Modeling preservice middle school science teachers’ reflective practice

Abstract We present an empirically based model for modeling the quality of pre-service teacher reflection. Conversations from twelve groups of a total of 47 pre-service teachers were video recorded and transcribed verbatim. First, we analyzed their conversations through practical epistemology analysis and an operationalization of Dewey’s definition of reflection to identify the moments of reflection that appeared in the pre-service teachers’ talk. Thereafter, we employed Dewey’s reflective attitudes, responsibility, open-mindedness and whole-heartedness, with the aim of understanding their roles in moments of reflection and how they relate to the quality of reflection. Our results show that the reflective attitudes played different roles and based on these roles and on particular patterns of the attitudes in our data it was possible to use them for modeling the quality of the pre-service teachers’ reflection. We suggest that presence of all of the attitudes corresponds to higher quality than absence of any of them. Moreover, identifying absence of one or more attitudes in PSTs’ reflection makes it possible to explicitly talk about what aspects of the reflection that should be improved, and why. Our results contribute with a first tentative model which may support teacher educators work with the development of pre-service teachers’ reflective practice.


Introduction
Reflection is a central component in teacher education programs around the world. Yet, we still lack clear conceptualizations of what reflective practices may look like, as well as conceptual tools for supporting PSTs to become good reflective practitioners.
In life, we constantly encounter new situations in which we need to find new solutions in the face of perplexity and feelings of shortcoming. This " [d]emand for the solution is the steadying and guiding factor in the entire process of reflection" (Dewey, 1910(Dewey, /1997(Dewey, /1997. Reflection may be regarded as "an important human activity in which people recapture their experience, think about it, mull it over and evaluate it" (Boud et al., 1985, p. 19). Reflection is part of peoples' daily life and of crucial importance for professional judgment (Coles, 2002;Dottin, 2009). Schön (1983) notes that "[w] hen a practitioner reflects in and on his practice, the possible objects of his reflection are as varied as the kinds of phenomena before him and the systems of knowing-in-practice which he brings to them. He may reflect on the tacit norms and appreciations which underlie a judgment" (p. 62). However, depending on which professional practices the practitioner belongs to, there are different professional judgements to be made. For instance, a lawyer has to make professional judgement concerning how to interpret laws and contracts and also about solving legal problems whilst a physician has to make judgement about diagnosis and treatment of patients.
In turn, teachers have to make professional judgements concerning many different things, e.g., what to teach and when and how to teach and assess it (Jank & Meyer, 2006). In this process of making professional judgements "reflection is considered as a key element of a teacher's professional judgement" (Shakra, 2022, p. 114). This need for reflection becomes even more challenging considering that the basis for teachers' professional judgement is constantly changing as schools are in a virtually constant state of reform (Carlgren, 1999;Collin et al., 2013;Grant & Zeichner, 1984). Consequently, the development of pre-service teachers (PSTs) into reflective practitioners has been a significant component in teacher education programs over the last 30 years (Beauchamp, 2015;Ruffinelli et al., 2020). As Tabachnik and Zeichner (1991) note, "there is not a single teacher educator who would say that he or she is not concerned about preparing teachers who are reflective" (p. 1). Conway (2001) goes as far as claiming that reflection has "become a guiding beacon" (p. 90) in teacher education.
Previous research demonstrates different ways to develop PSTs' reflective practice. Some studies indicate that reflection seems to benefit from being made jointly both with other peers who act as critical friends, and with teacher educators who may support PSTs in their reflective processes by creating a safe environment in which PSTs can identify opportunities and difficulties and develop suggestions on how to improve their teaching (Cavanagh & McMaster, 2015;Parsons et.al., 2005). This joint reflective work may be implemented in different ways. Rhine and Bryant (2007) provide an example of PSTs filming their teaching. Afterwards they watched the recorded teaching together and examined the consequences of what the PST and the students did, which gave the PST an opportunity to retest his or her reflection-in-action. Mulryan-Kyne (2021) suggested that the development of PSTs' reflective practice may be supported by integrating possibilities for reflection recurrently in all courses given in teacher education together with a constant support from peers and teacher educators. In addition, Roberts et al. (2021) showed how the development of PSTs' reflective practice may be supported by teacher educators presenting PSTs with reflective frameworks and also showing example reflections to elaborate what teacher reflective practice may look like.
Other studies instead indicate that PSTs seem to lack the ability to reflect systematically (Emsheimer & Göhl, 2014) and that many PSTs perceive the use of models for supporting reflection as something imposed and extraneous (Krapivnyk et al., 2021). When working with reflection, little consideration is given to PSTs questions and what they experience as uncertain, and at the same time a clear connection to the context in which PSTs' reflection is intended to take place is often missing. As a consequence, PSTs perceive reflection as a linear and mechanical process without any obvious relevance to teaching practice (Boud & Walker, 1998;Zeichner, 1993).
Despite the central place for reflection in teacher education, then, research seems to be divided as to the prospects of developing PSTs' reflective practice. The diversity of definitions does not offer a coherent and functional support for teacher educators in their quest for developing PSTs' reflective practice, leading to difficulties for PSTs to understand what it implies (Beauchamp, 2015;Bengtsson, 2007;Finlay, 2008). As Mann et al. (2009) state: "there is surprisingly little to guide educators in their work to understand and develop reflective ability in their learners" (p. 596). In this study, we address this knowledge gap by investigating preservice middle school science teachers' reflective practice as they plan a microteaching exercise together, in order to model the quality of their reflection. described as an individual process (Zeichner & Liston, 1996) it is sometimes also considered as something that may be carried out collectively (Collin & Karsenti, 2011). Rodgers (2002), with reference to Dewey, argues for the benefits of collaborative reflection, as participants' experiences get confirmed to a greater extent than when reflection is made individually, thereby broadening the field of understanding and supporting engagement in the process of inquiry (see also Cavanagh & McMaster, 2015;Parsons et.al., 2005). In an earlier study (Karlström & Hamza, 2018) we created a methodological tool for identifying such joint, collaborative reflection in an activity, viz "situations characterized by an experienced difficulty and a subsequent effort to solve the problem" (p. 5). This operationalization comes from Dewey's five steps for characterizing reflection (1910/1997, p. mw.6.236) "(i) a felt difficulty, (ii) its location and definition, (iii) suggestion of a possible solution, (iv) development by reasoning of the bearing of the solution, (v) further observation and experiment leading to its acceptance or rejection", In the study, we demonstrated that reflection was all pervasive in Swedish PSTs' conversations while jointly planning a microteaching exercise concerning sustainable development (Karlström & Hamza, 2018). Planning was definitely not a walk in the park for the PSTs, who frequently ended up in situations where they displayed "a state of perplexity, hesitation, doubt' (Dewey, 1910(Dewey, /1997(Dewey, /1997. These situations became starting points for processes of inquiry, constituting what we labeled "moments of reflection" that occured "in the midst of a performance" (Schön, 1983, p. 54).
Although these moments of reflection contained Dewey's five distinct steps for characterizing reflection (1910/1997, p. mw.6.236) it was noticeable that the different inquiries did not look the same. Intuitively and without being able to say precisely in what way, we came to perceive the PSTs' reflection as more or less developed. Others would perhaps say that they were of different "quality" (Roberts et al., 2021), "breadth and depth" (Lane et al., 2014) or "levels" (Mulryan-Kyne, 2021; Rhine & Bryant, 2007). These various ways of describing differences in reflection provides an understanding of that PSTs' reflections may be regarded as more or less complete. Dewey's model of reflection helped us identify and describe the PSTs' reflective processes and what they might learn from their moments of reflection (Karlström & Hamza, 2018) however the model did not provide any support for judging their reflections with a greater discernment nor creating an awareness of what it means to reflection with high quality. Leboskey (1993b) talks about reflection as being "carried out with the three attitudes of openmindedness, responsibility and wholeheartedness" (p.10) and that these attitudes are "integral to reflective action" (p.30). Thus, instead of talking about reflection as a means for solving problems we may direct our attention to how openmindedness, responsibility and wholeheartedness affect PSTs' actions when they end up in problematic situations.
In fact, these three attitudes, which Dewey introduced already in 1916 (Dewey, 1916), have been used frequently in previous research on reflection in teaching and teacher education. For instance, they played a crucial role when Loughran (1998) analyzed one teacher's reflective work through the teacher's diary, written feedback and responses from students and colleagues and students' interview transcripts and were used as a lens when So (1997) analyzed a language teacher's teaching with an aim to portray her reflective teaching as well as when Farell (2014) analyzed his own and other language teachers' teaching experiences and reflections on these. Dewey's three attitudes also become a basis when Grant and Zeichner (1984) discussed pre-and in-service teachers' classroom teaching, when Zoellner et al. (2017) analyzed PSTs' written professional development plans, and when Harrington et al. (1996) analyzed PSTs' written dilemma-based cases. Dewey (1933) means that, in contrast to the development of muscles as a result of repeated gymnastic exercises, there are no "set exercises" (p. lw.8.135) that will develop a good reflective thinker. Rather, a person needs to "cultivate those attitudes that are favorable to the use of the best methods of inquiry and testing" (p. lw.8.136). Accordingly, Dewey (1933) describes the three attitudes as follows: Responsibility "[T]o consider the consequences of a projected step" meaning a willingness "to adopt these consequences when they follow reasonably from any position already taken" (lw.8.138). Beingresponsible is the presumption "to carry something through to completion" (lw.8.138).
Open-mindedness "[F]reedom from prejudice" (p.lw.8.136) and when embracing the attitude of open-mindedness, it becomes easier to consider new problems with "hospitality to new themes, facts, ideas, questions" (p.lw.8.136) along with "an active desire to listen to more sides than one" (p.lw.8.136).
Although Dewey (1933) stresses the need to cultivate these three attitudes in the quest to become a good reflective thinker, the attitudes can also be used to conceptualize a collaborative reflective practice (Collin & Karsenti, 2011;Rodgers, 2002). This may give us a possibility to talk about reflective practice with a higher resolution and discernment, thereby making it easier to handle different aspects of what may described as completeness or quality of reflection. Thus, rather than treating responsibility, open-mindedness and whole-heartedness as individual characteristics we try to regard the attitudes as functions in ongoing moments of joint reflection which in the end, hopefully, may help both teacher educators and PSTs to better distinguish and talk about different parts of reflection, in order to develop the quality of reflection more easily.

Setting
During their third semester in the teacher education program at the Swedish university in question located in a larger city, PSTs can choose one of four elective subject and subject didactics courses. The PSTs in this study had chosen the course "Teaching Science and Technology in grades 4-6" (30 ECTS 1 credits) in which they both studied science and technology and how to teach these subjects. The course contains three different modules. The last one, focusing on science education for citizenship and socioscientific issues, includes a microteaching exercise (Bell, 2007). This exercise consists of an initial phase in which the PSTs plan a 20-minutes microteaching lesson collaboratively in small groups, a teaching phase where the PSTs implement their plan to their peers who are acting as students, and a reflection phase in which the PSTs and teacher educators evaluate and discuss the lesson. Before commencing planning, the PSTs were provided written guidelines. According to the guidelines that the PSTs received from the teacher educators, they had to provide a written plan including: • Information about the targeted students (e.g., age, previous teaching etc.), • A short description of purposes and learning goals for the microteaching lesson linked to the national curriculum goals, • A short account of selection of content and assessment methods, • A short description of the planned teaching activities and their purposes, and finally • Three questions to discuss with their peers after the microteaching, concerning pedagogical • Choices and dilemmas during planning and enactment of the 20-minutes lesson.
Before entering this course (i.e., during the first and second semester) the PSTs had taken four courses in general education (The history of education and its role in society, Perspectives on learning and development, Law and Ethics in school, and Social relations in school, amounting to 15 ECTS credits in total), three courses in subject and subject didactics: Mathematics I and II (7,5 ECTS credits, respectively) and Swedish [Home language] and language education (15 ECTS credits), and School Placements I and II (2,5 and 5 ECTS credits, respectively). In total the Swedish teacher education program for middle-school includes eight semesters corresponding to 240 European ECTS credits distributed as follows: subject and subject didactics (120 ECTS), general education (60 ECTS), school placement (30 ECTS), and diploma work (30 ECTS).

Study design
Our purpose was to explore the role of Dewey's three reflective attitudes in the PSTs' planning conversations, and based on these analyses suggest possible ways of modeling the quality of the PSTs' reflective practices. Therefore, we employed the didactic modeling framework as our overarching research design (Wickman et al., 2020). Didactic modeling aims at the production of models that support teachers' professional decision-making concerning purposes (the didactical question why), teaching methods (how) and/or content (what) etc. (Wickman et al., 2020). Didactic models are typically conceptual (cf. Thalheim, 2010), consisting of typologies or other categorical schemes, which help a teacher to better distinguish certain features of a teaching situation (Arnold, 2012;Jank & Meyer, 2006;Westbury et al., 2000;Wickman et al., 2020). Examples of didactic models are Klafki's (1995) critical-constructive didaktik, Duit et al. (2012) model of educational reconstruction, Schwab's (1962) Degrees of freedom, and Karlström and Hamza's (2021) planning model for teacher education.
Didactic modeling ideally proceeds through three phases: extraction, mangling and exemplifying (Wickman et al., 2020). In this study we are focusing exclusively on the first one-the extraction phase. This is the most creative part of the didactic modeling process, in which the researcher draws on both theoretical considerations and empirical data in order to extract components of a tentative model, a process very similar to Grounded Theory described by Glaser and Straus (1999) in which "one generates conceptual categories or their properties from evidence; then the evidence from which the category emerged is used to illustrate the concept" (p. 23).
For instance, we cannot rule out the possibility that teacher educators try to apply Dewey's idea concerning the cultivation of the reflective attitudes as a didactic model for creating opportunities for PSTs to develop their reflective practice (cf. Grant & Zeichner, 1984;Zoellner et al., 2017). However, this purely theoretical model provides teacher educators limited support because of the many unknowns, for instance the roles of the three attitudes, how they interact, if there are variations in how they are expressed, or if there are specific patterns of the attitudes creating different kinds of reflection. Wickman et al. (2020, p. 42) point out that "[t]he transaction between theory and practice in the extraction process may, apart from supporting the modeling procedure, also bring insight into what is missing in existing teaching traditions, and so supplement the model". The extraction phase results in a form of prototype model, in our case a model for developing reflective practice, that is both theoretically (Dewey's conceptual scheme of three attitudes) and empirically (our analyses of PSTs' reflection) based.
To find an answer to RQ1 we first needed to identify all moments of reflection that emerged. To be able to do this we applied a methodology developed by Karlström and Hamza (2018). Thereafter we operationalized Dewey's three attitudes (responsibility, open-mindedness and wholeheartedness) so that we could identify them in the conversations (i.e., during the moments of reflection that we identified). Finally, we tried to clarify the roles of responsibility, open-mindedness and whole-heartedness in the PSTs' reflections.
To find an answer to RQ2 we then searched for patterns in how responsibility, open-mindedness and whole-heartedness emerged in the conversations, and tried to relate these patterns to an assessment of the quality of the PSTs' reflection.

Data collection and processing
Data was collected through a naturalistic observation-based and non-reactive qualitative design (Bryman, 2018;Arthur et al., 2012). The first author presented the research project to the PSTs and received written permissions from each of them to make video and audio recordings of their planning conversations during the upcoming microteaching exercise. As the course was given on a regular basis but with a limited number of PSTs attending, we chose to gather data twice, initially from 22 PSTs and one year later from an additional 25 PSTs, i.e., a total of 47 PSTs (14 men and 33 women respectively). The additional sampling was made possible as the course was carried out in exactly the same way on both occasions.
During the microteaching exercise, the PSTs were divided into smaller groups, each group comprising 3-6 PSTs, which resulted in a total of 12 groups from the two rounds of data collection. The PSTs in each group had to collaboratively plan and implement a 20-minutes microteaching lesson on sustainable development. Planning conversations took place in different rooms on campus and we set up one fixed video camera that showed the entire group and put a voicerecorder on the table. The duration of the planning sessions varied between 60 and 120 minutes and some of the groups also chose to meet more than once to conduct their planning (in those cases, we only recorded the initial meeting in the group).
The first author watched and listened through the collected data thoroughly (Bryman, 2018) during the process of transcribing all data verbatim. The analysis was then carried out in collaboration between the first and second author in the following way. All transcripts were initially read repeatedly by the first author (Moser & Korstjens, 2018) and analyzed by adopting the reflective attitudes as described below in the analysis sections. This reading and analysis, in turn, was examined by the second author whereafter the two authors together followed up the analysis and arrived at joint decisions (Arthur et al., 2012).

Analyzing the role of the three attitudes in the conversations
In order to answer RQ1 we first needed to comprehend what the PSTs were doing during their planning conversations and as a second step, identify all emerging moments of reflection. For those two purposes, we applied a methodology developed by Karlström and Hamza (2018), which is a combination of Practical Epistemology Analysis (PEA; Kelly et al., 2012;Wickman & Östman, 2002) and Dewey's notions of reflection as a form of inquiry (1910/1997). In an initial step, PEA was used for identifying all the gaps that the PSTs notice. Operationally, a gap may occur as an explicit question or a marked break in the conversation. In order to proceed, PSTs need to fill the gap with relations, otherwise the gap will linger and the conversation takes a different direction. Alternatively, a gap may be implicit, in which case it is inferred analytically directly from the relations established in the conversation (Wickman & Östman, 2002). In the second step, those gaps leading to moments of reflection were distinguished. Dewey describes reflection as a process which starts in "a state of perplexity, hesitation, doubt" (Dewey, 1910(Dewey, /1997 whereafter an inquiry takes place with the purpose of "converting the dubious into the assured, and the incomplete into the determinate" (Dewey, 1925, p. lw.1.61). Based on these notions, it was possible to identify gaps constituting doubtful situations, which were followed by PSTs identifying the problem and finally dealing with it by implementing an inquiry-which is a working operationalization of reflection (Karlström & Hamza, 2018). The methodological use of this operationalization made it possible to follow PSTs' collaborative reflective practice. At the same time, it obviously did not make it possible for us to observe what was going on in the PSTs' minds. Thus, we do not claim to have detected all reflection that took place during PSTs' planning, but rather the joint and overt reflection occurring as part of the PSTs' conversations (see also Karlström & Hamza, 2018).
Having identified moments of reflection in the PSTs' conversations, we needed to analyze them in terms of Dewey's three attitudes. Therefore, we operationalized the three attitudes based on Dewey's (1933) notions as well as how they have been operationalized in later research (Farell, 2014;Grant & Zeichner, 1984;Rogers, 2002;Zoellner et al., 2017), as shown in Table 1.
All three attitudes did not appear in all inquiries. Therefore, we also recorded their absence in the PSTs' moments of reflection, labeling these parts of the conversations non-open-mindedness, nonwhole-heartedness and non-responsibility, respectively.
Having identified both responsibility, open-mindedness and whole-heartedness and the absence of these (i.e., non-responsibility, non-open-mindedness and non-whole-heartedness) in the gaps we finally tried to analyze and clarify their roles and how they were interconnected in PSTs' moments of reflection.

Modeling the quality of reflection
To find an answer to RQ2 we searched for and described different patterns in how responsibility, openmindedness and whole-heartedness (or their respective "non"-counterparts) emerged in the conversations. Thence, we tried to relate these patterns partly to a subjectively perceived quality of the PSTs' reflection, partly to the roles that we identified for the three attitudes through RQ1. Through this combination of different kinds of analysis (systematic and more subjective), we ended up with a tentative suggestion for how the quality of reflection may be modeled in pre-service teacher education.

Analyzing the role of responsibility, open-mindedness and whole-heartedness
Our results show that the attitudes played different roles when the PSTs ended up in problematic situations. Responsibility opened up for an inquiry to take place. In that way responsibility played a crucial role for whether there was going to be an inquiry at all. In that sense, responsibility can be compared to the starter motor in a car-if not working, the engine will not start either and as a result there will be no locomotion, even though the engine itself works perfectly and there is plenty of fuel in the tank. Responsibility also played a role when it came to the completion of the inquiry, as it made loose ends being tied together and invited the search for a final conclusion to the problem. Thus, responsibility also turns off the engine, as it were. Open-mindedness played a crucial role as it came to whether the conversation was going to be disputational or exploratory (cf. Mercer & Wegerif, 2004). Thus, open-mindedness may be said to allow material into the inquiry that can work as fuel. Whole-heartedness, finally, played a role of deepening the inquiry once there was fuel to get it going. In that sense, whole-heartedness might be likened to the clutch making the inquiry move forward once the engine has started (responsibility) and provided it gets appropriate, and enough, fuel (open-mindedness).

Reflective attitudes
Is seen when PSTs: Responsibility thoroughly investigate their beliefs, ideas and suggestions, whether or not leading into action, with an aim to understand the consequences of their actions.
Open-mindedness reexamine existing beliefs and practices without being condemning, and entertain and investigate multiple perspectives.
Whole-heartedness actively seek to overcome encountered uncertainties and show curiosity in investigating how ideas, suggestions and decisions may improve their teaching practice.
In the following two examples of PSTs' moments of reflection during planning we try make these different roles of responsibility, open-mindedness and whole-heartedness visible. This is done by analyzing their role both when they emerged in the conversation and when they were absent.
Example 1: When we enter the excerpt, the PSTs have decided that their students should be preparing themselves for a classroom discussion by searching for different arguments concerning the choice of different energy sources. At the end of the activity, the students should also account for any of the arguments that they have found and also what sources they have used during the preparation. When the PSTs summarize this, one of them, Ann, questions if source criticism really is one of the aims for the lesson (Turn 1).
The conversation starts with Cia summarizing what they have come up with so far, namely that the students should present their arguments and the sources for these. However, Ann points out that source criticism was not part of the aim for the lesson and, thus, implies that there is not good 1. Cia: [Reads aloud what they have decided earlier]"These four arguments and the sources are to be accounted for in class later on".

2.
Ann: But I'm not sure we had. . .eh. . .source criticism as an aim for the lesson.

Cia:
No. For that reason we will say that the sources have to be accounted for . . . eh . . .

Anton:
Or saved until the next lesson?

Cia:
No, but I'm thinking that. . .now they're only supposed to tell from where they got their arguments. We haven't put any evaluation in [interrupted].

Cia:
Because this is what we will bring forward the next lesson.

10.
Cia: Which will focus on source criticism. "Yesterday, when you got to argue pro and con, I wrote the names of the sources that you used on the white board".

12.
Cia: "What do you think when comparing these two sources?" 13. Kent: "And who really, is right?"

17.
Anton: This is a good way to make a connection between the lessons. alignment between aims and content of the lesson (turn 2). Taking this step back and questioning the initial assumptions of their plan, just as they are about to summarize it, includes central parts of the reflective attitude responsibility, namely to investigate suggestions and understand the consequences of certain actions (Table 1). As a result of this intervention, the conversation stops and an inquiry commences. Thus, they begin to reflect in the Deweyan sense. Then Anton (turn 4) asks if this should instead become part of "the next lesson" whereby Cia (turn 5 and 7) turns this into an idea of how to solve this problem. This is an instance of whole-heartedness, as the PSTs are actively seeking to overcome encountered uncertainties and investigating how ideas, suggestions and decisions may improve their teaching practice (Table 1). This new idea is treated positively by both Anton and Ann (turns 8-9). Not being condemning and acknowledging a new perspective on their lesson, their response aligns with the reflective attitude open-mindedness (Table 1). Cia will not let go of the idea of what continuous teaching may look like and pictures the "next lesson" whereupon the whole group shows curiosity and investigates how these ideas may improve their practice (turn 10-20) which, again, goes along with the reflective attitude whole-heartedness (Table 1), Finally, in turn 21, Cia summarizes what they have come up with concerning both the microteaching lesson and also an imagined following lesson, thereby making it possible to understand the consequences of their actions which also includes a central part of the reflective attitude responsibility (Table 1).
In this example, responsibility (R) played two different roles, first in the initiation of the inquiry (turn 2) and later on, also in its closure (turn 21). Once the inquiry was initiated, open-mindedness (O) brought ideas and experiences into the inquiry (turns 8-9) and whole-heartedness (W) promoted an exploration of these emerging ideas and experiences (turns 10-20).
Example 2: When we enter the excerpt, the PSTs have just been discussing teaching methods and decided that their students should work together in small groups. The situation becomes problematic as the PSTs start to discuss on what grounds they should divide their students into groups (Turn 22).

22.
Cia: We should put both stronger and weaker students together in a classroom, shouldn't we? Cia starts by asking her peers a question (turn 22) that signals a willingness to thoroughly investigate their beliefs, ideas and suggestions, which is a signum for the reflective attitude responsibility (Table 1). Thereafter, however, the conversation becomes disputational as the PSTs express different opinions on how to compose groups while making no attempts to pool their different opinions (turn 23-35). Here, then, we can observe rather the opposite of "reexamine existing beliefs and practices without being condemning, and entertain and investigate multiple perspectives", that is, we have an instance of non-open-mindedness. As the inquiry continues, the PSTs give examples of things that may be of importance when dividing students into groups (turn 36-41). But they present the different ideas more as a kind of inventory, without showing any curiosity in investigating them further. Thus, here we have a lack of the characteristics of the reflective attitude whole-heartedness, making this instead an instance of non-whole-heartedness. The inquiry was initialized by the question about how the PSTs should divide their imagined students into smaller groups and as we have seen, the ensuing inquiry is primarily a list of different personal opinions and ideas, with no attempts to tie loose ends together and decide on what grounds to divide students. On the contrary, the PSTs end up suggesting that "the teacher" should decide (turn 44-46). The encountered uncertainty embedded in the question about assigning students into groups still remains. Thus, here the PSTs do not follow through to consider any consequences of their ideas and suggestions, amounting to an instance of non-responsibility.
In this second example, responsibility (R) only comes to play the role of initiating the inquiry (turn 22). The appearance of non-responsibility (non-R) later on in the inquiry means that the PSTs omit the search for consequences of their actions and a solution to the problem. Thereto, non-(Continued)

34.
Ann: That the gap doesn't become too great.

35.
My: That's, I mean, I've been thinking about this so often that, maybe when I, during school placement, saw a strong student in Swedish [home language] that is put together with a student that is on the verge to pass the Swedish course . . . then the gap was too great. It feels like the strong student has to explain every word that is said and it becomes . . . there will be no progress.

Ann:
But. Isn't that more depending of the classroom environment than how students are put together in groups? open-mindedness (non-O) limits PSTs possibilities to seriously consider other perspectives (turns 23-35) at the same time as non-whole-heartedness (non-W) paralyzes momentum and implies that ideas that are aired are not being followed-up and further investigated (turns 26-41).
With these two examples we have tried to visualize when and how Dewey's reflective attitudes come into sight during PSTs moments of reflection. In summary, in our material as a whole, the roles of the reflective attitudes can be described as follows: • Responsibility is central for initiating the process of inquiry, as something that is experienced as unclear or inconsistent is brought up for discussion.
• Responsibility also makes room for closure, for instance as PSTs tie up loose ends, evaluate consequences and determine what to do.
• Non-responsibility at the beginning of a conversation means that experienced uncertainties stay invisible, which in turn implies that reflection is never really initiated.
• Thereto, the appearance of non-responsibility at the end of a reflective moment means that the consequences of the thoughts and actions dealt with during the reflection are not considered.
Given the presence of responsibility at the beginning, then: • Open-mindedness provides opportunities for an exploration of different beliefs and experiences.
• In turn, non-open-mindedness is connected to conversations instead turning disputational and lacking attempts to pool suggestions and ideas.
• Whole-heartedness in linked to opportunities to take care of and elaborate ideas in more depth.
• Non-whole-heartedness, finally, makes inquiry more of an inventory as when for example, PSTs simply list possibilities without investigating them any further.

Modeling the quality of reflection through Dewey's three reflective attitudes
As shown above, the reflective attitudes played different roles in the PSTs' reflection, each contributing a certain quality that we ventured to express in analogy with a car: getting the engine started and turned off (responsibility), fueling the engine (open-mindedness) and releasing the clutch so that the car moves forward (whole-heartedness). In that sense, presence of the three attitudes can be considered as an important feature of good reflective practice. Example 1 encompassed all three attitudes and may be considered the most developed/complete reflection found in the material. Correspondingly, the absence of an attitude meant that although reflection was initiated, it became thwarted as these roles were not played out properly. Example 2 lacked all three attitudes (except for Responsibility at the beginning, see further below) and may be considered the least developed/complete reflection.
In the most developed reflection, responsibility (R) emerged both initially and in the end of the inquiry, while open-mindedness (O) and whole-heartedness (W) took place in between (Table 1, pattern no. 1). In less-developed reflections, one or more of Dewey's attitudes were absent (Table 1, patterns no. 2-5). However, as seen from Table 1, responsibility always appeared at the beginning of a reflective moment. Obviously, if non-responsibility would appear initially, there would probably never become a reflection at all as there would be no hesitations or problematic situations that the PSTs needed to handle (i.e., there would be no moment of reflection). However, during an ongoing inquiry, then even if we find both open-mindedness and whole-heartedness, non-responsibility could appear by the end (Table 2, pattern no. 2), meaning that the PSTs never arrived at a conclusion or solved the problem.
The last three patterns all lacked the attitude open-mindedness (Table 2, patterns 3-5), thereby showing less complete reflection compared to the first two, in which all three attitudes were present. As we showed above, non-open-mindedness during reflection meant that ideas and suggestions were not received and dealt with in the group, leading to the conversation turning disputational as no one showed an interest in what the others were suggesting. Patterns no. 3 and 4 displayed an alternation between whole-heartedness and non-whole-heartedness. As shown above, non-whole-heartedness meant that to the extent that suggestions or ideas were received, they stayed unexamined or were "dealt with" by guesswork. Pattern no. 3 contained responsibility at the end, whereas pattern no. 4 lacked this kind of closure ( Table 2). As already stated, finally, pattern no. 5 virtually consisted in the initiation of reflection followed by a very thwarted inquiry lacking all three attitudes (Table 1).
These patterns of Dewey's reflective attitudes constitute a structured way of modeling the quality of PSTs' reflection, in conjunction with the roles that the attitudes were shown to play in the PSTs' conversations. It is obvious that all three attitudes contribute to the quality of reflection meaning that the three attitudes are important features of any good reflective practice (as stipulated theoretically already before our analyses) and, moreover, that they create conditions for each other (cf. our car analogy).
Thus, a tentative first model based on our empirical data would be that presence of all of the attitudes (pattern no. 1, Table 2) corresponds to a complete reflection containing all necessary elements: from initiation by taking on a perceived problem, over reception and recognition of alternatives and ideas and working on them, to closure through consideration of consequences. Absence of an attitude, on the other hand, indicates a need for improvement/development of the reflective practice. Moreover, the kind of improvement that is needed may be discerned and made explicit, as it corresponds to a certain attitude that needs to be included in the reflection to make it complete (sequence of patterns no. 2-5, Table 2). Thus, quality of PSTs' reflective practice may be modeled in two consecutive and interdependent "steps", as it were: first whether the reflection is complete (including all three attitudes) or not, and second which attitude(s) is/are needed to improve the quality of the reflection (discernment), and why (their respective functions).

Discussion
Our results suggest that it may be possible to distinguish qualities in PSTs' reflection using Dewey's (1933) reflective attitudes, that is, responsibility, open-mindedness and whole-heartedness. This is because the three attitudes seem to play different roles in the process of reflection while at the same time, they create conditions for each other. Thus, responsibility opens up for inquiry. In this way, responsibility seems to have a crucial role for whether reflection is implemented at all. Responsibility may also be of significance at the end of the reflection-although not as crucial as in the beginning, without responsibility there will be no real closure. Open-mindedness seems to be of importance for whether the reflection becomes constructive and exploratory rather than competitive and disputational (cf. Mercer & Wegerif, 2004). Whole-heartedness seems to matter

Table 2. The different patterns of presence and absence of the reflective attitudes Responsibility (R/non-R), Open-mindedness (O/non-O) and Whole-heartedness (W/non-W) identified in PSTs' conversations, and their suggested correspondence to the quality of reflection
No.
Pattern of reflective attitudes The presence of all three attitudes in PSTs' reflective practices seemed to be decisive. In the most developed or complete reflection found, responsibility emerged both initially and at the end of the inquiry, while open-mindedness and whole-heartedness took place in between, whilst in less-developed or complete reflections, one or more of Dewey's attitudes were absent. This is in line with Leboskey's (1993a) claim, that Dewey's three reflective attitudes "are more critical to the reflective process than the specific steps" (p. 26) as all the PSTs' reflections contained Dewey's five distinct steps for characterizing reflection (1910/1997) but showed differences in quality. In that sense, the three attitudes are important features of any good reflective practice.

Quality of reflection Excerpt
Our model is tentative and it is not our intention to ignore or replace other, more holistic, or more subjective ways to describe quality in PSTs' reflective practice. On the contrary, our model should be seen as offering teacher educators (and by extension, PSTs) a complementary, structured and systematic way to talk about qualitatively different reflections with greater discrimination and resolution. Reflection is made in relation to a specific content (Loughran, 1996) with the purpose to increase understanding about this content, and/or to improve this content (cf. Beauchamp, 2015). The quality of the outcome of reflection may be expressed in relation to how much understanding increased or to what extent content got improved. As a result of reflection, how much and to what extent may differ and these differences give us ways to talk about quality, both implicitly (i.e., referring to "gut feeling") or more explicitly by the use of a conceptual model like the one we have presented, through which it becomes possible to distinguish different components of reflective practice, judge to what extent these components are present or not, and explicitly point at missing components and what they would contribute the reflection. In that way, our model offers a possibility to relate the quality of the result or the product to what can be found (or not found) in the reflective process. This comparison may give us a possibility to understand both why quality differs and also how to improve and make PSTs' reflections more complete.
For instance, when reading through the second example one may wonder what if the PSTs had asked each other what they each meant by "stronger and weaker students" (turn 22), or in relation to turn 28, what research concerning grouping students that they had come across earlier in the program? Considering these and related questions through the lens of the model, we see that the first one aligns well with the reflective attitude open-mindedness, as it concerns an interest in and readiness to consider other opinions and ideas. In the same way, the second question is connected to the reflective attitude whole-heartedness, as it implies an interest in discovering what research actually has to say, rather than simply replying "mm" to the question (turn 29). Would the PSTs have reexamined their beliefs or investigated how their ideas may improve their teaching practice if these shortcomings of their reflection had been noticed? A teacher educator listening in on the PSTs' discussions through the lens of the model, would have an instrument for discerning elements of quality in the reflection that might not otherwise have been explicit to either the teacher educator or the students. In other words, instead of asking these and similar questions "at random", the teacher educator would be able to search for instances of these attitudes, as well as to distinguish less developed reflections (lacking one or more of the attitudes) from more developed ones. Moreover, the model would offer the teacher educator a way to visualize and make explicit opportunities for development in interaction with the students where possibilities for "perspective transformation" (Mezirow, 1981, p. 7) may happen.
Dewey talks about cultivating the reflective attitudes (1933, p. lw.8.136). To cultivate may be interpreted as to give something a possibility to grow and our result indicate that there seems to be a difference as to what extent the PSTs may have had opportunities to develop their reflective practice. By using the model, it may be possible to highlight the attitudes in different ways, either in the moment in interaction with PSTs, by introducing the attitudes to the PSTs and discuss what it means to show responsibility, open-mindedness and whole-heartedness (Harrington et al., 1996), or implicitly by scrutinizing tasks to be designed in ways that may encourage the realization of the attitudes. This kind of support and the possibility to work collaboratively is of importance for the development of students' habits to reflect (Mann et al., 2009). This, in turn, may constitute ground for a deliberate development of the quality of PSTs' reflective practice through their education, by talking about responsibility, open-mindedness and whole-heartedness as explicit elements affecting the quality of reflection.
This model should be seen as a prototype, still belonging to the extraction phase of the didacticmodeling process (Wickman et al., 2020). It needs to be further extended both to include patterns not identified in our material and to consider possible relative importance of the three attitudes in different parts of the reflective process. Moreover, the basis for the prototype model so far has been PSTs' contemporaneous discursive reflections made together while working in groups. One way to strengthen the prototype may be to investigate PSTs' moments of both retrospective and anticipatory reflection and also to explore how the model may support moments of reflection made individually by the PSTs, as well as reflection made in other ways than through conversation.
It is likely that if PSTs are given possibilities to develop the quality of their reflective practice in more systematic and explicit ways, they will learn from these experiences and successively develop reflection as a habit (Dewey, 1916) in the sense of "a form of executive skill, of efficiency in doing" and "an ability to use natural conditions as means to ends" (Dewey, 1916, p. m.w.9.51). Loughran (1996) emphasizes that reflection should be used as a "tool for unpacking and learning from the uncertainties of practice" (p.17) and our model offers a way to unpack and create opportunities for developing reflection into such a tool.
A possible complementary framework to our model could be Mezirow's (1981) theory about critical reflection which could add another level of thinking about the quality of one's reflection. Our model focuses on the quality of reflection in relation to certain content, here planning a science unit. By creating a critical consciousness, the PSTs may develop their reflective practice also by being critical to the judgements they have made, and to understand that their assumptions and what they have been taken for granted may not always be beneficial to their reflection. This critical consciousness may help them to adopt the new perspectives (perspective transformation) which our model entails creating possibilities for PSTs to develop also a "habit to reflect about their habit to reflect", as it were. However, the possibility of using Mezirow's theory of critical reflection needs to be further investigated empirically.
Attitudes, as a concept, are not unambiguous and may have different meanings. On one hand attitudes may describe a person's character and temperament, on the other, a person's mindset and position. Dewey (1929) describes attitudes as "modes of feelings and thinking" (p.lw.5.102). However, instead of treating the attitudes as characteristics of individuals, the model offers a way of treating them as discursive tools for discerning and making explicit different elements of reflective practice. However, one should not imagine that the model communicates this use of the attitudes directly to teacher educators. Therefore, in conjunction with moments of reflection teacher educators and PSTs need to discuss the meanings of responsibility, open-mindedness and whole-heartedness in relation to the specific content and purposes beyond Dewey's definitions and how we operationalized them for the purpose of analysis. Of course, this may require that other, more holistic and subjective aspects of reflection are also considered in conjunction with the three attitudes, such as Mezirow's (1981) theory of critical reflection (see above).
To be able to keep up in a changing world and to meet new challenges teachers need to be reflective practitioners (Carlgren, 1999;Collin et al., 2013;Grant & Zeichner, 1984;Ruffinelli et al., 2020) and to be able to challenge previous experiences (Griffin, 2003). We believe that by the use of the model, teacher educators and PSTs may work together with an overarching purpose to develop PSTs' reflective practice so that, eventually, it may become an "active capacit[y] to readjust activity to meet new conditions" (Dewey, 1916, p. mw.9.57) in other words, a habit.

Fia:
Then it will be a greater focus on that rather than on taking a stand . . . or?

125.
Steve: This is part of the dynamics of working together in a group, there are so many things, there has to be. If the two of us are having a hard time to make an agreement but we still need to produce something, you or I may say "yes" to what the other person thinks but without being convinced that the other persons opinion is right. (Continued)