Thoughts Are Free—Differences Between Unstructured and Structured Reflections of Teachers with Different Levels of Expertise

Abstract

In teacher education research, the primary source of data used to measure teachers’ reflective skills are written reflection products, which are often collected in the context of field experiences following specific structural guidelines (e.g., portfolio texts). However, it is unclear how appropriate written products are for this purpose, considering teachers’ everyday professional lives, in which reflection is a mostly verbal, highly self-directed process depending on the teachers’ level of expertise. Therefore, in our study, we analyzed how teachers’ free, unstructured reflections differ from reflections structured by model-based reflection prompts. In an exploratory qualitative research design with theoretical sampling, a total of 22 prospective teachers at four different levels of expertise were asked to reflect on two standardized fictitious vignettes using a think-aloud approach. For the first vignette, participants reflected in an unstructured way. For the second vignette, the reflection was structured using simple model-based reflection prompts. On average, the participants showed a significantly better reflective performance in the structured condition, but no significant differences in relation to the level of expertise were observed. The results contribute to a better understanding of the validity of typical reflections used in teacher education as an indicator of reflective practice in the professional field.

1. Introduction

Reflection is considered a fundamental competence of teachers that plays an essential role in their professional development and throughout teacher education (Beauchamp, 2015; Kaçaniku et al., 2019; Liu, 2015). On several levels, it enables students and teachers to critically question their own pedagogical actions, analyze professional challenges, and make informed decisions for future actions (van Beveren et al., 2018). On the individual level, reflection contributes to the development of a professional identity as teachers reflect on their beliefs, values, pedagogical content, and pedagogical content knowledge (e.g., Stenberg, 2010). On the interpersonal level, it improves teacher–student interaction by enabling a deeper understanding of students’ learning processes and perspectives (e.g., Martin, 2005). On the socio-structural level, reflection enables teachers to critically examine educational processes and better understand the influence of institutional frameworks, social inequalities, or power structures on teaching (e.g., Rahmawati et al., 2013).

Although the concept of reflection is fuzzy (Clarà, 2015), these diverse effects of reflection are systematically captured and analyzed through scientific models (Lenske & Lohse-Bossenz, 2023). While process models (e.g., Korthagen & Kessels, 1999) describe the course of reflection processes, stage models (e.g., Hatton & Smith, 1995; van Manen, 1977) focus on the qualitative depth of reflection. This depth is usually expressed by distinguishing between different levels of reflection, ranging from the mere description of a situation to the theory-based analysis and critical examination of educational structures. These models serve not only as heuristics for systematizing and measuring reflection processes, but they also serve as normative standards for assessing their quality, with deeper (higher level) reflection usually being considered of higher value (e.g., Sööt & Viskus, 2015). On this basis, structured reflection formats—such as written portfolios, teaching vignettes, or video-based reflection tasks—have been developed, particularly in teacher education, to create targeted opportunities for reflection and to promote evidence-based professional learning (Larrivee, 2008).

Although these formalized models of reflection are important for research and teacher education, they differ from actual reflection practices in everyday school life (Ryken & Hamel, 2016; Shoffner, 2008). While reflection in both empirical studies and teacher education is often promoted in the context of written, standardized (measurement) procedures, in professional practice, it is predominantly conducted orally, spontaneously, or in collegial discussion, often without external structuring and embedded in concrete teaching situations. This discrepancy raises the question of the extent to which unstructured reflection formats, which presumably come closer to authentic reflection practices, allow for reflection processes of comparable depth to structured ones, the importance of which is so well documented.

Closely related to this question is the ambiguity of how the degree of structuring of a reflection situation interacts with the teachers’ level of experience. While it is recognized that different formats of reflection serve different functions depending on the context (Jakfar & Rahmatillah, 2023), there has been little research on the extent to which the degree of structuring facilitates or restricts the ability to play out one’s professional reflective potential at different levels of experience. For example, it is conceivable that novices may benefit more from explicit reflection aids to achieve higher levels of reflection, while experienced teachers may prefer more flexible, unstructured reflection formats because they rely more on implicit knowledge (Jalilifar et al., 2014). However, the extent to which a higher degree of structuring is necessary or possibly even a limitation for experiential reflection remains disregarded.

Against this backdrop, it is essential to systematically investigate whether and how structured reflections differ in quality from unstructured, spontaneous professional reflections. This leads to two central research desiderata: First, it needs to be clarified whether structured or unstructured reflection formats promote deeper, i.e., higher quality, reflection processes. Second, there is a need to investigate the relationships between the level of experience, the degree of structuring of a reflection situation, and the depth of reflection.

This study examines these questions empirically by analyzing the reflection processes of (prospective) teachers with different levels of experience in a qualitative interview design. The following two reflection conditions are compared: an unstructured reflection without specific guidelines and a structured reflection with model-based reflection prompts. The aim is to identify differences in the quality of reflection between the two formats and to uncover any correlations with the level of professional experience. The results should contribute to the development of theory in the field of teacher reflection and the possible further development of reflection (measurement) formats, as well as provide practical implications for teacher training to adapt reflection formats specifically to the needs of teachers with different levels of experience.

2. Background

2.1. Reflection as Part of Professional Competence

In teacher education, reflection is often understood as an active, ongoing, and careful thought process in which one’s beliefs and actions are critically examined in regard to their bases and possible consequences (Dewey, 1933; LaBoskey, 1993). Reflection is seen as a key skill for professional teachers, as it not only contributes to individual development—for example, by reflecting on one’s own professional experience—but it also promotes higher quality teaching and a more critical examination of school conditions (Philp-Clark & Grieshaber, 2024). Professional reflection has this potential if it is used dynamically and contextually, as described by Hatton and Smith (1995) in their stage model. There, reflection is understood as a multifaceted process that ranges from simple description to dialogical and critical-reflective discussion. Closely connected to professional demands, reflection—in the sense of Schön’s (1983) concept of the reflective practitioner—is particularly effective when it occurs in specific professional situations and contributes to continuous professionalization. It is central to both post action analysis, reflection-on-action, and as an immediate reflection during action, reflection-in-action. More recent approaches extend this understanding, focusing on the anticipatory reflection of future action (reflection-for-action, “future-oriented reflection before action”, Conway, 2001, p. 90). Through this extension, professional reflection can be seen as a key competence for high-quality teaching practice that enables teachers to design lessons in the interplay of their accumulated professional experience and theoretical knowledge while responding flexibly to unforeseen circumstances (Cattaneo & Motta, 2021). Thus, reflective practice provides an opportunity for the continuous development of professional qualities.

2.2. Models for Structuring and Evaluating Reflection

In order to describe the professional reflection processes that provide this opportunity and that go beyond people’s basic ability to reflect (McArdle & Coutts, 2003), various models have been developed that differ in their objectives. Process models depict the process of reflection as a cycle of successive stages. Gibbs’ (1988) Reflective Cycle, for example, is conceived as a sequence of description, feelings, evaluation, analysis, conclusion, and action plan. In a similar structure, but with a focus on self-directed learning from professional experience, Korthagen and Kessels (1999) use the ALACT model to describe professional reflection as a cycle of (1) Action, (2) Looking back on the action, (3) Awareness of essential aspects, (4) Creating alternative methods of action, and (5) Trial.

Central to this article are stage models that focus on the qualitative depth of reflection. For example, Hatton and Smith (1995) distinguish the following four levels of reflection: descriptive reflection as pure description, dialogic reflection with initial attempts at interpretation, critical reflection with consideration of contextual and ethical aspects, and technical reflection with a focus on goal attainment and efficiency. Korthagen and Vasalos (2005) extend these aspects in terms of professional identity development with their Core Reflection Model, which includes the teacher’s values, beliefs, and personal concerns in addition to his or her actions. In this way, reflection focuses not only on concrete behavior, but also on the development of one’s own professional stance.

To create a common theoretical basis for the analysis of professional reflection and, at the same time, to increase the comparability of empirical findings in research on teacher education and professional development, this article is based on the understanding of reflection as a competence (von Aufschnaiter et al., 2019; Blömeke et al., 2015). This understanding is consistent with a cognitive psychological perspective that sees reflection as a particular form of goal-directed thinking (Aeppli & Lötscher, 2016).

Based on the competency model of Blömeke et al. (2015), reflective competence is understood as the interplay of individual dispositions (e.g., knowledge, beliefs, self-regulation) and situation-specific skills (e.g., professional perception) that bring forth reflective performance. This performance refers to observable reflective behavior and can be described in terms of process-related characteristics.

Following Aeppli and Lötscher (2016) and Korthagen et al. (2001), the situation-specific components of reflective competence can be divided into three central dimensions: (1) cognitive activities (e.g., describing, evaluating, justifying), (2) objects of reflection (e.g., teaching behaviors, student reactions, teaching objectives), and (3) goals or reasons of reflection (e.g., promoting learning processes, professionalization, self-regulation). This integrative model allows for a multidimensional view of reflective competence that considers both the individual prerequisites and the dynamic thought processes that lead to professional reflective performance (Kost, 2019).

2.3. Structured vs. Unstructured Reflection

The models of reflection outlined here not only serve to improve the understanding of reflection, but they also serve as a basis for deriving assessment standards for reflection quality (Roberts, 2016). In teacher education, in particular, such models are used (as benchmarks) to make reflection skills measurable and to improve these skills subsequently. Written reflection formats, such as portfolios, journals, or video-based analysis tasks, are often used in structured settings.

Structured reflections are usually characterized by specific guiding questions or predefined frameworks. For example, Toom et al. (2015) investigated video-based guided reflection discussions with student teachers, in which a selected teaching event was analyzed step by step from different perspectives. Michos and Petko (2024) show how reflection can be stimulated through a mobile app, for example, through videos, images, or notes created directly in the classroom, expanded by written reflections. These formats help, in particular, novice teachers to structure their observations and link theory to practice. Similarly, Husu et al. (2008) describe such settings as “guided reflection”, where learners are specifically guided through reflection processes.

In addition, Krauskopf and Fehn-Winterling (2023) show, in a quasi-experimental study, that written reflections based on videotaped lessons lead to more differentiated and emotionally deeper discussions of one’s own practice than those without video material. This supports the targeted use of structured reflection formats, especially in phases of competence acquisition.

However, Williams and Grudnoff (2011) show that experienced teachers use the same reflection models differently than novices. They rely more on implicit experiential knowledge and reflect more freely, while novices benefit from clear structures. This shows that the same reflection structures are interpreted and processed differently by novices and experts, which is a finding that argues for differentiated support according to one’s experience during teacher training.

In contrast to structured reflection, unstructured reflections are freer, oftentimes verbal, and frequently occur in informal exchanges, for example, in conversations with colleagues or as spontaneous thoughts after class. Mena Marcos et al. (2009) describe such reflective processes as part of everyday practice. Studies also suggest that oral reflections may contain more nuanced justifications. Guo (2022) shows that diary reflections often remain more superficial compared to dialogic forms, while Allas et al. (2017) found that oral reflections rely more on contextual, often emotional knowledge. Written reflections, on the other hand, more often involve theory-based and abstract thinking, which can lead to a more generalized examination of experiences.

2.4. Prompts and the Effect of Structuring

Building upon the distinctions between structured and unstructured reflections, this section focuses specifically on the role of prompts as a tool for systematically enhancing reflection depth. Prompts—i.e., structuring instructions or guiding questions—are regarded as central instruments for promoting and evaluating in-depth reflection. They guide the reflection process in a targeted manner and help novices to recognize complex interrelationships (Jung et al., 2022). Research following the aptitude–treatment interaction (Cronbach & Snow, 1977) also shows that the effectiveness of prompts depends strongly on the level of experience of the teacher; while novices benefit from clear guiding structures, experienced teachers sometimes find too-strong guidelines restrictive (Gentner & Seufert, 2020).

With regard to reflection processes, Guo (2022) identifies different types of prompts, including cognitive (“What alternatives would you have had in this situation?”) and metacognitive (“Why did you make this decision?”) prompts. Combining these types of prompts can help address different levels of reflection and increase their depth.

2.5. Level of Expertise and Depth of Reflection

However, research on the development of expertise reveals clear differences in the way teachers perceive and reflect on teaching situations. Dreyfus et al. (1986) and Berliner (2001) point out that novices rely primarily on explicit rules, theories, and models to guide their actions, while experts draw on extensive experience, implicit knowledge, and routine, situation-specific strategies (Stender et al., 2021) show that the depth of reflection differs between these groups. While novices often stick to descriptive reflections, experts are more likely to reflect critically and contextually.

These findings are supported by other research. For example, Allen and Casbergue (1996) show that the quality of reflection changes systematically with experience, from descriptive and explanatory to evaluative and strategic planning. Similarly, Jung et al. (2022) examined the use of different prompts in structured reflection settings and found that standard and concept-based prompts, in particular, contributed to deeper reflection among less experienced teachers. Building on this, Jung (2012) found that experienced teachers predominantly engage in flexible and open-ended reflection processes, focusing on adapting their practices to classroom dynamics and reconstructing their professional beliefs, rather than relying on structured reflection formats.

Differences such as these have direct implications for the design of reflection tasks. While novices seem to need structuring reflection prompts to arrive at deeper reflections, experts benefit from open formats that are more closely tied to authentic classroom conditions and allow for greater freedom (Jakfar & Rahmatillah, 2023).

These conclusions are supported by findings on the so-called expertise reversal effect. While learners with low expertise benefit from clear instructions, these can lead to cognitive overload for experts and reduce the efficiency of their learning process (Rey & Buchwald, 2011). When applied to reflection processes, this means that novices need clear guidance structures in order to develop reflective skills, while experienced teachers should be given more autonomy in the choice and design of their reflection strategies.

Building on these findings, the present study aims to examine how unstructured and structured reflection formats interact with different levels of teaching expertise in shaping reflection quality.

2.6. Research Questions

Against the background of the theoretical approaches and empirical findings described above, the following central research desideratum emerges: it has not yet been sufficiently clarified how different reflection formats—in particular structured versus unstructured—influence the quality of professional reflection, and how the level of experience of teachers plays into this. While existing studies address individual aspects of these questions, a systematic analysis that considers both factors at the same time is lacking.

Because research on the interactions between degree of structuring, quality of reflection, and level of experience is still inconsistent and open to theory, we deliberately refrained from formulating precise hypotheses. Instead, this study pursues an exploratory, theory-based interest in knowledge with the aim of gaining differentiated insights into the organization of professional reflection processes.

This leads to the following research questions:

• How do structured and unstructured reflection formats differ in terms of reflection performance (e.g., in terms of depth of reflection achieved, content addressed, and theoretical references)?
• To what extent does teachers’ level of experience influence the quality of reflection in structured and unstructured reflection formats?

3. Method

3.1. Design

To explore the differences between unstructured reflections and those structured through model-based reflection prompts among (prospective) teachers, a qualitative interview design was chosen. Participants were asked to reflect on the teaching of other teachers, described in two standardized, fictional vignettes used in previous research (Stender et al., 2021). To authentically capture reflection processes, a think-aloud approach was employed, in which teachers verbalized their thoughts during the reflection process. The rationale behind this approach was to place participants in a context that more closely resembled reflection in teachers’ everyday practice (e.g., oral rather than written reflection, shorter or limited time frame, no mandatory documentation) in the sense of a more authentic assessment of reflection performance (Gulikers et al., 2008). However, it is still a reflection on another teacher’s lesson and not on one’s own.

Both vignettes were designed to display typical classroom situations. Vignette 1 describes a chemistry lesson, in which the teacher reviews experimental instructions with the class but fails to deal effectively with disruptions caused by students. As students proceed to conduct the experiment independently, misunderstandings arise, leading to a disorganized and partially unsuccessful implementation. Vignette 2 describes a math lesson in which the teacher initiates group work but fails to intervene when students work individually rather than collaboratively, resulting in only high-performing groups successfully solving the task. Although situated in two concrete subjects, both vignettes describe pedagogical difficulties without reference to the content of the lesson. Therefore, participants do not need to have any specific content knowledge or pedagogical content knowledge, and the vignettes can be used to evaluate the reflective performance of (prospective) teachers of all subjects and teacher education programs (Stender et al., 2021).

The interviews were conducted in the following two phases:

• Unstructured Reflection—Participants freely reflected on the first vignette without specific structural guidelines. The exact prompt for all respondents was: “Please read the situation description and reflect on the teacher’s behavior. Include theoretical approaches in your reflection. Please verbalize all your thoughts aloud—everything that comes to mind.” Participants were informed beforehand that, for methodological reasons, the interviewer would not react during this phase and that they should engage in expressing all their thoughts freely. After participants had articulated their responses, a brief moment of silence was observed before they were asked if they had completed their reflection. The interviewer then transitioned to the second phase of the interview.
• Structured Reflection—The reflection on the second vignette was guided by model-based reflection prompts provided by the interviewer. The instruction in this phase was: “Please read the situation description. Unlike in the first scenario, we will not begin the reflection immediately. Instead, please read through the situation first and let me know when you are done.” Subsequently, all participants were asked the following guiding questions (in that order):

  (a)

  What stands out in this situation? [Description]

  (b)

  How do you assess the teacher’s behavior? [Evaluation]

  (c)

  Are there criteria that led you to this assessment? [Explanation]

  (d)

  What do you think the students would think about this behavior? What effect do you believe the teacher’s behavior has on the students? [Analysis]

  (e)

  Have you encountered theoretical content related to [vignette 1/2] during your studies, and can you apply it to this situation? [Theoretical Reference]

  (f)

  How would you have reacted in this situation?/If you were to give advice to this teacher, what would it be? [Planning]

The vignettes were evenly distributed across the two interview phases and participants. Each participant therefore had to reflect on both vignettes, either in the first or in the second interview phase. All interviews were recorded, transcribed, and subsequently analyzed using structured qualitative content analysis according to Mayring (2021). The three overarching categories of analysis were developed deductively based on the following established models of reflective competence (Aeppli & Lötscher, 2016; von Aufschnaiter et al., 2019): (1) Objects of Reflection—Content Aspects; (2) Cognitive Activities—Level of Reflection; and (3) Target Dimension of “Professional” Reflection—Theoretical Reference. A comprehensive coding manual was developed to define, describe, and illustrate the main and subcategories with anchor examples. The manual was validated through expert interviews (Stender et al., 2021). An excerpt from the coding manual, presenting the categorization system used in the analysis along with essential descriptions of individual categories, is displayed in the

Table 1. Extract from the coding scheme.

Category	Subcategories: Vignette 1 (Chemistry)		Subcategories: Vignette 2 (Math)
Content Aspects	(1) Classroom management		(1) Homogeneous learning groups
	(2) Teacher clarity and lesson organization		(2) Time on task
	(3) Time on task		(3) Teacher clarity and lesson organization
	(4) Scaffolding		(4) Adaptive teaching competence
	(5) Lesson goal		(5) Lesson goal
	(6) Other		(6) Other
Level of Reflection	Level 1: Description	Level 1 of the reflection level category (Description) represents the ability to describe the situation or the teacher’s and students’ behavior in the vignette in one’s own words (What is the case?). In such a description (descriptive reflection), one’s perception is expressed by assigning characteristics to the situation or behavior without making a value judgment about the observed aspects. That is, the focus lies on stating what something is or how it functions, rather than evaluating whether it is good or bad.	Anchor example: “At the beginning, there is a disruption of the lesson that somehow happens with these paper balls. It’s mainly one pupil who is apparently disrupting the lesson, but this in turn has a certain ripple effect on other pupils, who are then also distracted, so in principle a certain ripple effect is triggered by this disruption. And so everyone is admonished, yes.”
	Level 2: Evaluation	Level 2 of the reflection level category (Evaluation) represents the ability to assess the situation or the teacher’s behavior in terms of its appropriateness for high-quality teaching as a standard (How is it?). This includes indirect negative evaluations through imperative statements (e.g., “She should not have done it this way”). Distinction from Level 5 (Planning) Participants often do not evaluate the situation “directly” (using adjectives such as “good,” “bad,” “hindering,” “beneficial,” etc.) but rather “indirectly” by naming alternatives derived from the situation description itself. These alternatives do not constitute a constructive (independent) development of action alternatives but merely indicate that a characteristic of the situation “should have been different.” For instance, if participants demand “more student-centered instruction,” this implies an indirect evaluation, which is coded as Level 2. Level 5 would only be coded if participants elaborated on the means or methods by which a higher degree of student-centered instruction could be achieved.	Anchor example: “Okay, so what I find problematic about this group division is that I immediately noticed that each student works on their own. Of course, it’s often the case in groups that everyone has to understand the task themselves first. But somehow the students don’t interact in their groups at all. And especially how the groups are divided up. So, for example, that the high-performing groups, the high-performing students are in one group and the low-performing students are in another, because I think that’s somehow not the point of group work, because the others aren’t pulled along or supported at all. And that’s why I think the teacher should divide up the groups a bit more fairly. Of course, that doesn’t always work, for example, if there are methods where groups are somehow thrown together or numbers are counted. But somehow, I don’t think the group division here is very well chosen. It’s just stupid that the low achievers end up understanding almost nothing. Yes, there should definitely be a way for the lower achievers to at least ask questions at the end or for things to be explained to them better. For example, one approach would be to mix up the groups again afterwards so that they can explain it to each other. But I think that’s rather a bad idea.”
	Level 3: Explanation	Level 3 of the reflection level category (Explanation) represents the ability to substantiate the evaluation of the situation or the teacher’s behavior by referring to underlying reasons (Why is it good or bad?). These reasons are anchored in the appropriateness of the behavior for achieving high-quality teaching (as defined in Level 2).	Anchor example: “So the teacher has now opted for homogeneous learning groups, i.e., a division with more people of similar ability, pupils, exactly. But that may not necessarily have been beneficial in any case. So, as I understand it here, the same text tasks or the same text task were given to all groups. So that doesn’t mean differentiated by level. And some students then had the problem that they were not able to solve the task in their group, because this group would be relatively low-achieving, for example, because they couldn’t understand it. And yes, in this respect, perhaps a heterogeneous division, i.e., with higher to lower-performing students, would have been better so that they could support each other in the learning process.”
	Level 4: Analysis	Level 4 of the reflection level category (Analysis) pertains to the ability to establish connections between the teacher’s actions and the resulting effects on the students in the classroom. This level focuses on understanding the causes and consequences of the teacher’s behavior, linking observed instructional practices to their impact on student learning and classroom dynamics.	Anchor example: “I suspect that they might either still see her as a little unsure […] I don’t know how close she really is to the students or whether she’s somehow teaching past them. And then the students don’t really feel like they’re being picked up on and then somehow lose interest in the whole thing and don’t take it as seriously or don’t feel like having to have everything explained to them again and again and the teacher is then simply the only one who can help them somehow and a feeling of dependency somehow arises. So yes, it’s not ideal, but in any case, they respect her as the person who should know. They then ask about the details, so they already have the feeling that the teacher definitely knows how to do it. Only in terms of authority, perhaps and transparency, the students might think yes, maybe she could have explained it differently or something.”
	Level 5: Planning	Level 5 of the reflection level category (Planning) refers to the ability to propose justified alternatives to the teacher’s classroom management behavior (How could it be improved?). Distinction from Level 2 (Evaluation) The suggested alternatives must be substantiated to distinguish them from indirect evaluations without a constructive component (Level 2). Justifications are not present when participants merely call for “more student-centered teaching” without elaborating on the methods or strategies that could achieve higher student engagement. The key criterion here is the constructive aspect of generating alternative actions.	Anchor example: “But I also think that the fact that the students ask about details in more detail, i.e., it doesn’t say here that they are asking about the basic implementation, but only about details, which for me would be rather minor aspects that may not have been understood, I also thought directly. [...] But when I read that, a lot of alternatives simply go straight through my head. And I would have used marble phases on the one hand. But above all, I would have tried to visualize the instructions for the experiment in some way using fixed material, which would perhaps remain somewhere on the blackboard, so that the whole process could really be retraced at any time. Because if the whole thing is only done orally and that’s how it seems here, since she clearly demands this concentration again and again, saying “Please concentrate, because this is important now!”, so basically appealing to the students, so you are doing this now so that it can go on well right away, it sounds as if it would have happened mainly orally and yes, I will also directly think of an alternative or a criterion why I see the whole thing relatively, yes, critically, that the whole thing is perhaps not sufficiently visualized and only explained orally.”
Theoretical Reference	Theoretical Reference	A theory relevant to the vignette is explicitly referenced. Participants must explicitly refer to a theory or empirical findings on teaching quality or use a concept or term that is the subject of a theoretical framework. Potential references include dimensions and empirical findings on teaching quality, classroom management techniques, and approaches to handling classroom disruptions.
	Lack of Theoretical Reference	The unit of analysis contains no explicitly mentioned theoretical foundation.

.

The five levels are defined by an increasing level of complexity and depth of reflection. Level 1 (Description) merely involves stating observations without judgment. Level 2 (Evaluation) adds a critical component, implicitly or explicitly judging the appropriateness of behaviors based on teaching standards. Level 3 (Explanation) further deepens reflection by requiring justifications of these judgments, explicitly linking them to standards of quality teaching. Level 4 (Analysis) moves beyond simple justification to a detailed examination of causal relationships between teaching actions and student outcomes. Lastly, Level 5 (Planning) distinguishes itself by focusing on constructive, actionable strategies aimed at improving future teaching practices, requiring substantiation of the proposed methods or approaches.

Each interview was thus coded three times, as follows: (1) with one or more addressed content areas (with a maximum of six content areas for both vignettes, as no double assignment of the same content area was made per unit of analysis); (2) with the highest level of reflection reached across the entire unit of analysis (even if multiple levels of reflection were identifiable); (3) with the indication of whether a theoretical reference was established within the unit of analysis (rather than the number of theoretical references). All transcripts were coded by two trained, independent raters. The interrater reliability was κ = 0.826, which can be interpreted as a good level of agreement (Landis & Koch, 1977). Appendix A provides a summary table of the codings for all cases.

Due to the small sample size, we used nonparametric tests to analyze differences be-tween variables (e.g., level of reflection) in different conditions. For example, we used the Wilcoxon signed-rank test to analyze differences between two conditions (e.g., unstructured vs. structured reflections) and the Kruskal–Wallis test to analyze differences between more than two conditions (e.g., levels of expertise). These tests do not require the variables to be distributed in a certain way and only require ordinal scaling.

3.2. Sample

In order to investigate differences in reflective performance in relation to teachers’ level of expertise, participants were selected from the following four successive career stages in accordance with the purposeful sampling approach (Patton, 2014): teacher education students in bachelor programs, teacher education students in master programs, teacher trainees in the induction phase after studies, mandatory in the German context, and experienced teachers. The four levels differ mainly in the amount of experience participants have been able to acquire regarding the professional tasks of teachers in an everyday school context. As part of their bachelor’s degree, teacher education students are required to complete a six-week work placement in a school. On the master’s program, student teachers take part in a one-semester, five-month school internship. In the induction phase after studies, teacher trainees spend 1.5 years training in a school under the supervision of experienced teachers, while also attending courses at centers for practical teacher training. Experienced teachers fulfill all the tasks of a teacher in the full complexity of school conditions without mandatory further professional development (professional development is voluntary). Over the course of the four levels, the focus of teacher education changes from a stronger emphasis on theoretical content to a greater emphasis on developing skills for professional duties in the workplace.

In general, the recruitment of participants followed convenience sampling, albeit also being stratified to guarantee that all four career stages are represented as described above. The participants were informed about the study before taking part voluntarily; they were free to withdraw at any time without disadvantages. Their details were kept confidential.

The sample consists of N = 22 (prospective) teachers, including six bachelor’s teacher education students, six master’s teacher education students, four teacher trainees, and six experienced teachers. At the time of the survey, the respondents had a mean age of 26 years (SD = 4). A total of 19 participants identified as female, three as male. Nine of the (prospective) teachers who resp. studied to work in elementary schools, ten in secondary schools.

4. Results

4.1. Performance in Unstructured and Structured Reflections

We conducted two Wilcoxon signed-rank tests to test whether participants’ reflective performance varied between the two vignettes regardless of the interview phase. The first analysis compared the number of achieved content areas in each vignette (max. six) and revealed no statistically significant differences between the two vignettes (W = 2.071, p = 0.280). The second analysis compared the highest reflection levels (1–5) achieved in each vignette and also found no significant differences (W = 1.833, p = 0.310). In terms of references to theory, only two participants made such a reference in vignette 2. However, as the participants made almost no references to theory overall, it is likely that this difference between the vignettes is more of a coincidence. We therefore assume that differences in reflective performance are not due to the type of vignette and that the vignettes can be used to analyze the difference between unstructured and structured reflective performance.

The descriptive results of the reflection performance in relation to the two interview phases are presented in

Table 2. Reflection performance: unstructured vs. structured reflection.

Category		Unstructured Reflection	Structured Reflection
Content Aspects	Median	3.00	3.00
	Mean	3.05	2.95
	SD	1.09	0.95
Level of Reflection	Median	2.50	3.00
	Mean	2.64	3.32
	SD	0.79	0.95
Theoretical Reference (in %)	yes	4.5	4.5
	no	95.5	95.5

.

To examine differences between unstructured and structured reflections, two Wilcoxon signed-rank tests were conducted. The first analysis compared the total number of achieved content areas and revealed no statistically significant difference between the two conditions (Z = −0.384, p = 0.701). The median number of achieved content areas was 3.00 for both structured and unstructured reflections, suggesting that participants achieved a comparable number of content areas in both task types. The second analysis compared the reflection levels (1–5) achieved in the two conditions and found a significant difference (Z = −2.385, p = 0.017). The median level achieved in unstructured reflections was 2.50 (evaluation), while in structured reflections, it was 3.00 (explanation), indicating that participants tended to achieve higher levels in structured reflections. Achieving a higher reflection level indicates a shift from mere description or evaluation toward more in-depth explanations and analyses of teaching situations, which is associated with higher professional reflection quality.

Regarding the use of references to theory, no differences could be observed between the two reflection processes. In both the unstructured and structured reflections, participants made almost no reference to theory, indicating a consistent pattern of non-use of theoretical or empirical resources. In qualitative analyses of the structured reflection processes, it can be seen that the participants tended to respond to the prompt to include theoretical references by repeating the explanations of the previous prompts in different words, but they did not adopt a theoretical perspective on the vignettes.

4.2. Differences Regarding Level of Expertise

To examine differences between levels of expertise, two Kruskal–Wallis tests were conducted. The first analysis compared the mean of the number of achieved content areas across the two vignettes (max. six) and revealed no statistically significant differences between the four levels of participants’ expertise (W = 0.543, p = 0.909). The second analysis compared the highest reflection levels (1–5) achieved in the two vignettes and found no significant differences (W = 0.855, p = 0.839).

However, in order to explore possible differences between levels of expertise in more detail,

Table 3. Reflection performance: level of expertise.

Category		Bachelor		Master		Trainees		Experienced
		unstr.	str.	unstr.	str.	unstr.	str.	unstr.	str.
Content Aspects	Median	3.00	2.50	3.50	3.00	3.00	3.00	2.50	3.00
	Mean	3.00	2.83	3.33	3.17	3.00	2.75	2.83	3.00
	SD	0.63	0.98	1.21	1.47	1.63	0.50	1.17	0.63
Level of Reflection	Median	2.50	3.00	2.00	3.00	3.00	3.50	2.50	4.00
	Mean	2.50	3.00	2.67	3.33	2.75	3.50	2.67	3.50
	SD	0.55	0.90	1.21	1.03	0.50	1.29	0.82	0.84
Theoretical Reference (in %)	yes	0.0	0.0	16.7	0.0	0.0	25.0	0.0	0.0
	no	100.0	100.0	83.3	100.0	100.0	75.0	100.0	100.0

presents the descriptive results for unstructured and structured reflection performance for each level of expertise.

Descriptive analyses by expertise levels suggest that structured reflection prompts led to higher reflection levels across all groups, with the most pronounced increase observed among experienced teachers.

Even though the differences between unstructured and structured reflection regarding the number of content areas achieved are not statistically significant, it is worth noting that, on average, bachelor and master teacher students as well as teacher trainees addressed a slightly lower number of content areas in structured reflection than in unstructured reflection. However, the results are the opposite for experienced teachers. Regardless of their level of expertise, all participants showed a higher average level of reflection in the structured reflection. It is interesting to note that the difference in the mean values was greatest for the group of experienced teachers. Descriptively, no systematic differences in the four levels of expertise with respect to theoretical references could be observed, analogous to the previous analyses.

5. Discussion

In this interview study, we investigated the reflective performance of teachers with different levels of expertise in a setting that is fairly typical of everyday reflective practices of teachers, predominantly spontaneous, verbal, and based on specific teaching situations (Mena Marcos et al., 2009). Notably, such reflection is seldom documented in written form. Therefore, everyday reflection is also likely to be less systematic and structured compared to written reflection formats often used in teacher education. To explore this assumption in more detail, we conducted a think-aloud interview study to investigate the extent to which teachers’ oral reflection performance on two teaching vignettes with different levels of expertise differed depending on whether they reflected freely and spontaneously or whether the reflection was structured by (simple) cognitive interview prompts.

Overall, the participants showed on average a significantly better performance in the structured reflection in terms of the level of reflection. These findings are consistent with previous studies in which (prospective) teachers used structuring tools for written reflection (e.g., Michos & Petko, 2024; Krauskopf & Fehn-Winterling, 2023). The results also show that (prospective) teachers very rarely refer to theoretical knowledge in their reflections. Studies of written reflections have also found this to be the case (Stender et al., 2021). This persistent finding emphasizes the following significant challenge: theoretical knowledge seems to remain predominantly abstract and detached from teachers’ immediate professional practice (Korthagen & Kessels, 1999). Consequently, future interventions in teacher education might need to systematically bridge theory and practice through authentic, situationally embedded reflection formats (Toom et al., 2015).

In our study, however, we did not observe differences in terms of content aspects. Structured reflection, therefore, appears to influence how teachers reflect rather than what they reflect on. Since our prompts explicitly targeted reflective depth, this finding is not surprising. Nonetheless, it suggests the following potential methodological limitation: the reflection prompts might not have sufficiently encouraged participants to explore additional or alternative content aspects. Future research should therefore test specifically designed content-related prompts, though this might limit direct comparability between structured and unstructured reflections.

There were no overall differences between structured and unstructured reflection performance with regard to the level of expertise, contradicting previous findings (Allen & Casbergue, 1996). Nonetheless, the descriptive data provide several noteworthy insights. In terms of content aspects, the results align with prior research on expertise differences (Williams & Grudnoff, 2011). Under everyday professional conditions, i.e., during unstructured reflection, experienced teachers tend to analyze or reflect on situations holistically, often unconsciously. Their experience allows them to focus more rapidly on aspects they deem important. This is reflected in a lower average number of content areas addressed in their responses. Structured reflection prompts may encourage experienced teachers to consider additional content aspects. In contrast, novice teachers tend to consider a wider range of aspects equally and have greater difficulty prioritizing them during reflection. This leads to a higher average number of content areas addressed in their unstructured reflections. Accordingly, the structuring prompts had different effects depending on the level of expertise. The prompts of the structured reflection can encourage beginners to focus on content areas and explore them in more depth. Regarding the level of reflection, the prompts had a positive effect across all four levels of expertise. The most pronounced difference between unstructured and structured reflection was found among experienced teachers, which is a result that was unexpected in light of previous research (Jung et al., 2022). One explanation for this unexpected result could be that structured prompts reactivated formal reflection processes or theoretical frameworks, which experienced teachers had internalized but seldom explicitly accessed due to their routine-based reflective practices (Berliner, 2001; Rey & Buchwald, 2011). Experienced teachers might therefore particularly benefit from occasional structured interventions that reintroduce a more explicit, analytical mode of reflection into their habitual practice.

These findings strongly suggest the importance of a differentiated approach in de-signing reflection opportunities according to teachers’ expertise. Specifically, structured prompts seem universally beneficial for enhancing reflective depth, but their function and frequency should differ by expertise level: novices require systematic scaffolding to focus their reflections, whereas experienced teachers could profit from periodic structured prompts to activate and refresh their explicit theoretical knowledge (Toom et al., 2015; Rey & Buchwald, 2011).

Several limitations of our study should be taken into account when interpreting these results. Although the sample was well balanced across expertise levels through theoretical sampling, its size remained relatively small, which generally complicates statistical analysis. Furthermore, while participants were recruited from two different teacher education institutions in Germany, they are not representative of all (prospective) teachers. To validate the findings, future studies should include larger sample sizes. As with all think-aloud studies, there is an inherent risk that participants may not verbalize all of their thoughts, potentially leading to an underestimation of reflective performance. Additionally, the specific content of the vignettes may have influenced participants’ responses. It is possible that different results would have been obtained if the vignettes had focused on other content areas. However, the data provide no clear indication of such an effect. Therefore, future studies should systematically vary the contents and contexts of vignettes to assess the robustness and generalizability of our findings. Additionally, future studies with similar designs could further investigate interpretations based on our results by validating data gathered in think-aloud interviews using additional methods, such as member checking.

6. Conclusions

Overall, it can be concluded that, even in oral reflections, which are characteristic of teachers’ everyday reflections, relatively simple prompts can enhance the quality of reflection. Such prompts would be easy to implement in practice (e.g., in the form of a brief set of guiding questions to consult after a lesson), making their use more a matter of habit and routine. Given their simplicity and practical feasibility, teacher education and professional development programs should systematically integrate brief structured prompts into routine reflection practices, thereby fostering the habitual use of theory-based reasoning in everyday teaching situations (Michos & Petko, 2024).

Over time, these low-threshold interventions could potentially have a greater practical impact than the more complex written reflection formats commonly used in teacher education programs. This is not to suggest that written reflection is unimportant for the development of reflective competencies. However, teacher education curricula should be adapted to better prepare prospective teachers for the kinds of reflection that are typical in day-to-day classroom practice. Additionally, in recognizing the differential needs of teachers across expertise levels, professional development for experienced teachers should increasingly offer adaptive and flexible reflection formats, balancing implicit experiential knowledge with explicit, theoretically grounded reflection.

Regardless of the format, a persistent challenge remains, which is that teachers often struggle to incorporate theoretical knowledge into their reflections, or they may consciously choose not to do so. Future research should therefore directly explore under which circumstances teachers spontaneously and effectively integrate theoretical knowledge into their professional reflections. This includes examining how initial teacher education, continued professional learning experiences, and specific reflection contexts influence theoretical integration in practice (Korthagen & Kessels, 1999). By addressing the gap between research-oriented structured reflection and the realities of everyday reflective practice, this study provides new insights into designing reflection formats that are both authentic and conducive to professional growth. Ultimately, a differentiated and context-sensitive approach to fostering reflective competencies emerges as essential for supporting sustained professional development and high-quality teaching practice.