Introduction

Over the last decade, literature has described the assessment process in the classroom as a dynamic process in which both teacher and student are involved (Christoforidou et al., 2014). In the assessment process, many scholars make a distinction between summative assessment (drawing conclusions about learning at the end of the learning process) and formative assessment (using assessment to guide further learning). Formative assessment (FA) is described in this study as an interactive process between student and teacher to gather, analyse and interpret student responses in learning to make better decisions for follow-up actions (Black & Wiliam, 1998). In formative assessment, interactions between teachers and students are the key to making learning visible for both teachers and students (Carless, 2017). One of the reasons for teachers in secondary education to apply FA in the classroom is to stimulate students’ self-regulated learning (SRL) (Panadero et al., 2020). However, teachers find it difficult to use FA in a way that SRL is stimulated, and such practices—for example, using self and peer assessment—are not often seen in the classroom yet (Kippers et al., 2018; Wolterinck, 2022). A solution to these difficulties of implementing FA to stimulate students’ SRL for teachers could be to use FA to create co-regulated learning with students (Heritage, 2018). Co-regulation in FA means that FA strategies are regulated, not only by either the teacher or by the students themselves, but by joint activities or shared responsibilities in the classroom through interactions between the student and the teacher, peers and the learning context (Allal, 2020). When implemented successfully, teachers and students together regulate learning in the FA process to move learning forward. This joint effort, in turn, offers opportunities to stimulate students’ self-regulated learning (Allal, 2020; Panadero et al., 2020). However, only a few studies have investigated how teachers create co-regulation through FA to stimulate student SRL and, as Panadero et al. (2018a, b) advise, ‘… the intersection between FA and co-regulation deserves more empirical study. […] It is time to study the relationship in classrooms’ (p. 28). More research is needed to uncover how teachers design and implement the interactive FA process in their secondary school classrooms to further help them create co-regulated learning and, thereby, students’ self-regulated learning. This will help teachers to implement more interactive FA that potentially stimulates students’ SRL and will help researchers to better understand the concept of co-regulation in FA and its impact on improving students’ SRL.

FA and Self-Regulated Learning

In many educational theories, such as Vygotsky’s zone of approximal development, learning is described as a process done by the student and guided by the teacher with the aim to stimulate student agency (Andrade et al., 2021; Marshall & Drummond, 2006). For many teachers, guiding students to become agentic self-regulated learners is a goal of using FA (Panadero et al., 2020; Wolterinck, 2022). When students use self-regulated learning, they actively plan, monitor, evaluate and adapt learning to reach learning goals (Heritage, 2018; Zimmerman, 1986). Recent studies have shown that with the use of FA in the classroom, students’ SRL could be stimulated, especially when FA strategies are implemented that focus on self- and peer assessment (Beekman et al., 2021; de Vries, 2022; Panadero et al., 2020). Even more so, in a study of Andrade and others (2021), it was described that FA strategies that activate students and use scaffolding enhance student motivation that is necessary for students to self-regulate their learning. When students are active agents and work together with their peers and teacher in the FA process, their SRL will improve (Heritage, 2018). Still, while FA offers many opportunities to help students improve their SRL, a lot of teachers have difficulties in implementing these FA strategies properly (Veugen et al., 2022; Kippers et al., 2018), and students rarely perceive these FA practices in the classroom (Veugen et al., 2021; Wolterinck, 2022).

FA as Co-regulated Learning

Taking responsibility as a teacher to help and guide students with the use of FA to become more agentic learners is seen as exploiting the spirit of formative assessment (Andrade et al., 2021; Marshall & Drummond, 2006). As our previous study (Veugen et al., under review) indicates, teachers succeed in implementing this spirit of FA when they are able to implement FA that activates and engages students and, sometimes, even is designed together with students. Interaction between teacher and students is necessary to implement FA that encourages students to self-regulate their learning (Carless, 2017). In other words, students and teachers need to use interactive FA that focusses on co-regulated learning, for students to become successful self-regulated learners (Allal, 2020; Heritage, 2018; Panadero et al., 2020). In FA that creates co-regulated learning, students and teachers together monitor, evaluate and adjust learning (Heritage, 2018; Panadero et al., 2020). For example, teachers could stimulate students to assess their own work, to provide feedback to themselves and indicate the next steps in learning. In this way, students learn to compare their own work with the success criteria and stimulate inner feedback processes (Nicol, 2020). The FA processes then become vital for students to regulate their own learning (Andrade & Brookhart, 2020). Students themselves will become better assessors and will become less teacher-dependent. However, teachers must actively design FA that stimulates or prompts students to engage in these co-regulatory activities (Nicol, 2020). How these interactive FA practices that create co-regulation take place, in terms of teacher and student behaviour, in the classroom needs to be further investigated to better understand the processes of co-regulated learning and to be able to improve students’ assessment skills and learning (Allal, 2020; Panadero et al., 2020).

The FA Cycle to Identify Co-regulated Learning

To be able to further identify the interactive FA teacher and student behaviours in co-regulated learning, the formative assessment cycle (FA cycle) has been chosen as a framework. Within the FA cycle, five phases are distinguished that, aligned together, form the FA process (Gulikers & Baartman, 2017): (1) clarifying expectations in learning goals and success criteria; (2) eliciting student responses to collect information about the learning process; (3) analysing and interpreting these responses; (4) communicating with students about the responses; (5) and adapting teaching and learning by taking follow-up actions (see Fig. 1). Each phase consists of teacher FA strategies, describing what teachers could do in the classroom to implement FA, including how to engage students in these phases. Previous research, using the FA cycle as a framework to measure teachers’ FA, provided useful information on which to build on this study. First, previous research has shown that the FA cycle is a helpful framework for mapping the implementation of FA in the classroom (Veugen et al., 2021). Second, former research has shown that the FA cycle helps teachers to formulate FA practice that creates co-regulated learning by explicating first student FA behaviour and then formulating teacher FA behaviour necessary to let students show the student FA behaviour (Gulikers et al., 2021). Third, another study (Veugen et al., under review) showed that teachers implement the FA cycle phases in a more teacher- or student-centred way. With teacher-centred FA, it is meant that the teacher is the most active agent who steers the FA process and the students respond to the FA strategies initiated by the teacher. Student-centred FA, on the other hand, includes the student as the most active agent of the FA process (Heritage, 2018; Veugen et al., under review). In this form of FA, the teacher initiates FA that activates the student to take responsibility for the FA process and learning. This could be student-activating forms of FA, such as letting students self-assess their work with the success criteria or even students designing FA themselves, such as students choosing ways to monitor their own learning. These results indicate that the FA cycle can provide a useful framework for teachers to help them plan and implement FA behaviour that should focus on (mostly) student-centred FA practice to create co-regulated learning. Using the framework, in the present study, we further examine how teachers plan and implement FA behaviours in each of the five phases of the FA cycle in the classroom that creates co-regulated learning and thereby fosters students’ SRL.

Fig. 1
figure 1

The formative assessment cycle (Gulikers & Baartman, 2017) (permission to use from authors)

Full size image

Implementing Planned and Unplanned FA

When using FA in the classroom, it is the teacher who designs and implements the FA strategies that activate and engage students in the assessment process (Boud & Molloy, 2013; Carless & Winstone, 2020). The teacher should think of and design the educational environment in which FA could be implemented before the lesson is given, but should also be able to provide an educational environment in which FA can take place during teaching (Heitink et al., 2016). By doing so, planning FA beforehand helps teachers to think of how to reach the aims they have with implementing FA, such as how to collect information that gives them insights in where learning is to adjust teaching accordingly (Wiliam, 2014). Also, during teaching, teachers need to be able to adjust the learning environment to students’ needs to make sure FA will succeed in helping students to improve their learning (Wiliam & Leahy, 2015). For FA to be effective in, among other goals, stimulating students’ SRL, teachers need to plan and implement the entire FA process with the five phases aligned (Gulikers & Baartman, 2017; Veugen et al., 2021). They need to know what learning goals are set for students, what data about student learning they want to collect and how they want to use these data to improve learning (OECD, 2005; Wiliam, 2014; Wiliam & Leahy, 2015). However, teachers also need to be able to interpret information about student learning as it occurs in practice to adjust teaching and learning accordingly and reach learning goals (Gulikers & Baartman, 2017; Heitink et al., 2016). In our earlier study (Gulikers et al., 2021), we showed how teachers describe teacher and student FA behaviour within the five phases that creates co-regulated learning, but not how teachers plan and implement such practices. The present study will investigate whether teachers are able to plan teacher and student FA behaviour that creates co-regulated learning before teaching and what FA teacher and student behaviour happens (planned or unplanned) during teaching.

Study Context and Research Questions

The present study was conducted in the context of a FA learning network, in which teachers of thirteen different secondary schools in the Netherlands were trained in implementing FA in the classroom between 2018 and 2022. The FA cycle was used as a theoretical model to help teachers make FA more concrete and to implement the FA strategies of the five phases into classroom practice. At the time of this study, teachers in this network were already able to implement FA strategies of all five phases of the FA cycle (Veugen et al., 2021). However, some teachers wanted to improve their FA practice even further towards more actively engaging students and stimulating their SRL. Therefore, in the present study, these teachers further professionalised in how to use FA in their classroom that would create co-regulated learning and students’ SRL. Two schools with three teams of the learning network were chosen as the context of this study. The study answers the following research questions:

  1. 1.

    How much planned and unplanned teacher and student FA behaviour of the five phases of the FA cycle is observed in the classroom, when FA is implemented to stimulate students’ SRL?

  2. 2.

    What does teacher and student FA behaviour look like that creates co-regulated learning and stimulates students’ SRL?

Method

Design

In this qualitative study, teacher and student FA behaviour that creates co-regulated learning was researched in the classroom. The variables of this study were the FA behaviour of teachers and students, the FA behaviour that shows co-regulation and the SRL behaviour of students in FA. Teacher and student FA behaviours in our study are defined as behaviours that fit one of the five phases of the FA cycle (see Fig. 1). Teacher FA behaviour includes certain FA strategies in each of the five phases, as previously defined in earlier studies (Gulikers & Baartman, 2017; Veugen et al., 2021). Teachers could plan to use these strategies in their lesson beforehand or apply these strategies unplanned during the lesson. FA student behaviour is also described from these five phases and includes similar FA strategies to those of the teacher, only from a student perspective (Gulikers et al., 2021). FA as co-regulated learning was defined and measured as the instance where both teacher and students regulate learning by using activities in the FA process and the teacher and student FA behaviours matched, for example, when a teacher stimulates students to analyse each other’s work and students analyse their peers’ work. With SRL behaviour of students in FA is meant that students show FA behaviour that includes planning, monitoring or evaluating their own learning. To measure this, we determined if the observed FA behaviours of teachers and students that create co-regulation were more teacher or student-centred, for student-centred FA will stimulate SRL of students more than teacher-centred FA (Veugen et al., under review). We then checked if the student-centred FA practices included student behaviour that showed students planning, monitoring or evaluating their own learning.

Procedure

In this study, three teacher groups of two secondary schools, who participated in the FA learning network, implemented FA behaviours as co-regulated learning. The three teacher groups took part in two professionalisation sessions to help teachers design and implement interactive FA as co-regulated learning to stimulate students’ SRL. At the start of the period, moments were set for the sessions to take place at both schools between October 2021 and February 2022. All teachers were asked to fill out a consent form before the study was conducted, in which they provided their consent to use their notes and record audio fragments during the sessions. Teachers participated in the two sessions. In the first session, teachers were asked to think of what student behaviour they would expect of students in each of the five phases of the FA cycle that showed SRL and what teacher FA strategies were needed to stimulate this behaviour (also see Gulikers et al., 2021). Using their collaboratively discussed ideas on expected student behaviour and required teacher behaviour, all teachers planned FA as co-regulated learning for a certain lesson. In doing so, they explicated the kind of teacher and student behaviour they planned to implement and wanted to see in their lesson. Between the two sessions, teachers implemented their own planned FA strategies in a lesson. A teacher colleague, who also participated in the study, observed the lesson and both teachers evaluated the lesson afterwards. In the second session, teachers reflected on the given lessons and observed FA strategies to discuss whether and how co-regulated learning occurred and stimulated SRL of students. To answer the research questions, the planned and unplanned teacher and student FA behaviour was qualitatively analysed based on the lesson plans, observation notes and audio fragments of the training sessions.

Participants

The participants included fourteen secondary school teachers of two schools who participated in the FA learning network. Of School A, two teacher groups took part in the professionalisation sessions, and of School B, one teacher group took part. In the first group of School A, seven teachers participated; in the second group of School A, three teachers participated; and in the teacher group of School B, four teachers participated. All teachers of this study were already familiar with the five phases of the FA cycle and the strategies within these phases. In the total group, eight teachers were male and six teachers were female. Teachers taught different subjects and different grades. In secondary education in the Netherlands, four education types exist: practically-oriented pre-vocational education [Dutch abbreviation: vmbo-bg]; theoretically oriented pre-vocational education [vmbo-tl], senior general secondary education [havo]; and pre-university education [vwo] (Government of the Netherlands, 2020). The types of education of School A were havo and vwo, and the types of education at School B were vmbo-bg and vmbo-tl.

Instruments and Gathered Data

In this study, data was collected during the professionalisation activities via recording audio of the two training sessions and collecting lesson plans and observation notes of teachers. The first step in professionalisation was for teachers to formulate teacher FA strategies that could be used to stimulate students’ SRL and what student behaviour in FA they wanted to enact and would show SRL for each of the five phases of the FA cycle. These collectively formulated descriptions of teacher and student FA behaviour were added to a lesson plan format by the researchers, in which teachers would describe their planned lesson. Teachers chose what FA teacher behaviour they wanted to use and what student FA behaviour they wanted to see in their planned lesson and noted these in their lesson plans. These FA behaviours were described in this study as the planned FA teacher and student behaviours. The teachers then carried out their lessons, and a teacher colleague studied the lesson plan and observed the lesson. The teacher who observed the lesson noted what FA teacher and student behaviour he or she saw happening in the lesson on a worksheet. After the lesson, the teacher and the observing colleague reflected together on whether the planned teacher and expected student FA behaviours were present in the lesson or not and whether other unplanned FA behaviours were used and seen in the lesson. They noted their reflections on the worksheet. These worksheets were gathered and used as data to analyse what FA teacher and student behaviour happened in the classroom. Finally, in the second training session, teachers reflected on what happened in their own and each other’s lessons regarding FA and discussed what FA practices stimulated students’ SRL. The audio recordings of this session were transcribed and used as data to analyse what FA teacher and student behaviour was implemented in the classroom that showed co-regulated learning and stimulated SRL according to the teachers.

Analyses

To answer the research questions, the observation notes and audio fragments in which teachers shared their observed and implemented FA behaviours were analysed. First, the planned teacher and student FA behaviours, as written in the lesson plan, were collected per teacher. Then, the observed teacher and student FA behaviours, as noted by the colleague observer on the observation sheet and discussed by the teacher and observer in audio fragments, were collected per teacher. The data were then divided into fragments of one or two sentences describing FA behaviours of teachers and students in one of the five phases of the FA cycle. The first two authors analysed each fragment to uncover what FA behaviour was described, to which phase of the five FA cycle phases this behaviour belonged and whether this behaviour was planned by the teacher or not. To answer Research Question One, the planned and unplanned FA teacher and student behaviours, which were found in the teachers’ data, were noted per teacher in the Appendix. Each FA behaviour was noted only once per teacher, meaning that certain FA behaviours could have occurred more than once during the lesson. To answer Research Question Two, first, the FA behaviours that created co-regulated learning were determined, and then it was analysed which of these behaviours also showed student-centred FA. FA as co-regulated learning was measured when FA teacher and student behaviour was implemented that showed shared responsibility and matched each other, for example, when the teacher stimulates students to formulate success criteria and students formulate success criteria (Phase 1). If the implemented teacher FA behaviour did not match, or did not occur with, the observed student FA behaviour, there was no FA practice as co-regulated learning noted.

Results

Research Question One: Planned and Unplanned FA Behaviour of Teachers and Students

In the total dataset, there were 129 fragments coded that showed planned and unplanned teacher FA behaviour in the datasets of the fourteen teachers. These coded fragments described 31 FA teacher behaviours (see the Appendix). Of the total set of fragments, there were 43 fragments (33%) that included teacher FA behaviours that were planned in the lesson plans by teachers and also observed in the lesson by their colleague. There were some FA teacher behaviours (9 fragments, 7%) that teachers planned in their lessons that did not occur in practice. The remaining FA teacher behaviour fragments (77 fragments, 60%) described unplanned FA. These FA behaviours were not noted down by the teachers as planned FA on the lesson plans, but did occur in classroom practice according to observer notes and reflections of teachers in the audio fragments. In Fig. 2 is illustrated how these teacher FA behaviour fragments were divided over the five phases and what FA behaviour was planned and unplanned per phase.

Fig. 2
figure 2

Division of fragments (f) of planned and unplanned implemented FA teacher behaviour over the five phases

Full size image

In datasets of the teachers, 114 fragments were made of planned and unplanned FA student behaviour. These fragments described 26 different FA student behaviours (see the Appendix). Of these fragments, there were 32 (28%) that described student FA behaviour that was planned and implemented during the lesson. Some fragments (18 fragments, 16%) of FA student behaviour were planned in the lesson plans, but were neither observed during the lesson by the observing colleagues nor mentioned as implemented during the lesson by the teachers in the audio recordings. The majority of fragments (64 fragments, 56%) described FA student behaviour that was not planned (i.e. not described in the lesson plan), but was implemented in the classroom. In Fig. 3 is illustrated how these student FA behaviour fragments were divided over the five phases and what FA behaviour was planned and unplanned per phase.

Fig. 3
figure 3

Division of fragments (f) of planned and unplanned implemented FA student behaviour over the five phases

Full size image

Comparing the FA teacher and student behaviours, both patterns showed that most behaviour was unplanned and implemented. Moreover, FA strategies regarding gathering student responses (Phase 2) were described the most in both teacher and student FA behaviour. The difference between these two divisions in behaviours was that teachers planned and implemented FA behaviour with respect to adapting teaching and learning the least (Phase 5), while for students this holds for clarifying expectations (Phase 1).

Research Question Two: FA to Create Co-regulated Learning

To find out if teachers used FA as co-regulated learning in their lessons to stimulate self-regulated learning, we analysed per teacher if the teacher and student FA behaviour matched per phase. The behaviour of FA was matched when teachers and students jointly took responsibility for the FA strategy used in a phase of the five phases of the FA cycle, for example when the teacher stimulates students to ask questions and students ask questions (Phase 2). This comparison led to the insight that thirteen of the fourteen teachers succeeded to implement FA as co-regulated learning. On average, teachers implemented four different FA behaviours in the classroom that created co-regulation, with the lowest being zero behaviours (teacher two) and the highest being nine different behaviours (teachers five and thirteen). In total, twenty different FA behaviours that showed co-regulation were observed and are listed in Table 1. Teachers and students mostly used FA behaviours to create co-regulated learning in Phase 2 (25 behaviours) and the least in Phase 5 (4 behaviours). The most commonly used FA behaviour that created co-regulated learning (n = 7) related to Phase 3: the teacher stimulating the students to analyse each other’s learning or work and the students analysing the learning or work of their peers (see Table 1). In general, as can be seen in Table 1, there was no clear pattern in how teachers implemented FA as co-regulated learning. In other words, the FA behaviours that created co-regulated learning as observed in the classroom seemed to vary between teachers and the FA practice of one teacher could be totally different compared to another teacher. In the end, 114 fragments of the total of 243 fragments (47%) in the total dataset of teacher and student FA behaviour described FA as co-regulated learning.

Table 1 Observed teacher and student FA behaviours that created co-regulated learning
Full size table

The goal for teachers was to use FA as co-regulated learning to stimulate students’ SRL. To analyse whether and how teachers were able to stimulate SRL student behaviour with FA as co-regulated learning, we first analysed if the FA behaviours described more teacher-centred or student-centred FA. As could be seen in Table 1, thirteen of the twenty FA behaviours were categorised as student-centred FA. In these student-centred FA behaviours, teachers initiated FA that made the student the most active agent in the process and let them plan, monitor and evaluate learning in the FA process. Students were stimulated by their teachers to formulate learning goals (Phase 1); to ask questions, provide answers, help others, discuss learning and choose ways of learning (Phase 2); to analyse their own and peers’ learning (Phase 3); to formulate feedback, receive feedback and interpret feedback (Phase 4); and to think of follow-up actions in learning (Phase 5). The other seven FA behaviours that created co-regulated learning were more teacher-centred, meaning that the teacher mostly steered the FA process and the students responded to what was asked of them to do. In these teacher-centred FA behaviours, students were actively involved in the FA process but were not specifically stimulated to take responsibility for planning, monitoring and evaluating their own learning. In these behaviours, teachers clarified expectations (Phase 1), asked questions, observed learning, provided tasks to students, collected work (Phase 2) and provided feedback to students (Phase 4).

We will now provide an example to further illustrate what FA that created co-regulated learning looked like in the classroom. In this example, the teacher of a second-year havo gymnastics class created groups of three students for a climbing exercise: one student to climb, another student to control the ropes and a third student to observe how the exercise was acted out. The colleague who observed this lesson noted that students were really taking ownership for their own and peers’ learning by listening carefully to the instructions of the teacher, so they understood the expectations and instructions of the task, and then during the task, they helped each other by providing peer feedback. This colleague mentioned that at a certain point during the lesson, he observed that the climbing student was stuck and the partner on the ground with the ropes did not know how to help the student anymore, then the observing student ‘… knew what to do and helped them by saying “you should do this and that”’, before they asked help from their teacher. This is an example showing both teacher-centred and student-centred FA behaviours that created co-regulated learning. First, a more teacher-centred FA behaviour was that ‘the teacher clarifies and discusses expectations (learning goals and success criteria) and students listen or read the expectations’ (Phase 1). In the interaction between the teacher and the student in this FA behaviour, the teacher steered the students to understand the expectations. Then the more student-centred FA behaviours followed: ‘the teacher stimulates students to/students help each other with their learning or work together’ (Phase 2), ‘the teacher stimulates students to/students analyse each other’s learning or work’ (Phase 3), ‘the teacher stimulates students to/students give peer feedback’ (Phase 4) and ‘the teacher stimulates students to/students think of a follow-up action by themselves’ (Phase 5). In this example, the teacher (and the observing colleague) observed how the students helped each other learn. The student with the observer role analysed the situation, provided feedback and thought of a follow-up action for his peers. The students were doing this because the teacher initiated this situation to happen—he planned for students to help each other learn, to monitor and evaluate their own and each other’s learning without directly asking help of the teacher. Therefore, students were the most active agents in the FA process, who planned, monitored and evaluated their own and each other’s learning, thus showing SRL behaviour.

Discussion

This study examines what interactive FA to create co-regulated learning can look like in a secondary school classroom. It studies what student and teacher FA behaviours teachers (who wanted to stimulate students’ SRL) planned and implemented in their classrooms (RQ1) and how these FA behaviours reflected interactive co-regulated learning (RQ2). Regarding Research Question One, results showed that teachers mostly displayed unplanned FA behaviour (60%) and also observed mostly unplanned FA student behaviour (56%). Most of the FA behaviour that was planned was also implemented and observed. Regarding Research Question Two, the conclusion was that thirteen of the fourteen teachers used FA behaviour as co-regulated learning in one or more of the five phases of the FA cycle, resulting in twenty different FA behaviours that created co-regulated learning (see 1 in the "Results" section). Of these twenty FA behaviours, thirteen behaviours were more student-centred, and the other seven behaviours were more teacher-centred. In the student-centred FA behaviours, students were planning, monitoring and evaluating their own and/or each other’s learning process with the support of their teachers, which are processes related to students’ self-regulating learning (Clark, 2012). An example of this student-centred FA was ‘The teacher stimulates students to/students give peer feedback’ (Phase 4). In the more teacher-centred FA behaviours, teachers mostly steer the FA process, for example by clarifying learning goals, while students listen. In this form of FA as co-regulation, teachers and students co-regulate learning, but the teacher is the most active agent, and the students are more responding to what is asked of them by the teacher. Therefore, this study shows that teachers, who were already used to implementing FA in the classroom, were able to implement FA practice as co-regulated learning and potentially stimulate students’ SRL.

Planned and Unplanned FA Practice

The first aim of this study was to further investigate how teachers would plan and implement FA as co-regulated learning in terms of teacher and student behaviour. The teachers in this study were shown to be able to implement their planned FA, as well as to use unplanned FA, which they mostly displayed in their practice. Previous studies have shown that it is important for teachers to plan FA strategies for all five phases of the FA cycle to implement the full FA process (Gulikers & Baartman, 2017; Veugen et al., 2021) and that the framework of the FA cycle also helps teachers to formulate FA as co-regulated learning in terms of teacher and student FA behaviour (Gulikers et al., 2021). Other research underlines the importance for teachers to consciously plan FA before teaching, to decide what they want students to learn (setting learning goals), to know what data they want to collect and how they want to use this data to adjust teaching and learning (OECD, 2005; Wiliam, 2014). On the other hand, it is also important to apply FA more ‘on the fly’ and for teachers to be able to adjust teaching to students’ needs during FA (Heitink et al., 2016; Nieminen et al., 2021). Often, these more short-cycle FA processes continuously take place in teaching and are important to move learning forward (Wiliam, 2006; Wiliam & Leahy, 2015). Teachers in this study have shown to be able to adjust teaching and learning directly during the lesson to the needs of their students, which is essential for FA to be used to its full potential (Gulikers & Baartman, 2017; Heitink et al., 2016). The fact that teachers mostly used unplanned FA could indicate that teachers seem to use FA as a daily teaching method. An explanation for teachers to be able to do so is that teachers in this study participated for already several years in a FA learning network, in which they were professionalised in applying the five phases of the FA cycle. If our argumentation holds true, it would confirm that long-term professional development helps teachers to implement various FA practices, such as planned and unplanned FA, into daily teaching. Teachers who apply mostly unplanned FA in the classroom could also indicate that teachers are not always aware of the FA that they actually use in practice. Still, when teachers plan FA, they almost always implement the planned FA practice, indicating that teachers are able to consciously plan and implement various FA practice.

Stimulating SRL of Students Through Co-regulated Learning in FA

The second aim of this study was to further investigate what FA as co-regulated learning looked like in the classroom and how this FA practice had the potential to stimulate students’ SRL. This study offers further knowledge about the process of FA stimulating co-regulation and how the FA behaviour that shows SRL behaviour in the classroom could be better understood (Allal, 2020; Panadero et al., 2020). FA as co-regulated learning occurred when teachers and students regulated FA together and their FA behaviours matched. The underpinning of this was that teachers were able to implement FA in the five phases that activated students and made them participate in the practice of FA. Through the implementation of student-centred FA as co-regulated learning, in which students actively participate in goal setting, monitoring, evaluation and adjusting their learning (that is, the various phases of the FA cycle), student SRL is potentially stimulated (Allal, 2020; Clark, 2012; Zimmerman, 1986). Teachers in this study show to be able to implement these student-centred FA behaviours, for example by letting students analyse each other’s work (Phase 3) and providing peer feedback (Phase 4). This confirms the findings of our previous study, in which it was found that teachers vary in using mostly teacher-centred FA strategies to using student-activating FA strategies (Veugen et al., under review). A key process of stimulating SRL via FA is that students are stimulated by their teachers to be active agents in the assessment process (Heritage, 2018). Only then can FA become a vital form for students to self-regulate their learning and can FA be used in its spirit (Andrade & Brookhart, 2020; Andrade et al., 2021; Marshall & Drummond, 2006). This study shows what FA teacher and student behaviours to create co-regulated learning can be used in practice to achieve this goal.

Practical Implications

Several important practical implications could be identified from this study. First, it is important for teachers to both use planned FA as well as unplanned FA in the classroom. With planned FA, teachers will more consciously implement all five phases of the FA cycle and FA that stimulates co-regulated learning. With unplanned FA, teachers are able to steer the FA process directly to meet students’ needs and help them further in their (co-regulated) learning. When implementing unplanned FA, teachers need to pay special attention to complete the process of communicating about student responses (Phase 4) and taking follow-up actions (Phase 5) to continue the process in (re)setting expectations (Phase 1) (Gulikers & Baartman, 2017). A second implication of this study is the added value of letting a colleague observe FA practice in the classroom, as well as teachers reflecting afterwards on their implementation of FA in their lesson, to make teachers more aware of the FA practice they use in the classroom and to more consciously implement FA to improve FA practice. This is an important realisation and potential driver to make the implementation of FA in classrooms more feasible and more interactively co-regulated. A third implication of this study is that this study offers concrete examples of FA strategies for teachers who want to use more FA in their classroom to stimulate students’ SRL by explicating a variety of FA practices that created co-regulated learning, such as letting students analysing each other’s learning or letting students provide feedback to each other (see Table 1 in the "Results" section).

Limitations and Future Research

For this study, certain limitations must be taken into account. The first limitation is that the participants were a preselected group of teachers who already had prior experience and skills in using FA in the classroom. When teachers without previous experience in how to implement FA would do the same training in this study, this might result in teachers using more teacher-centred forms of FA. Teachers need time to grow in their FA practice from using mostly teacher-centred FA towards student-centred FA (DeLuca et al., 2012). However, the question is whether it really does matter how much professionalisation in FA and time to develop FA practice a teacher has had for FA practices to differ between teachers. This question has not been researched yet. Furthermore, future research could further uncover whether FA behaviours that create co-regulated learning differ in primary and higher education, for example, if in higher education more student-centred forms of co-regulated FA take place, since students are expected to already be more self-regulatory.

A second limitation is the size of the sample of participants in this study. Although the fourteen teachers did show many similar implementations of FA behaviours, thus indicating an overlap in FA practice and less variety, more (varied) participants of different schools (outside the learning network) would have made the results stronger.

A third methodological limitation of this study is that the conclusions of this study were based on the observations and reflections of the teachers themselves. If we had also used more objective classroom observations done by researchers themselves this could have added to the strengths of the findings. On the other hand, the method used in this study, in which teachers observed each other’s lesson, had an important impact on teachers’ learning about FA in that it helped teachers to realise that more FA happened in the classroom than that they themselves would have noticed. Furthermore, by using teachers’ own reflections on their lessons and observations of their colleagues, teachers became more aware of the FA practice they used in the classroom leading to more fragments of FA practice in the data of teachers’ reflections. Also, teachers might have felt less pressured by having a colleague observing their lesson rather than a researcher, keeping their teaching practice as close to normal everyday practice (Munson, 1998).

A fourth also methodological limitation is that the qualitative analyses were done by two researchers and conclusions were based on the determination of how these two researchers defined the FA practices that resulted in co-regulated learning. In other words, the study helped to further determine what FA practice is that stimulates co-regulated learning. There were no earlier specific definitions in other research of what these FA practices looked like. To strengthen the conclusions of this study, the analyses done in the study could be repeated by other researchers to see if the same conclusions are found of what FA practice is that stimulates co-regulated learning. Another implication for future research to strengthen the results of this study is by using more quantitative measurements by using for example teacher perceptions on how their FA practices stimulate co-regulated learning via a questionnaire.

Last, the focus of this study was on the implementation and development of FA practice by teachers and less on the impact of these FA practices on the results of students. Future research could also improve student participation in research, for example by letting students reflect on provided lessons. Still, this study did let teachers reflect on student behaviour and therefore some conclusions about what students do when they use FA in the classroom were drawn. In conclusion, this study has shown that teachers can implement various FA practices as co-regulated learning in their classrooms. Some of these practices were more teacher-centred, while most were more student-centred, activating students to plan, monitor and evaluate their learning. As such, these more student-centred FA practices that created co-regulated learning have the potential to stimulate students’ SRL. Thus, this study contributed to the body of literature on SRL and FA, making much more tangible how FA can stimulate SRL, via implementing co-regulated learning practices in FA.