Students’ independent requests for feedback during collaborative productive work in their mother tongue subject

Feedback that supports students’ learning and self-regulation has mainly been investigated in a teacher-led feedback practice. This article investigates how students’ independent requests for feedback during collaborative productive work in their mother tongue subject may be identified and characterised. The study was conducted as an ethnographic case study in the context of Danish compulsory education and video recordings were used to observe students’ production processes. The results show that requests for feedback extend beyond the goals set by the teacher, and thus beyond the criteria. A request for feedback may appear implicit in a dialogue or explicit as a question. The students ask for feedback regarding their tasks, processes and self-regulation. This article calls for paying greater attention to students’ goal setting in the study of feedback processes.


Introduction
Extensive research emphasises the positive connection between feedback and learning (e.g. Black & Wiliam, 1998;Hattie & Timperley, 2007;Kluger & DeNisi, 1996;Shute, 2008). The development of feedback practices at all educational levels has mainly been investigated in activities initiated and organized by the teacher. This study centres around feedback activities initiated by students by investigating how students' requests for feedback during productive work should be identified and characterised. The empirical data comprise video-recorded observations of lower V. Christensen 46 secondary students' independent work in the mother tongue subject. These observations were conducted as part of a larger study investigating feedback processes in students' development of multimodal reading and writing competence.
In the following section, I present some results of previous research on feedback in general, before focusing on the results of criteria-based feedback and results that connect feedback and self-regulation, particularly in writing and text production.
Understanding feedback Feedback is defined as information given to a learner regarding his or her actual performance or understanding related to learning goals, and used by the learner to improve performance (Black, Harrison, Lee, Marshall & Wiliam, 2004, p. 16;Hattie & Timperley, 2007, p. 87;Sadler, 1989, p. 120). From a long-term perspective, good feedback supports students' ability to provide feedback to themselves i.e. to become self-regulated learners (Black et al., 2004;Hattie & Timperley, 2007;Sadler, 1989, p. 122). Useful feedback answers three questions: 1) Where does the learner have to go? 2) How is he or she doing? 3) Where does he or she have to go next? (Black & Wiliam, 2009, p. 8 ;Hattie & Timperley, 2007, p. 87). Feedback may be provided by various agents, such as teachers, peers, the student themself, books and other learning resources (Hattie & Timperley, 2007, p. 81).
Criteria-based feedback The positive impact of feedback is closely related to the existence of goals or success criteria for the task and the learning outcomes, students' knowledge and understanding of the goals or success criteria and teachers providing feedback on these goals or success criteria (Hawe, Dixon & Watson, 2008;Timperley & Parr, 2009). Furthermore, the quality of the feedback that the teachers offer affects the students' learning (Parr & Timperley, 2010). However, students use only a small part of the feedback offered, because they may not understand the information as feedback, they may not understand the content of the feedback or they may not find the feedback relevant to the task and their learning (Carless, 2006;Christensen, 2015;Higgins, Hartley & Skelton, 2001). Though research indicates that students learn as much from feedback from peers as from teachers (Gielen, Tops, Dochy, Onghena & Smeets, 2010;Karegianes, Pascarella & Pflaum, 1980), a lack of faith is associated with peer feedback (Gielen, Tops et al., 2010). Scaffolding is recommended to ensure the feedback recipient's confidence in the feedback received and the quality of the feedback provided by peers, for example, as training in criteria-based feedback or in the form of rubrics (Gan, 2011;Gielen, Peeters, Dochy, Onghena & Struyven, 2010).
Though there is reliable evidence of the positive relationship between goal-oriented feedback and the recipients' performance or learning, there are those who express reservations. While reporting on lack of involvement if students have no influence on the goal setting, Murtagh (2014) also observes students' becoming 47 over-dependent on clear goals. Winstone, Nash, Parker and Rowntree (2017) mention the need for research on the learners' engagement with the feedback they receive, which is currently underrepresented in the literature (Winstone et al., 2017, p. 17). Sadler (2015, p. 1) has reservations about presenting judgment criteria before an assessment, because the students themselves should be able to make an appraisal. Sadler (2015, p. 9) suggests that students work with an open set of criteria, invoke the criteria relevant to the current assignment and discuss their conclusions with peers and their teacher.  suggest starting feedback sessions with the learner asking a question. Before providing criteria-based feedback, the peers answer the question. Initial questions were used during a peer feedback session in the large project of which this study is a part. Results show that, on the one hand, the students found it difficult to ask this question, and on the other hand, they used the feedback they received in response to the question in their ongoing work (Christensen, 2015).
Feedback and self-regulation Feedback and self-regulation may be investigated from psychological and pedagogical perspectives. Psychologically, feedback is understood as inherent in and a determiner of self-regulation (e.g., Butler & Winne, 1995). Butler and Winne's model of the cognitive system's role in academic tasks includes monitoring knowledge of domains and strategies, motivation and belief, goal setting, strategies used and the outcome of a task. The internal monitoring results in internal feedback on all aspects, and subsequent use of strategies emphasise personal agency (Zimmerman, 1989(Zimmerman, , 1990. Hattie and Timperley (2007) distinguish between feedback on tasks, processes, self-regulation and self (cf. the more elaborated presentation of feedback levels in the following section). Most feedback -more than half -is provided at the task level, less is provided at the process level, 5-10% at the self level and 1-2% at the self-regulatory level (Hattie, 2012, p. 123). Black and Wiliam's (1998) review of formative assessment in the classroom accentuates the importance of feedback and self-assessment. They find that students who monitor and regulate their work themselves gain more than students working with other sorts of feedback and that self-evaluation and reflection on one's own learning enhances learning. With respect to writing development, Graham and Harris (2000) find that developing self-regulatory strategies, including monitoring, evaluation and revision, are important for becoming a skilled writer. Furthermore, self-regulation increases with age and schooling, and teaching self-regulatory strategies accelerates this development.
To summarise, research on feedback emphasises a positive connection between criteria-based feedback and student learning. Furthermore, feedback regarding self-monitoring and self-regulation is more supportive than other sorts of feedback but less frequent. There are those who express reservations with respect to students' uptake and use of the feedback, their engagement and agency. However, some results indicate that the use of feedback increases when the students are given agency by asking the initial feedback question. We do not know whether and how they request feedback when they work independently, and this article contributes to filling the knowledge gap by asking the question: How should students' independent requests for feedback during productive work be identified and characterised?
Theoretical framework Wiliam (1998, 2009) focus on formative assessment in the classroom. They describe five strategies in a model of assessment for learning, one of which is feedback moving the learner forward. Along with Harrison, Lee and Marshall, they stress that feedback functions formatively only if the feedback information is used by the learner to improve his or her performance (Black et al., 2004, p. 16). Furthermore, Black and Wiliam describe the learning process as an interaction between external stimulus and feedback and internal production (Black & Wiliam, 2009, p. 11). Other three of the strategies in the model are of special interest in preparing to identify students' independent requests for feedback. The first is the teacher's responsibility for clarifying 'learning intentions' and success criteria, the second is the students as learning resources for each other and the third is the students' ownership of their learning (Black & Wiliam, 2009, p. 8). Students are included as owners of their learning and as learning resources for each other in Hattie and Timperley's feedback model (2007) as well in their description of the feedback agents. According to Hattie and Timperley, feedback is given in relation to goals and/or criteria and receiving feedback provides the opportunity for students and/or teachers to set more challenging goals (Hattie & Timperley, 2007, p. 88-89).
As mentioned in the introduction, Black and Wiliam and Hattie and Timperley agree on good feedback answering three questions but differ with respect to how specific the direction of the learning in the first question must be expressed. Hattie and Timperley emphasise the need for specific goals, whereas Black and Wiliam use the broader term 'learning intentions'. Alternatively, Sadler (2015) suggests an open set of general criteria. In this study, the three understandings are conceptualised as 'embedded'. The specific goals indicate the desired and planned learning, whereas the term 'learning intentions' acknowledges adjustments and supplements when meeting students in practice. The second feedback question, regarding the students' actual performance, involves self-evaluation when assessing the actual level of understanding. The third question, concerning how to proceed, means helping students execute the actual task or reach their actual goals (Hattie & Timperley, 2007) and leads to new goals (Hattie, 2009).
According to Hattie and Timperley's model the three questions are answered at four levels: the task level, process level, self-regulation level and self level.
Feedback at the task level includes information about the quality of the performance of the assignment, for example, knowledge and correctness. Feedback at the process level focuses on the work process, and includes knowledge of how to complete a task, for example, detecting errors, using learning strategies or finding alternative strategies if stuck. Feedback at the self-regulation level 'addresses the way students monitor, direct, and regulate actions toward the learning' (Hattie & Timperley, 2007, p. 93). Whereas the previous levels focus on a specific task with specific content, the self-regulation level addresses learning in a more abstract sense. Feedback at the self level is personal and directed at the person, rather than the task or learning (Hattie, 2009). Because of the uncertainty regarding the influence of feedback at this level, it is excluded from the following account. In the analysis below, feedback is categorised according to the other three feedback levels to offer insight into the nature of the requested feedback.
The connection between the external feedback and the internal mental processes mentioned above is addressed by Nicol and Macfarlane-Dick (2006). They offer a model connecting the external feedback provided by teachers, peers or other sources with internal feedback processes. The internal feedback is not a simple translation of external feedback. Initially, the students interpret the externally stated task and/or the goals and criteria based on their domain knowledge, strategy knowledge and motivation and beliefs. This work results in student internal goal setting, which shapes the tactics and strategies employed to generate internal learning. The internal learning leads to an externally observable outcome. The internal work is monitored by self-regulatory processes which provide internal feedback. The model distinguishes between internal and external learning outcomes. The external sign is the only visible element and is, in continuation of Nicol and Macfarlane-Dick's ideas, the pivotal starting point in the process of identifying and analysing student-initiated feedback-seeking.
The general definition on feedback in the introduction requires two conditions for categorising information as feedback: 1) The information is given with the intention to be feedback and 2) the information is used by the learner to improve performance or understanding. Feedback initiated by the giver implies a risk for not being perceived as feedback by the recipient, which makes it difficult to categorise the information as feedback according to the definition. When the initiator of feedback is the learner, the two conditions are present in the learner at the beginning. The question is how to distinguish requests for feedback information from other forms of communication. Asking a question is an explicit method of feedback-seeking. But feedback-seeking might occur implicitly in dialogue as well. Mercer's (1996Mercer's ( , 2004 categories of talk describe various sorts of dialogue. 'Cumulative talk' occurs when speakers build 'positively but uncritically on what the other has said' (Mercer, 1996, p. 369). Cumulative talk offers an opportunity to observe how knowledge is collaboratively constructed. 'Exploratory talk' happens when 'partners engage critically but constructively with each other's ideas' (Mercer, 1996, p. 369). Exploratory talk adds a meta-reflective dimension, as the academic content or understanding is also discussed.
The explicit framing of feedback which comes with teacher-led feedback activities is absent in the students' independent feedback-seeking, thus the feedback information needs to be delimited. Adie, van der Kleij and Cumming (2018) distinguish between feedback conversations and feedback messages but the former is too vague to operationalise and the latter not enough to identify feedback. I establish the intermediary category feedback incidents. Feedback incidents are units of communication aiming to support learning as part of an extended dialogue (cf. Mercer, 1996). The dialogue makes it possible to distinguish an initial question as a request for feedback.
To summarise, feedback is information regarding the learner's performance or understanding for the purpose of being used to improve the performance or understanding. A request for feedback is an external sign of internal work (Nicol & Macfarlane-Dick, 2006). The feedback-seeking may be expressed explicitly by asking a question or it may be concealed in dialogue (Mercer, 1996). Similarly, the goals for the learning associated with the request for feedback may be explicitly expressed in the question or implicitly occurring in the dialogue (Black & Wiliam, 2009;Hattie & Timperley, 2007;Sadler, 1989). Feedback may be requested at all feedback levels (Hattie & Timperley, 2007).

Context and method
Research context This study is part of a larger project that investigated planned feedback and students' spontaneous requests for feedback on productive work with multimodal texts (Christensen, 2015). The research was carried out at the Danish lower secondary school in two 8 th grade classes (students aged 14 to 15) and focused on two highperforming and two low-performing students in each class (Christensen, 2015). The data for this study comprises video recordings of two high performing girls' productive work.
The class was involved in a course on multimodal texts, which was a continuation of a course on advertising. A few months before this course, the researcher observed the same class during a writing course to get some idea of the feedback practices in the class. The multimodality course was designed by the teacher, who discussed it with the researcher and subsequently adjusted it based on former observations of feedback practices. The class analysed websites before producing a mock-up. 1 Through the work on the mock-up assignment, the students were to be able to: • argue for the content they selected, with reference to the envisaged target group • give reasons for their choice of modes, drawing on their knowledge of modal affordances • demonstrate skills related to and reflections on layout and aesthetics Design and participants The methods of ethnographic classroom research were applied (Atkinson & Hammersley, 1994;Erickson, 2011;Gudmundsdottir, 1998;Klette, 1998). Video recordings were used for observation, and video-recorded data was supplemented with field notes and documents. All the students in class agreed to participate in the project and obtained their parents' consent, as all the activities in the Danish class were observed over a period of three weeks. A pair of high-performing and a pair of lowperforming readers were chosen as focus students. They were identified by the teacher based on their results on a national reading test. From each group, the teacher recommended two students that she expected would enjoy working together. These four students agreed to videotape their independent work on the assignment for research only. Content logs were produced, selected parts of the recordings were transcribed, and all participants were anonymised. The large project included four cases. This study focuses on one of the four cases: two high-performing girls named Sharon and Sylvia.
The researcher had the role of participating observer, and the video camera allowed the researcher to distance herself from the participants (Erickson, 2006(Erickson, , 2007Jewitt, 2012). Video-recording provided the opportunity to follow two groups working simultaneously on the assignment. Being positioned behind the camera also allowed the researcher to generally leave communication and guidance to the teacher although occasionally being asked for guidance by students. Recordings of the two girls' independent work in school lasted 8.5 hours. This design provided an unusual opportunity to gain insight into students' independent work.

Data selection and analysis
Coding of the recordings of Sharon and Sylvia's work on their assignment identified seven feedback incidents. Furthermore, three incidents were difficult to accept or reject. Of the seven definite feedback incidents, three were chosen and are analysed below due to their complexity.
Reliability was ensured by discussing data coding with research colleagues. I also had the opportunity to check my own coding, because I accidently coded the same video twice.

Results and analysis
Initially, I present the three feedback incidents. Next, I analyse the incidents with regard to the goals of the requested feedback, the level and content of the feedback and finally the feedback resources and the use of the feedback.

Feedback incidents
Incident one: The meaning of the word 'quality' Sharon and Sylvia are producing a website for an imaginary fish restaurant. They want the restaurant and the mock-up to be elegant and of high quality, as they put it [the researcher would call it an 'upmarket' restaurant]. They frequently discuss their aesthetic choices, referencing the concept of quality. They have planned their mock-up, received feedback from their teacher and discussed and used some of the feedback. The dialogue below takes place during the subsequent concentrated work period: Sharon: What does 'quantity' mean? Is it poor quality? Sylvia: Quantity? Isn't it when there is a whole lot of something?
Sylvia searches the Internet and finds the term 'amount'. With a laugh, Sharon then jokingly notes: The next time we need to use the word 'amount', we'll instead say, 'I have a large quantity'.
Incident two: The creative process Sharon and Sylvia are beginning their productive work. They are about to design an elegant logo for their restaurant. For inspiration, they make a Google image search. They switch between three screens with images of logos. The Batman logo appears at the centre of one of the screens. They point out various logos and exchange views on what they consider 'exclusive'. They do not choose a specific logo as a model, and after a few minutes, Sylvia opens some drawing software, and makes two ovals, one inside the other. Sharon watches, and the following short dialogue takes place.

Sharon:
How did you get that idea? Sylvia: What idea? Sharon: That you could just do that? (Sharon is referring to the two ovals) Sylvia does not answer, and the girls continue to discuss the design of the logo.
Incident three: The difference between websites and advertisements The girls have finished their website mock-up and are going to present the result to their classmates and the teacher. They are planning a video presentation. They have completed a Prezi (software for presentation) of the mock-up, and Sylvia thinks they are ready to supplement the presentation with speech. Confusion arises because they do not distinguish between advertisements and websites. The mock-up is visible on Sylvia's screen. At first, Sharon's screen is dark.
Sylvia: Should we try to shoot it? Sharon: Yes, but first we must plan something to say. Let's find the model for advertising analysis.
Sharon is going to prepare the presentation based on the analysis model. On Sharon's screen, they open a digital resource for advertising analysis used in the previous course. Sharon haphazardly clicks on various links in the model and reads about the sale of goods, and objectivity and subjectivity. While looking at the screen and reading about influence on the reader, she asks Sylvia whether they are cheating.

Sylvia:
Cheating? Sharon is reading: It says here: 'affects the recipient's sense'. Sylvia: And emo… [she is about to say 'emotions'] Sharon: Oooohhh, it is because it is not… shouldn't it be a website advertising analysis? Isn't there a difference? Sylvia: I would say there is a difference, but it is a bit the same.
Sharon is still clicking on the advertising analysis model on her screen. Then she looks at their mock-up on Sylvia's screen and back to her own. She seems confused.
Sylvia: It's just not… I don't know.
The girls continue looking from screen to screen. This silent consideration seems to clarify a difference for Sylvia, because the next question she asks concerns when the teacher switched from the previous topic to the actual topic. The girls find the sheets that confirm that they should be working with web pages and not advertisements.
The goals of requested feedback Initially, I will discuss why the three incidents are requests for feedback. All the incidents involve chunks of communication. In incident one, Sharon asks a question, and she accepts Sylvia's answer without discussion. The talk is cumulative, though Sylvia tends to have an exploratory talk with herself. In incident two, Sharon asks another question, but receives no answer. One explanation is that Sylvia is too focused on designing the logo, another is that she does not know what to answer. A third possibility is that she does not decode the question as a request for feedback. In the Danish context, this sort of question is also a way of praising and expressing appreciation. Yet, I construe Sharon's question as a feedback question regarding her understanding of creative processes based on former statements on her lack of creativity and digital skills. In incident three, the feedback request develops along with the dialogue and is most clearly phrased by Sylvia, who states that she does not understand when they went from advertising to websites. They discuss differences between advertisements and websites through exploratory talk. Because they are seeking information related to their understanding and ongoing learning, the three incidents are external signs of internal work (Nicol & Macfarlane-Dick, 2006). The three requests for feedback are difficult to categorise according to the criteria for the assignment. In incidents one and two, the learning goals are implicit in questions: the meaning of quality and understanding creative processes. In incident three, the immediate goal is to prepare a digital presentation of a mock-up, realising the objectives of the assignment, but the goal evolves throughout the conversation and is transformed into how to understand the type of text they are working on and distinguish it from the texts in the preceding course. The implicit goal changes when Sharon says, 'Oooohhh, it is because it is not… shouldn't it be a website advertising analysis? Isn't there a difference?', and Sylvia answers: 'I would say there is a difference, but it is a bit the same'. In all three incidents, the students' learning goals are implicit but revealed by a question or statement. As 'learning intentions' or goals underlie their questions, the incidents are requests for feedback. Their questions relate to the purpose of and the competence objectives for the subject, Danish (their mother tongue), and thus indicate broader but relevant learning goals. They set the goals for their actual learning through their conversation, and teacher agency is replaced by student agency. Described through Nicol and Macfarlane-Dick (2006), there has been internal student goal-setting in addition to and in continuation of the teacher's goals.
The content and level of requested feedback Three levels of feedback are requested (Hattie & Timperley, 2007) in the feedback incidents. In incident one, Sharon's goal is to understand the meaning of 'quantity', and she asks for feedback at the task level. In incident two, her goal is to obtain understanding of a creative process, and she asks for feedback at the process level. In incident three, Sharon and Sylvia try to understand the relationship between the content in the current course on multimodality and the previous course on advertising. The dialogue shows them monitoring their understanding and asking for feedback at the self-regulation level.
From a broader perspective, the request for feedback is a result of monitoring in all three incidents. In the first incident, Sharon identifies her doubts about her understanding of the meaning of a term. Knowing the meaning of specific words targets disciplinary knowledge in a broader sense than the objectives for the assignment.
Similarly, in incident two, she assesses her (lack of) ability to design a logo before asking how to do. She asks for specific strategies for processing creative work. This question targets processes related to the subject, Danish, and to aesthetic matters in a broader sense. In the third incident, the two girls explore limitations and connections between content areas of the subject. Advertising and websites are established as two different categories, and they now recognise similarities and connections. They struggle to make inferences. The three questions asked are a result of monitoring, which is a central element of self-regulated learning (Nicol & Macfarlane-Dick, 2006), and a request for feedback that supports further learning. The monitoring leading to feedback questions appears in the girls' dialogue as reflections on generalisations, inferences and patterns in the content of the discipline, and may be signs of skilled writers (Graham & Harris, 2000).

Use of feedback and resources
The three incidents differ in how the feedback requests are met. In incident one, Sharon receives an answer to her question and develops her understanding. Though joking, she even explicates her understanding in her concluding comment, 'Next time we say the word "amount", we'll instead say: "I have a large quantity"'. The girls' use of cumulative talk (Mercer, 1996) makes it possible to observe Sharon's learning. In incident two, Sharon again asks Sylvia for feedback. Unfortunately, she does not get feedback, and is not provided with information for further reflection and learning. Incident three is longer, and differs from incident one and two, because it is unclear whether they receive the requested feedback. Initially, they seek feedback from each other, then from the Internet, and finally, from teaching materials and, to some extent, from the teacher. During their dialogue, their goal evolves from that of making a presentation to understanding the differences between advertisements and websites. When they find the instructional material handed out by the teacher, they switch back to the presentation goal, and it is uncertain whether they get information addressing their difficulty in understanding the distinction between advertisements and websites. If the requested feedback is given, it is used in the preparation of the presentation.
Seeking feedback is a result of monitoring one's own learning. Initially, Sylvia and Sharon ask each other for feedback, then they use various kinds of resources, such as the Internet, and finally, they ask the teacher for feedback. The collaborative work on the assignment gives them ownership of the assignment and their learning, and they become instructional resources for each other.
The most important results of this analysis are that student-initiated feedback is the result of self-regulated learning and a starting point for further learning. Giving students agency leads to the emergence of student goals that may differ from the goals set by the teacher. Questioning is an explicit way of initiating feedback, but requests for feedback may take the form of a dialogue. Feedback questions are mainly addressed to the peer. Feedback at all levels is requested, and, if an answer is provided, the information is used.

Discussion
The design of this project allows the students to show rather than tell whether and how they use feedback to support their learning. Video-recorded observation provides unique access to the students' work when they are on their own. A limitation of this method is that only naturally occurring externalised signs are available. In this study, dialogue and gaze are investigated. Internal self-regulatory work, in the form of thoughts, was inaccessible. By interviewing the students and asking for their thoughts, some thoughts could have been externalised. In other words, the number of feedback incidents during the work on the assignment may be higher.
As students monitor their learning, they set up internal goals. If a request for feedback is expressed as a question, the learning goal appears in the question itself. If the request develops along with dialogue, the goal may be implicit. A student's internal goal may extend beyond the goals and criteria set by the teacher. It seems necessary to conceptualise the goals set by the teacher as providing a direction for the actual work and learning, and the purpose and competence objectives for the entire discipline as the horizon. In practice, the disciplinary horizon is narrower, as the feedback request is situated. In the case of Sylvia and Sharon, they ask questions mainly within the framework of the current and previous courses. An implication of this understanding is that it is necessary to be familiar with the content of an entire discipline to identify questions as feedback requests. In other words, the academic content of a discipline is central (Bennett, 2011, p. 20).
Sylvia and Sharon seek feedback at all levels and related to the three feedback questions. Incident three encompasses feedback requests related to all the questions. They discuss what they are expected to present -where they are going. They judge their mock-up to be finished and ready to share with their classmates and are aware that the speech of the presentation is still missing -how they are doing. Finally, they discuss how to finish their assignment. Furthermore, the three incidents are requests for feedback at the three feedback levels. The three levels form a progression (Hattie & Timperley, 2007) that makes the self-regulation level the most complex. Monitoring one's own learning before asking for feedback is a complex process. Asking for feedback at the task level requires assessment of the current assignment, and the process level requires assessment of the actual work process, but feedback at the self-regulation level requires meta-reflection on the assignment, as well as on one's work on the assignment. As self-regulation develops with age and schooling (Graham & Harris, 2000), scaffolding the development of self-regulated learning by scaffolding self-assessment and feedback may encourage assessment of one's performance (Gielen, Tops et al., 2010). Sylvia and Sharon seek feedback mostly from each other and the learning materials and, to a lesser extent, from the teacher. Based on this study, it is impossible to conclude whether this pattern is the result of setting a collaborative task, a result of their trust in each other because of their high level of performance or a result of something else.

Conclusion
The two students in this study, whose reading skills are demonstrably strong, seek feedback independently during collaborative work. Their requests for feedback are a result of self-regulative work and an activity that supports further learning. The feedback requests appear as explicit questions or as statements evolving from dialogue characterised as cumulative or exploratory talk. The exploratory talk provides the opportunity to reveal how the recipient of the feedback will proceed with his or her learning. Mostly, they ask each other for feedback, and sometimes they get the feedback requested, sometimes not. Getting no answer from the peer does not necessarily result in asking the question to the teacher.
A student's request for feedback is evidence of ongoing learning and personal goal setting. These goals may transcend the goals set by the teacher. The teacher's goals and criteria define the direction of the actual learning, whereas the purpose and competence objectives of a discipline constitute the horizon of the questions. Feedback requests are made at all three levels described above. At the most complex level of feedback, the self-regulation level, the subject matter is reflected in expressions of wondering, generalising, and pattern-identification.
This small case study is only a first step in the investigation of student feedback-seeking. There is a particular call for intervention studies in which scaffolding is provided for learners' self-initiated feedback seeking. As indicated by the present study, this kind of feedback processes bear witness of ongoing learning and pave the way for the development of self-regulated learners.