. Hargreaves (2014) identifies and describes categories of teachers’ autonomy-promoting feedback, noting how teacher feedback can potentially support children’s development as independent learners. Elsewhere, Boistrup (2015) construed feedback discourses including on the one hand, ‘do it quick and do it right’ – where the feedback focus was on matters such as correct answers or the number of tasks accomplished by the student – and, on the other, ‘reasoning takes time’, which was characterised by advocating a slower pace and a focus on content and processes within the school subject. Some of the classroom assessment described in the literature is informal, implicit and often non-verbal (Jordan and Putz 2004), as is much of the feedback communicated in dialogues between teacher and student (Ruiz-Primo 2011). Research on feedback is not, though, restricted to aspects of academic performance. Studies of the link between feedback and student behaviour go back at least as far as White (1975), who found teacher approval as most common for addressing instructional behaviour and almost non-existent for addressing managerial behaviour. Later research on feedback on behaviour often highlighted approval and disapproval (e.g. Swinson and Knight 2007), and specifically the use of praise (e.g. Brophy 1981; Chalk and Bizo 2004). Indeed, Torrance and Pryor (1998) described teachers’ everyday assessment practices as including assessment on behavioural performances as well as academic achievements. Tunstall and Gsipps (1996), in their typology, described assessment-related feedback as being about classroom and individual management as well as performance, mastery and learning orientation.

There are various definitions of the term ‘feedback’. In the present study, we adopt a broad understanding of feedback, including what Hattie and Timperley (2007) label ‘feed back’ (response on performances), ‘feed up’ (goal-oriented feedback) and ‘feed forward’ (feedback about the next step). Feedback is defined as interpersonal, communicating responses to performances, academic and behavioural.

The present study: aims and purpose

From the existing literature, it is evident that much has been discussed about how feedback is communicated, how feedback can be enhanced, different feedback focuses and the effects of interventions. However, research specifically concerning how teachers conceptualise feedback (e.g. Brown, Harris, and Harnett 2012), and, hence, studies focussing on teachers’ feedback rationales that give teachers the opportunity to discuss their own understanding of and perspectives on feedback in their own words, seem limited.

The present study, therefore, aimed to explore and conceptualise primary school teachers’ rationales for giving students feedback. The following questions were asked:

•	What are the main concerns and rationales that teachers describe for how feedback is communicated by them in the classroom?
•	How are the relationships between those main concerns and rationales manifested?

In line with a symbolic interactionist framework (Blumer 1969; Charon 2007), we viewed it as essential to study the teachers’ ways of reasoning about their classroom feedback to their students, in order to enhance our knowledge about teacher feedback in the everyday classroom. We adopted a grounded theory methodology, since its explorative approach is especially suitable for studying topics where studies are limited. We asked teachers to rationalise the feedback strategies they use in classroom interactions with students, and then analysed their narratives. The findings offer a framework for further analyses of and reflections on teacher rationales for giving students feedback, both for researchers and teachers.

Method

Interviews were carried out and grounded theory analysis was conducted in order to examine, qualitatively, the teachers’ rationales for giving students feedback. As Morcom (2014) stated, ‘qualitative research methodology endeavours to understand the world of the participant by situating the researcher with all their values and assumptions in that world’ (21).

Ethical considerations

We have carefully followed the ethical principles and guidelines as stated by The Swedish Research Council (2017). We obtained informed consent from all participants in the study. They were also informed that they had the option of withdrawing from the study at any time. In order to ensure confidentiality for the participants, all participants, including people mentioned in interviews, and schools and locations, have been given fictitious names in transcriptions, field notes and publications.

Participants

Thirteen teachers ( 10 women and 3 men) with 4 to 40 years of teaching, working in Swedish primary schools with students aged 7–9 years-old (grades 1–3), participated in the study. An open sampling procedure was adopted to recruit the teachers, as it ‘seeks to maximise variations in experiences and descriptions using participants from contrasting milieus and backgrounds’ (Hallberg 2006, 143). In line with this quotation, we wanted to discover and categorise phenomena common among participants from a broad range of contexts. The participating teachers taught in 11 different schools, in 8 different municipalities. Some worked in larger inner-city schools and others in small rural schools, in schools in low and mixed socio-economical areas, in schools with ethnically mixed student groups and schools with almost only students from non-ethnic minority backgrounds. In this way, the data were based on a diverse student population: socio-economically, ethnically and socio-geographically.

Data collection

Interviews were conducted by the first author. Field notes, classroom observations and classroom recordings from four of the teachers’ classrooms, analysed in a previous study (Eriksson, Boistrup, and Thornberg 2017), served as a source in the process of constructing interview questions. The four teachers from Eriksson, Boistrup, and Thornberg (2017) were interviewed first; two of them together, the others individually. During analysis, new questions arose, which led to additional interviews with three of the teachers. These interviews were supplemented by interviews with nine additional teachers. The interviews were conducted in Swedish. In total, 15 semi-structured interviews were conducted, with an average length of 55 min. Coding and analysis, carried out in parallel with the data collection, resulted in additional interview questions and re-interviews in order to examine the constructed codes and categories (this iterative procedure is described as theoretical sampling in grounded theory, see Charmaz 2014).

In the interviews, the teachers were asked: in what situations they perceived that they gave students feedback, how they described their feedback in terms of aim, focus, strategies and priorities, and what things they saw as affecting their feedback. The interviewer used probing and follow-up questions (e.g. ‘How come?’, ‘What do you mean?’ and ‘Tell me more’) and took a non-judgemental approach (e.g. Kvale and Brinkmann 2009). In the interviews with the first four teachers, some questions were situated, linked to classroom feedback that had been observed. Each interview was recorded and transcribed verbatim. Quotations from the transcribed data presented in the findings have been translated from Swedish into English by the authors.

Data analysis

The data analysis was guided by grounded theory methods (Charmaz 2014; Glaser and Strauss 1967). Analysis was accomplished by initial and focused coding (Charmaz 2014), theoretical coding (Glaser 1978), constant comparisons, theoretical sampling and memo writing (Charmaz 2014; Glaser 1978; Glaser and Strauss 1967). While adopting grounded theory, we tried to encounter the data with what can be described as an open mind rather than an empty head (Dey 1993). This meant that we initially tried to put what we already knew aside, while staying close to data during the analysis. In this way, we could construct codes that were not only derived from data, but were also realised in wordings that were authentic in relation to the data. In a later part of analysis, comparisons were made with previous research. Thus, constant comparison was made within the interview data, and between codes and categories.

In line with a constructivist grounded theory approach, symbolic interactionism was used as an open-ended theoretical perspective combined with curiosity and openness (see Charmaz 2014). Accordingly, definition of situation was used in terms of how the participants interpret a situation they are in which influences their actions (Thomas and Thomas and Thomas 1928); main concern addresses what the participants are mainly occupied with (Glaser 1978) in relation to feedback; and rationales were regarded as the participants’ motives or logical explanations for acting in different ways (Alvarado 2003). We also adopted various categories of teacher feedback strategies, including deliberating, i.e. promoting explorative thinking and discussions; mirroring, i.e. a neutral, non-evaluating response, merely confirming having noticed a strategy; and steering, i.e. communicating that there is a desirable strategy or answer, developed by Eriksson, Boistrup, and Thornberg (2017) as sensitising concepts (cf. Blumer 1969). In line with a constructivist grounded theory approach, we have used the literature as a possible source of inspiration, creative associations, critical reflections and multiple lenses (Thornberg 2012).

After six interviews, we had constructed a set of focused codes, which were further elaborated by the three interviews that followed, and strengthened and confirmed by the last six interviews, ensuring saturation and trustworthiness. The different categories of feedback rationales occurred in all participating teachers’ descriptions, with slight variations in frequency. Of course, it must be recognised that the particular schools that the teachers worked at might have affected how frequently different feedback rationales were used. That was, however, not the study’s focus, as we were interested in describing the common feedback rationales the teachers used.

Findings

From the data analysis, we concluded that teachers’ feedback practices depended on what they interpreted as perceived needs. We defined perceived needs as any need that the teachers described that they had to address with feedback, in the classroom situation. Thus, it seemed that teachers’ main concern in their feedback practice was to meet perceived needs in the ongoing classroom situation. In line with grounded theory, the terms used to describe the feedback rationales were constructed from the data as such. Consequently, they are not terms already established within the research field. In this way, the findings clearly address participating teachers’ rationales, providing authenticity in the sense of a grounded theory approach.

Two patterns of general perceived needs for giving feedback

Two recurrent patterns of general perceived needs for giving feedback to students were constructed and coded as two main categories representing different kinds of needs: (a) students’ academic needs and (b) students’ behavioural and emotional needs. Both categories of needs could be perceived by teachers to be at an individual level as well as group level. They evoked a variety of feedback rationales that influenced and guided teachers’ feedback practice, as well as the choice of necessary feedback strategies.

In line with our adopted prototypical model of categorisation (Dey 1999), we recognised overlaps between our developed categories, as in the following excerpt:

Isabelle has extremely low self-esteem, although she is one of the best in the class in algebra. But she has no belief in herself. She also has really low status in the group. So, I am struggling with what kind of support might help her. (Alex)

When Alex stressed that the student was one of the best in algebra, we analysed it as a rationale addressing the student’s academic needs. The teacher’s concern regarding the student’s low status in the group was analysed as a rationale addressing the student’s behavioural and emotional needs. The lack of ‘belief in herself’ was perceived by the teacher as something inhibiting the student in her otherwise excellent learning (i.e. considering her academic needs), but, at the same time, an emotional need in terms of the need to improve her self-esteem or self-confidence.

Because a prototype is presented as a set of typical characteristics of a category, membership of a category was a question of the degree of family resemblance to the prototype rather than sharing the full set of common features (Dey 1999). Therefore, overlaps between categories were possible in our analysis, which reflected greater sensitivity to the nuanced complexity of teachers’ feedback rationales. As mentioned above, teachers’ feedback rationales were analysed into the two main categories of student needs depending on how the needs were emphasised by the teachers in relation to the various feedback rationales. The excerpts chosen to illustrate a certain category of teachers’ feedback rationales may come from a single or just a few teachers, but are broadly representative of the participants as a group. Although all categories were present in all teacher interviews, sometimes a category would be clearly evident and at other times it maybe linked to other categories, which made some excerpts more suitable than others for exemplification purposes.

Pattern 1: addressing academic needs

Academic needs were related to students’ skills, knowledge, learning process and the ability to sustain concentration on school subject activities. This main concern addressed what kind of feedback would best fit the group, or individual students, in order to address and optimise academic learning and activities. The teachers’ perceptions of their students’ knowledge tended to set parameters for how they planned their lessons and teaching, including their rationales for feedback.

Feedback rationale (1): academic encouragement

When using the feedback rationale identified as academic encouragement, the teacher viewed the student as someone in need of support and encouragement in order to progress. This rationale was mostly associated with giving praise (e.g. ‘For those who are unconfident, you have to give praise for every sentence or letter’ (Alex)), and ‘mirroring’, which, as explained earlier, refers to a neutral, non-evaluating response, confirming that the teacher has noticed what kind of strategy was used or answer the student has given (e.g. ‘They need to feel noticed in order to push themselves into doing more’ (Chris)). Thus, the teachers expressed an encouraging approach in which they recognised and showed care more frequently to those students who needed it the most by praising and encouraging their academic achievements regarding the amount of work done and its quality. As one teacher, Tove, explained, ‘They would need someone beside them. Someone who confirms that I [the student] am doing things correctly, or someone that gives a slight push’.

The teachers particularly emphasised this encouraging rationale when dealing with student groups in which there were students in need of extra learning support or with insufficient language skills. Showing faith in students’ academic ability was highlighted as an essential part of academic encouragement, as this teacher quotation suggests:

They need to feel that there is someone there to help them. / – / Without, for that matter, coddling them, telling them that ‘yes but, well it’s no big deal, but it’s a pity that you find it difficult’. But rather that ‘we can do it, if we decide we can do it, but it will require hard work.’(Mio)

While some students were described as needing much academic encouraging, students confident in their learning were described by the teachers as having a lesser degree of need for such feedback. Nevertheless, the teachers emphasised every student’s need to be seen; as one commented, ‘Everyone needs to be seen anyway. So that they know, that they know that I know what they are working with’ (Kim).

Feedback rationale (2): individualising

While the academic encouragement rationale was based on a perception of the students’ need for positive and encouraging support, the rationale individualising refers to fitting the feedback to the individual student, which took place when the teachers provided each student with feedback adapted for their specific academic needs. Robin described the rationale as having to do with:

meeting the students and giving them feedback related to where they are in their learning process. / – / Pointing out what is good and then ‘Well, how can you proceed?’ / – / And then they can discover it by themselves. (Robin)

Robin reported a need for the teacher to be receptive to the individual student’s knowledge and learning processes in order to meet their individual needs. In this rationale, teachers generally highlighted the fact that all students have different areas of knowledge and are, therefore, in need of learning and improving different skills.

I can’t tell Tomas that he has done a good job, if it is not good based on his [work], You can’t compare Tomas with someone else [another student]. You must compare Tomas with what he has achieved earlier. (Toni)

As in the excerpt above, the rationale individualising was focused on the particular student’s individual academic progression. To manage such feedback, it was necessary, according to the teachers, to ensure that they knew their students, as well as what kind of feedback worked best for each student.

Feedback rationale (3): peer learning modelling

When adopting the rationale peer learning modelling for giving feedback, the teacher recognised desirable knowledge and skills among the students. They did this by highlighting a student’s or a group of students’ performance or learning activity as a model for the others, or for a specific student – in other words, drawing students’ attention to those who performed well or better, in the hope that this might lead them to adopt the model themselves, thereby learning from their peers. As Kirsten explained, ‘they need to get some sort of model. Otherwise they are supposed to guess what it might look like’.

The rationale was also evident when teachers merely underlined the advantages of reviewing others’ work and working with peers. As Alex put it, it was important ‘To change focus from me [the teacher] as the only one who knows’; Pim similarly observed, ‘Because it’s not as if it [the feedback] always has to come from me’. Mio explained the advantages of using peer learning modelling in this way: ‘You may need different ways of saying something. And then it might be the way the peer said it that made them learn quicker’. The teachers argued that, by seeing peers as ‘knowers’, the students would, to greater extent, seek help from one another.

Furthermore, the teachers considered peer learning modelling as a possible source of inspiration and motivation. For instance, Toni stated, ‘Hopefully they became a bit inspired by each other to continue working on their texts’, and added that the students seemed to like reading their texts to their peers. Kirsten, in turn, described the students as more motivated to do a task ‘if they know that it has a recipient’. Thus, peer learning modelling was considered to be stimulating and helpful to both the observer and the model.

Feedback rationale (4): task controlling

The feedback rationale task controlling concerned the teacher wanting to ensure that students’ activities corresponded with what was planned, and making sure that the students focused on what was intended. Task controlling appeared to be prominent during group discussions where the teacher, through feedback, tried to make sure that the discussion stayed focused on the intended topic, as illustrated by the excerpt below:

If you lose grip of the group during a conversation, and it [the discussion] takes off in a different direction–, I think it is about how the group works as a group. With some groups, you can have those deliberating conversations and stick to your intentional idea. (Alex)

In this excerpt, task controlling is revealed as the teacher reflects on how certain kinds of groups needed feedback so as to make sure that the students maintain their focus on the topic or task and to avoid ‘losing grip’ in the classroom situation as a teacher. According to the teacher reports, feedback driven by task controlling was accomplished using a high degree of ‘steering’, in which the teachers gave students controlling questions and instructional feedback.

Feedback rationale (5): classwork flowing

The teachers often encountered classroom situations in which several students simultaneously expressed the need for individual feedback concerning their academic needs. This was experienced by the teachers as a constant call for help, leading to the rationale classwork flowing setting in. According to the teachers, students’ aims in asking for help could be: wondering what to do, how to do something, or merely wanting confirmation. The teachers believed that a failure to address these needs would result in working and learning interruptions and also a higher risk of off-task behaviour. This rationale motivated the teachers to try to help as many students as possible by giving short, specific feedback that was assumed to be sufficient for the students to continue their work. They coordinated several short feedback inputs to make classroom work flow without interruptions. As Vanja explained:

So, then the feedback perhaps might not be so well thought out. You might not have the time, or take the time to reflect for yourself, or be that accurate and specific and calmly make time to say what you want to say to the child you are standing beside.

The classwork flowing rationale compelled teachers to be time-efficient and, therefore, often motivated them to use feedback such as ‘steering’, in which the teacher piloted a student by presenting a strategy that would help complete the task, or by correcting or confirming a student’s performance.

The need to be time-efficient, by giving short responses to as many students as possible, both demotivated and dissuaded the teachers from using the more time-consuming and meta-cognitive supportive ‘deliberating’ as feedback strategy. Furthermore, if teachers in such classroom situations adopted deliberating feedback by posing a question that invited the student to reflect, it risked not being followed up by the teacher, who needed to leave them in order to help someone else, and then might not have time to return.

Feedback rationale (6): time-for-reasoning

In contrast, in some situations, teachers considered classwork flowing as a rationale for feedback as less adequate. These situations included: when working with a student group with fewer students in need of extra support, if there were more teachers in the classroom, or when all students seemed to be confident in their tasks. Chris compared the different classroom situations:

It does not become this, swoosh (illustrating a constant running between students by waving the arms back and forth), that I have to go there, and then I have to be there, and then I need to go there. Instead, I can sit down with a student and discuss. (Chris)

Chris described how a less stressful classroom situation created space for time-for-reasoning to emerge as a significant feedback rationale. This occurred when the teachers found themselves having plenty of time and, thus, were more able to focus on a specific student or a pair of students, and to sit down and discuss things. They underlined the importance of giving all students, regardless of needs, time for reasoning, when opportunities arose. In addition, time-for-reasoning was also connected to making a fair assessment, as Toni explained in the context of maths teaching and learning:

I find it insufficient merely to look at [the answer]; you have to listen to how they reason. Not just look at–, it is like–, to complete those tasks on time, that is one way. But you want to know how they think too in order to make a correct assessment, I think. Not only that you–, you can’t just move numbers, you have to think–, the students’ strategies for reaching the answer. (Toni)

The time-for-reasoning rationale was associated with the feedback of ‘deliberating’. In addition to creating a pedagogic dialogue to support students’ learning and reflection of learning, this helped the assessment of learning by providing the teacher with more insight into students’ strategies and knowledge by assessing how they reason in the dialogue.

Pattern 2: addressing behavioural and emotional needs

Behavioural and emotional needs refer to students’ needs related to developing and maintaining desirable behaviour, emotional well-being and adjustment, such as: positive self-esteem, appropriate social norms, positive classroom climate, feeling included in a group or class.

Feedback rationale (1): need-for-order

When it came to the main concern of students’ behavioural and emotional needs, different kinds of need-for-order rationales were emphasised by the teachers. It sometimes had to do with creating order, where different rationales for doing so were highlighted in the teachers’ statements. At other times, it was about maintaining order. In the data analysis, three different need-for-order rationales were constructed. These were termed (a) fire-fighting, (b) peer order modelling and (c) without-order-no-learning.

(a) Fire-fighting

The metaphorical term fire-fighting was used for preventing disorder, or maintaining a rather fragile order in situations where potential disorder was anticipated. Taking action to maintain classroom order was then considered necessary. The teachers explained it by emphasising the need to be one step ahead in order to maintain order in the classroom. This rationale could take the form of the teacher preparing specific students beforehand in terms of what the lesson will be about; for example, ‘Then I’ll have to catch them on the way in. Tell them what’s going to happen now’ (Pim). This need-for-order rationale was, however, mostly described as the teacher needing to be there the moment a student has finished a task, to give feedback directly on the performance and directing the student towards the next task. Vanja said that ‘Sometimes it’s about whether I should decide that it becomes chaotic, or shall I decide to give these children help quickly so that they get started and it doesn’t become chaotic’. The teachers stressed that being one step ahead of some students, helping them before others, might not be fair. But, the thinking was, since chaos was avoided, everyone benefitted from it. Alex, Robin and Pim all described it metaphorically as ‘it’s sometimes all about extinguishing fires’. This rationale also comprised teachers noticing and managing small cases of disorder just after they appeared, preventing them from escalating into wider disorder, perhaps involving more students, and becoming more difficult to control. The teachers described a need for having so-called ‘eyes in the back of the head’. This connects with Kounin’s (1970) reference to teachers’ ‘with-it-ness’ in the classroom, which thus leads to being alert to signs that they may need to act on.

(b) Peer order modelling

Because students vary in their compliance with rules and standards, teachers sometimes adopted the peer order modelling rationale. This was based on the assumption that students can learn to behave more appropriately by observing how peers who display desirable behaviour will produce positive consequences for themselves or others. As one teacher explained:

If I give a student a response, like ‘Well, how nice that you show me that you learnt how to raise your hand and be quiet and wait for your turn. Good! That’s why you’re getting to speak now.’ Then I can see how some other student, who is sitting there, talking when they are supposed to be quiet and listening, reacts like ‘Oh!’ (Robin)

Robin described how praising one student could make another student react, and hopefully adapt. Equally, Kirsten commented that ‘It’s a really good help if you have students who don’t really know how to behave in a classroom’; continuing, ‘instead of complaining’. As in the rationale of peer learning modelling, there was the aspiration of a social learning effect. This rationale can be associated with Bandura’s (1977) concept of vicarious reinforcement, and was often described as enacted through praise. However, if praise was overused, the teachers considered it to be at risk of not been perceived as genuine among the students.

(c)Without-order-no-learning

This rationale accentuated the need-for-order as something that had to be fulfilled in order for the students to learn, and was often communicated as a joint class project: ‘Let’s just put down everything and try to make this work, so it doesn’t have to be like this’ (Mio), and, ‘It’s hard for anyone to concentrate on work if the sound level in the classroom is this high’ (Robin). The teachers emphasised the wider possibilities of giving more ‘deliberating’ feedback, promoting explorative discussions and thinking through dialogue when there was order in the classroom, which was confirmed in the classroom observations. When there was disorder in the classroom, a considerable part of the teachers’ focus and feedback efforts were on creating order. Hence, the teachers considered classroom order to be the superior need, because a failure to address this behavioural need at group (classroom) level would severely block their ability to address students’ academic needs as well as other behavioural and emotional needs.

Feedback rationale (2): Caring

Students whom teachers attributed with low self-esteem tended to attract teachers’ caring rationale for feedback, and were often seen as being in need of extra praise aimed at improving their self-esteem: ‘Regarding feedback related to more social aspects, I think it’s important to know the students well. / – / Some you have to be really careful with’ (Pim). The teachers sometimes described some students as appearing extra sensitive to criticism and, therefore, this was something to be aware of and take into consideration when giving feedback.

There are some that I know that I praise intentionally. Mia is such a girl–, that I need to lift–, and Clara, I have to lift her too. With her, I try to remember to be friendly, extra kind, if it–, and highlighting her answers [during gatherings] as good. (Alex)

Low self-confidence, despite good subject knowledge, was perceived as a reason for giving some pupils an extra amount of praise. Caring was manifested by the teacher using praise, ‘mirroring’ and taking time to talk to a student, highlighting strengths and showing that efforts had been noticed. It was also apparent in teachers being clear and specific; making sure the student understood that feedback on performance was not the same as feedback on a person. The perceptions of what kind of, and how much, praise different students were in need of, and how frequently they needed praise, varied according to the teachers, and were linked to academic needs as well as behavioural and emotional needs.

The relationships between teachers’ feedback rationales

Depending on how different needs were weighted in relation to each other, and consequently which rationales the teacher decided to act upon, the teacher could choose different feedback strategies. In ongoing classroom situations, teachers had to take into account and coordinate various feedback rationales based on their ongoing assessment of the present academic, emotional and behavioural needs. In this process, there was a constant interplay between feedback rationales, in which without-order-no-learning, classwork flowing, and individualising seemed to be the three primary rationales. The coordination of these three rationales guided teachers in their adoption of the other feedback rationales.

Meeting every student’s individual needs in the moment

The teachers were deeply concerned with teaching, supporting and giving feedback based on students’ individual needs. They described striving to be able to meet all students’ individual needs at the same time. The focus was not limited to academic or behavioural and emotional needs, but rather to all needs identified at a specific moment. Yet, in alignment with the teaching task, this predominantly concerned students’ academic needs. The teachers acted upon striving for this using different rationales: which ones were used depended on which needs were identified. When there was order in the classroom and everyone seemed to be working, there were few acute needs identified. Fulfilling those needs was then seen as rather easily manageable. When there were several hands raised at the same time, the striving for individualising took shape by means of the teacher helping students in pairs or groups, by simultaneously switching between two students, or by rushing between students in order to provide short, specific feedback.

Through analysis, it became apparent that the rationale individualising informed and guided the feedback rationale academic encouragement. The teachers expressed their belief that students performing poorly were in need of more frequent support, while students performing well were not. Individual differences regarding the amount and frequency of feedback needed was a central aspect of teachers’ attempts to meet all students’ needs at the same time. Whereas individualising focused on particular individual needs, time-for-reasoning was guided by a more general idea of pedagogical dialogue as something that helped every student. In fact, according to the teachers, a combination of these rationales – i.e. individualised time-for-reasoning – emerged in the analysis as ‘the gold standard’ rationale of classroom feedback. However, too often it was too difficult to fulfil because of the need to construct classroom order: there being too many needs at the same time, and a lack of time. Teachers’ attempts to meet all students’ individual needs manifested itself in the sense that there were a few main rationales, and at the same time, the rationales that seemed to be needed at that moment were the ones used and seen as equally important when opportunities arose and compromises were few.

However, the teachers drew attention to the difficulty of meeting all students’ individual needs at the same time by emphasising a lack of time and resources and feelings of inadequacy. When teachers found themselves in situations where they perceived too many student needs at the same time, the primary feedback rationale individualising came into conflict with the primary feedback rationale classwork flowing. The perceived problems of not meeting all students’ needs in the moment were: lack of learning – taking into account the students who had to wait too long to get feedback – and the risk of off-task behaviour and classroom disorder. These were problems that, in turn, evoked the primary feedback rationale without-order-no-learning that motivated the teachers to adopt classroom flowing in the moment before individualising. Thus, the rationale classroom flowing seemed to be often driven by both without-order-no-learning and attempts to meet all student needs in the moment.

Superior and subordinate needs

Classroom management can be defined as ‘establishing and maintaining order in a group-based educational system whose goals include student learning as well as social and emotional growth’ (Emmer and Sabornie 2015, 8). In line with such classroom management literature, teachers were strongly focussed on creating and maintaining classroom order that was conducive to learning: the very core of the feedback rationale without-order-no-learning. Analysis indicated that, although this primary feedback is a need-for-order rationale addressing behavioural needs (counteracting disruptive behaviour and promoting appropriate behaviour in the classroom), a clear category overlap exists. The fulfilment of the behavioural need-for-order was assumed to be a necessary condition for the fulfilment of academic needs. The ideal was a kind of classroom order in which all students were engaged in on-task behaviour: in other words, involved in academic engagement and learning.

The primary feedback rationale without-order-no-learning motivated teachers to adopt classwork flowing and fire-fighting as feedback rationales in order to prevent disorder and off-task behaviour. Thus, they constituted efforts to maintain students’ attention and focus on on-task behaviour in the classroom. The without-order-no-learning rationale also motivated the peer order modelling based on social learning. In addition, when teachers encountered classroom disorder, or perceived in other ways a strong need for constructing classroom order conducive to learning (e.g. interpreted that order was fragile), the without-order-no-learning rationale appeared to overrule feedback rationales such as individualising, academic encouragement, time-for-reasoning, which were aimed at addressing academic needs. Similarly, if the three primary feedback rationales without-order-no-learning, classwork flowing and individualising were perceived to be in conflict with each other in an actual classroom situation, need-for-order rationales and trying to keep all students focused on academic tasks in the moment took precedence over rationales concerning other needs. However, although the without-order-no-learning rationale was a superior rationale concerning behavioural needs, it was in alignment with the teaching task predominantly concerning students’ academic needs, since its aim was to create a better learning environment.

Discussion

This study explored and conceptualised primary school teachers’ rationales for giving students feedback. Our broad approach was based on a symbolic interactionist framework (Blumer 1969; Charon 2007), where knowledge about teachers’ perspectives is crucial in order to better understand their actions in everyday situation, such as feedback.

We examined primary teachers’ main concerns and rationales about their everyday classroom feedback practice. Analysis of their reports allowed us to generate a set of feedback rationales. Among these, three feedback rationales were considered to be the most significant: without-order-no-learning, classwork flowing and individualising. The teachers’ main concerns were to address academic needs as well as emotional and behavioural needs. The findings contribute to the field by offering a framework for the further analysis of teacher rationales in relation to feedback.

The reported feedback rationales highlighted teachers’ awareness of the impact that feedback may have on self-esteem (cf. Gipps 1995). This awareness was also articulated in their desire to be receptive as to how students interpreted and made use of feedback. This receptiveness was often described as most easily fulfilled when there was order in the classroom, creating opportunity to use rationales such as individualising and time-for-reasoning. At the same time, feedback rationales addressing emotional needs were less described and less elaborated on by the teachers as compared with feedback rationales addressing academic and behavioural needs. Thus, the teachers appeared to be more concerned with academic learning and externalising behavioural problems in the classroom than with internalising emotional problems among the students (cf. Lane et al. 2004). In line with Lane et al., our study suggests a need for more elaborated awareness and feedback rationales among teachers regarding students’ emotional needs beyond low self-esteem, since research has found academic achievement to be negatively associated with negative emotions and internalising problems, and positively linked to positive emotions (for a review, see Valiente, Swanson, and Eisenberg 2012).

In the current study, we identified an overlap between classroom assessment and classroom management in respect of teachers’ main concerns and rationales for classroom feedback. We found that without-order-no-learning was the dominant feedback rationale that tended to overrule a range of other rationales when classroom order was fragile or had already become disorderly. Order and on-task behaviour was seen as the norm by the teachers, and was understood as the basic classroom condition for learning. It was assumed that making room for an ideal feedback meant focusing upon academic needs and rationales such as time-for-reasoning and individualising. Classroom management has to be understood as integrated with, and part of, classroom instruction (cf. Muijs and Reynolds 2005). The current study demonstrates how this integration may happen, through its analysis of primary teachers’ rationales for their everyday classroom feedback.

It is noteworthy that in the rationale classwork flowing, the teacher gave short, specific responses and often tended to use ‘steering’ feedback. This was a time-efficient but instructionally scanty approach that was assumed to cope with a lack of time and a large amount of student need at the same time, and construct classroom order conducive for learning (or at least on-task behaviour). This can be compared to the rationale time-for reasoning, which the teachers described as providing for communication on processes such as inquiring and reasoning, but also as something that was not often achievable. This tension between rationales is similar to the two assessment discourses ‘do it quick and do it right’, and ‘reasoning takes time’ reported by Boistrup (2015). Hence, although teachers in the current study had very good intentions of giving individualised, deliberative, meta-cognitive supporting and autonomy-promoting feedback (cf. Hargreaves 2014), the findings indicated that sometimes teachers could not give feedback in accordance with their primary intentions, since they had different needs to take into consideration. In addition, constructing classroom order might, in particular, overrule such intentions in the actual classroom situation.

Limitations and implications for future research and professional development

In this small scale, qualitative study, the small sample size, of course, limits the transferability of these findings. Nonetheless, possible future research on a larger scale should include teachers at other school levels as well as teachers from different countries. Another important limitation is that the findings do not provide information on how teachers actually act on their feedback rationales, but a constructed conceptualisation of how teachers describe their feedback rationales and how they act on them. Students are represented in the study as described by the teachers. Students’ need for feedback as presented is, thus, necessarily an image of how teachers perceive them. Teacher feedback is, particularly from a symbolic interactionist view, a complex interactional pattern between teachers and students, meaning that it is crucial to examine students’ perspectives to better understand teacher feedback. However, social research is always an issue of, and limited to, perspectives (Charon 2007). Larsson (2009) talks about ‘generalisation through recognition of patterns’ in which the reader judges the generalisability. In addition, Glaser (1978) argues that findings are not a fixed end-point, but are constantly open for modification as new data are gathered. In accordance with constructivist grounded theory research (Charmaz 2014), we do not claim to offer an exact picture – but, rather, an interpretative portrayal of the phenomenon studied.

Conclusions

Despite the limitations and cautions considering generalisability mentioned above, we suggest that the findings have implications for teachers and teacher education. The study contributes to the literature as well as to teachers and teacher educators by presenting a systematic conceptualisation of teachers’ rationales for giving students feedback in the classroom. Based on systematic analysis of – and thus grounded in – qualitative interview data with primary teachers, the current study offers a framework for surveying, understanding and discussing teacher feedback. It uses some explicit concepts that maybe helpfully included in teachers’ professional language, in terms of their teaching practices of classroom assessment and feedback. Furthermore, and in line with Torrance and Pryor (1998) and Tunstall and Gsipps (1996), our findings showed that the everyday practices of classroom assessment and classroom management overlapped, which, in turn, underlines the importance of both classroom assessment training (cf. Black and Wiliam 2009) and classroom management training (cf. Dicke et al. 2015) in teacher education. In order to support professional development, teachers and those involved in teacher development may find it valuable to examine the overlaps and integration between them with reference to the rationales and feedback strategies set out and discussed in our study.

Disclosure statement

No potential conflict of interest was reported by the authors.