During collaboration, two or more learners are involved in interaction with each other as well as with tools and resources, working toward a shared goal (Bedwell et al., 2012; Dillenbourg, 1999; Patel et al., 2012). Within this process, students actively engage in constructing and maintaining a joint problem space through sharing, expanding on and negotiating ideas (Borge and Mercier, 2019; Roschelle and Teasley, 1995). Ideally, the collaborative group achieves both a common solution to solve the problem as well as skill and knowledge gain of each group member (Schürmann et al., 2024).

The importance of collaborative working and learning approaches has already been known for a long time and research on collaboration has a long tradition. Collaborative settings encourage students to participate in discussions that not only deepen their understanding of the material but also help them develop shared mental models of complex phenomena, promoting positive interdependence that nurtures collaboration and mutual support in achieving common goals (Johnson et al., 2014). As emphasized by Scager et al. (2016), cognitive benefits (e.g., conceptual understanding) as well as an improvement of social skills are provided through collaborative learning approaches. Therefore, using collaborative learning approaches might function as both a mean (collaborating to learn, i.e., knowledge construction, solving problems) and an objective (learning to collaborate, i.e., development of social skills; see Salmons, 2019). These dimensions represent two traditions in the collaboration literature. One perspective is grounded in the constructivist theories of Piaget and Vygotsky, which argue that knowledge is actively built through social interaction and dialogue, and the other perspective represents a shift toward recognizing collaboration as an essential human skill (see Child and Shaw, 2019; Rowe, 2020).

While several studies highlight the effectiveness of learning and working together in teams (e.g., Lou and Macgregor, 2004; Nokes-Malach et al., 2015; Tao and Gunstone, 1999), collaborative learning is not always superior to individual learning (Retnowati et al., 2018). Moreover, research and practice show that students are indecisive, predominantly skeptical about group work because of interpersonal conflicts, logistical challenges such as scheduling meetings and time management issues as well as unequal contributions of group members (Donelan and Kear, 2024; Scager et al., 2016; Wilson et al., 2018). This is particularly true for students at higher educational levels as they often work on challenging, ill-defined tasks characterized by ambiguity while receiving less support compared to earlier stages of education. Their collaboration is mostly self-organized outside class hours and in absence of instructors (Scager et al., 2016). Additionally, higher education students already bring with them a range of prior experiences and attitudes toward group work which, in turn, can influence team orientation, conflict resolution and outcomes (Fransen et al., 2013; Pauli et al., 2008).

Although collaborative learning approaches are ubiquitous, students at various educational levels are often not properly instructed on how to work in groups (Le et al., 2018; Leopold and Smith, 2020). Research suggests that instructors often engage primarily in logistical aspects of group work, such as assigning tasks and group size, rather than actively facilitating collaboration (Xu, 2024). Consequently, higher education students often report a lack of collaborative skills leaving them feeling ill-prepared for collaborative work—a phenomenon observed globally, including in both Western (e.g., Wilson et al., 2018) and Eastern contexts (e.g., Xu, 2024). As summarized by Borge and Mercier (2019), “collaboration is a collective cognitive endeavor that is hard, requires sustained effort from participants to work well, and is prone to breakdowns caused by different socio-emotional, cognitive, metacognitive, and socio-metacognitive problems” (p. 220). Typical problems encountered in student group work include different task understandings, altering levels of interests and commitment, or incompatible working and interaction styles, among others (Näykki et al., 2014). Especially in interdisciplinary student groups, collaboration issues often arise due to conflicting schedules, competing commitments (e.g., extracurricular activities), and differing priorities (see Hussein, 2021). Furthermore, the rise of virtual and hybrid learning settings since the COVID-19 pandemic has introduced additional challenges for students, particularly in managing their socially shared regulation of learning (e.g., Donelan and Kear, 2024; Oshima et al., 2024; Wildman et al., 2021).

But not only students face various obstacles with regard to collaborative learning, so do higher education instructors. For instance, they are struggling with crowded curricula and limited resources. Besides, although many higher education instructors are strong subject matter experts, they have received little pedagogical training. This impedes the selection of appropriate collaborative settings and tasks as well as giving instructions on how to work together successfully. Therefore, the focus of their teaching often is on disciplinary knowledge while collaborative learning goals take a back seat, with collaboration skills typically treated as a mere byproduct of group work (Boud and Bearman, 2024). Yet, even if instructors aim to develop both cognitive and collaborative skills through applying collaborative learning practices, they face challenges in assessing the collaborative processes and skills (Boud and Bearman, 2024; Le et al., 2018).

Based on John Dewey’s criteria for reflection, Rodgers (2002) describes reflection as “a meaning-making process that moves a learner from one experience into the next with deeper understanding of its relationships with and connections to other experiences and ideas” (p. 845). Hence, reflection is a purposeful thinking process involving affect and cognition, which is controlled by a learner (Nguyen et al., 2014; Yukawa, 2006). This process engages critical thought about an experience and generates insights that can inform future learning and development (Quinton and Smallbone, 2010). Besides reflecting on one’s own thoughts and actions, it is promising to reflect in interaction with others about collaborative processes. Collaborative environments provide greater opportunities for reflection compared to individual settings, as they facilitate the exchange of diverse perspectives and promote deeper critical engagement with experiences (Yang and Choi, 2023). Reflecting on peer feedback is expected to enhance group members’ awareness of their own behavior, its impact on others, and the potential need for adjustments (Eshuis et al., 2019; Phielix et al., 2011). By doing so, not only self-regulation but also mutual understanding and coordination within the group can be fostered. For instance, in medical education, reflective practice often takes place in form of orally conducted collaborative discussions giving students the opportunity to prepare themselves for their career including learning to collaborate (Johansson et al., 2017; Paige et al., 2021). This technique is referred to as debriefing.

Originally stemming from high reliability industries such as military and medicine, debriefings are a well-established and systematically applied concept for learning in teams in these contexts (Allen et al., 2018; Keiser and Arthur, 2021). Debriefings are structured group reflexivity interventions, in which teams are asked to reflect about recent events during which they collaborated (Allen et al., 2018). This process is often led by a facilitator but is associated with comparable effectiveness if self-led (Boet et al., 2011; Keiser and Arthur, 2021). Particularly with the advent of agile working methods, similar concepts can be found under different terms in many other working contexts (e.g., retrospective meetings conducted by scrum project teams). What they have in common is that they are based on the concept of experiential learning (Kolb, 1984), and combine feedback, reflection, and discussion (Keiser and Arthur, 2021). The purposes of group reflexivity interventions are manifold inclusive of improving teams’ problem-solving and decision-making processes, establishing psychological safety, enhancing group identity, and reducing failures (e.g., Allen et al., 2018; Schippers et al., 2014; Schippers and Rus, 2021). Their effectiveness in terms of several criteria such as individual and team performance has been confirmed in a number of studies including meta-analyses (e.g., Keiser and Arthur, 2021; Tannenbaum and Cerasoli, 2013).

Considering the importance of reflective practices for lifelong learning, reflection in higher education is receiving increased attention in recent years (e.g., Tan, 2021; Veine et al., 2020). As demonstrated by Merkebu et al. (2023), factors enhancing reflection are metacognition and emotional regulation. In general, the concepts of metacognition and regulation (e.g., self-regulation, socially shared or emotional regulation) have been advancing in significance within educational theory, research and practice (e.g., Hadwin and Oshige, 2011; Järvelä et al., 2015; Järvenoja et al., 2013; Kaplan, 2008). However, as of yet, reflective practices are predominantly common and studied in medical (e.g., Dornan et al., 2019; van Braak et al., 2021) and teacher education (e.g., Mortari, 2012; Slade et al., 2019) but rather scarce in other disciplines (Chan and Lee, 2021).

Moreover, research and practice revealed that students tend to struggle with reflection tasks (e.g., Moreland and McMinn, 2010; Tan, 2021). In their literature review, Chan and Lee (2021) refer to misconceptions by both students and teachers. Challenges at the student level include little motivation, lacking understanding of and ability to engage in reflection, as well as emotional impact through feelings of stress, vulnerability and anxiety (Chan and Lee, 2021; Tan, 2021). Examples for challenges at the instructor level are difficulties with pedagogical logistics like large classes and tight teaching schedules, choosing appropriate reflection approaches as well as assessment and feedback issues due to subjectivity and lacking criteria (Chan and Lee, 2021). Typically, students are asked to reflect individually and in written form such as through learning journals, diaries and portfolios (Chan and Lee, 2021; Heymann et al., 2022). Nevertheless, some research findings indicate that writing is not students’ preferred mode of expression (see Chan and Lee, 2021).

Although some students might have a natural tendency to reflect on a collaborative learning experience, it is rather unlikely that students reflect together as a group and in a systematic way about what happened and what could have been improved (see Fanning and Gaba, 2007; Harrison et al., 2003). Considering the positive findings of group reflexivity interventions (i.e., debriefings) in medical education and workplace settings, the current study transfers their systematic approach to higher education context.

Asking and supporting students to reflect regularly on their collaboration seems to be promising in various ways: Engaging in a joint meaning-making process of experienced collaborative work might help students to overcome challenges like task understanding and altering levels of interests or working styles. Additionally, students’ (re-)configuration of internal collaboration scripts (i.e., learner’s knowledge that guides understanding and action in collaborative practices, Fischer et al., 2013; Kollar et al., 2007) could be fostered. As students often find it difficult to raise sensitive issues (e.g., incompatible interaction styles or sharing the feeling of frustration due to low commitment of some collaborative partners) in collaborative work, debriefings provide a good opportunity for students to address these issues as they are explicitly asked to. Furthermore, the systematic and regular implementation of reflexivity tasks might help to raise students’ understanding of and ability to engage in reflection.

Debriefings should be ideally designed in accordance with the participants’ needs (Allen et al., 2018), such being provided with a clear structure and guidance during joint reflections (Schürmann et al., 2025). Behaviorally Anchored Rating Scales (BARS) might be promising to this end. In this rating format, performance dimensions and scale values are defined in behavioral terms (Schwab et al., 1975). Typically, rating scales ranging from five to 10 points are employed, incorporating behavioral anchors to illustrate varying levels of quality within a given construct, with higher scale values indicating more effective behaviors (Debnath et al., 2015). Research indicates promising findings in terms of their use for self- and peer-evaluation purposes (Ohland et al., 2012) and their feedback potential (Hom et al., 1982). Moreover, BARS have been successfully applied in team-based contexts, particularly for assessing team member effectiveness and evaluating team adaptation processes (Georganta and Brodbeck, 2020; Ohland et al., 2012). By providing information on what constitutes a poor, a satisfying and a good performance, using BARS might (re)configure students’ knowledge about collaboration (see Ohland et al., 2012) and stimulate reflection processes. Accordingly, BARS might be a beneficial tool to support and structure students’ self-led debriefings.

Taken together, the current study explores the application of structured and regularly conducted group reflexivity interventions (i.e., debriefings) in higher education context. We attempt to understand whether and how this approach helps students working in collaborative learning environments. In addition, this study investigates the potential of BARS as a tool to support students’ debriefings. We aim to answer the following questions, thus yielding valuable implications for higher education practitioners:

Our assumption is that students’ collaboration improves through regular reflection. We further postulate that the debriefings are perceived as useful for learning about how to collaborate. While RQ 2 explores the general effectiveness of regular debriefings in fostering collaboration within the context of higher education, RQ 3 delves more specifically into the role of BARS and examines their value in enhancing the quality and impact of debriefings for students.

The study was conducted in a one semester-long, interdisciplinary project course offered annually at a university in Germany. In this course, all faculty students being in their fifth or higher semester can choose from a range of topics presented by instructors of various disciplines. The learning objectives include encouraging students to discover new topics and enabling them to work together in international and interdisciplinary teams in authentic learning settings. Each summer term, students receive basic information about the project options; afterwards, they can enroll in their favored project taking place in winter term. The interdisciplinary projects take places throughout the entire winter term, i.e., 15 weeks.

The project investigated in the current study is one of 16 interdisciplinary projects (N = 199 participants) that took place in winter term 2022/2023. It dealt with creating a podcast for supporting students during thesis writing. Prior to enrollment, students were informed that this project is to be accompanied by a case study. The course limit was set to 16 participants. After the first week, one student withdrew her enrollment due to personal reasons.

At the beginning of the course, all students received an instruction on collaboration. The aim of the instruction was to inform students about the main characteristics of student collaboration while highlighting behaviors representing an effective collaboration (see Eshuis et al., 2019) based on the framework on collaboration in higher education by Schürmann et al. (2024). It was delivered by the course instructor (VS). Moreover, one team member of each team volunteered as so called multiplicator. The four multiplicators were psychology students who were instructed on how to implement, conduct and document a debriefing (focus: collaboration) within their teams. They received a document with explanations and guiding questions for the debriefing. Furthermore, they were compensated for their additional efforts with research participation credit. As depicted in Figure 1, we combined several research methods to collect quantitative and qualitative data on both individual and team level. Data were collected before (t₁), during (t₂-t₅) and after (t_6, t₇) student teams’ main working phase. The first author took field notes throughout the project course to document her observations.

Table 2 and Figure 2 depict how student teams’ collaboration as rated during the debriefings developed over the semester. Findings indicate a positive development of students’ self-assessed collaboration quality for three of the four teams. Looking at the sub-facets of collaboration, the teams felt most confident in Planning Activities, Handling Tools and Resources and Establishing a Positive Atmosphere and Cohesion. On the other hand, Monitoring and Reflecting, Managing Emotions and Individual and Joint Participation were rated lowest (see Table A1).

Team 1 initially showed an increase in collaboration scores, which then fell intensively during the final debriefing. Taking the qualitative data into account, it becomes evident that the interviewees underestimated potential problems in the beginning and team members did not raise their issues in the debriefings until the last one. Two sub-groups (psychology, business) were formed with one sub-group feeling more actively involved in the group work than the other. Although this was thematized in the beginning, the students found it challenging to repeat on addressing this issue. For instance, one team member highlights that one hindering factor was that they got along very well personally. It seems like they did not want to risk the positive atmosphere with addressing negative issues. However, during the last debriefing they discussed their issues very honestly. One member of team 1 believes “that if this last reflection meeting had not taken place, it would have remained unspoken in the end.”

Team 2 also experienced the formation of two sub-groups (psychology, business). Prior to the first debriefing, the psychology students sought dialogue with the course instructor to complain about the unbalanced division of work and the passive behavior of their teammates. They raised their issues within the first debriefing leading to a slight improvement afterwards (see D2). However, coordination and joint participation remained a challenge until the course end (see Table A1). For instance, during the group interview, interdisciplinarity came up as a potential reason.

Team 3 reported the highest self-assessed collaboration values. This is supported by the open text-field answers and interview data in which team members report that they functioned very well as a group.

Team 4 behaved rather unobtrusively during the semester. Despite having the lowest collaboration values for D1 to D3, they did not report on any challenges or troubles during group work during the semester. After initially increasing slightly, their collaboration values rose sharply toward the fourth debriefing.

In addition to the quantitative data gained in the debriefings, the analyses of the field notes, open-text fields and interview data further supports these findings. Specifically, two teams (teams 1 and 2) reported challenges and difficulties during group work. During coding, we also collected information on whether the debriefings lead to any modifications during student’s group work. Whereas member(s) of team 1 experienced hardly any recognizable change other than poorer interaction and atmosphere, statements of team 3 and 4 suggest behavioral adoptions after the debriefings (“In the first reflection it turned out that I contributed a little less to the group, which could be discussed immediately and improved for the next reflection,” member of team 4). Furthermore, they reported an improved communication (e.g., member of team 3) and highlighted that the debriefings helped to immediately clarify disagreements (e.g., member of team 4). Taken together, the qualitative findings provide a nuanced perspective that complements the teams’ self-assessments gathered during the debriefings.

During data analysis, it became evident that the responses to RQ 2 and RQ 3 were intertwined, making it challenging to draw clear boundaries between them. Consequently, we will combine our findings related to both research questions. In the upcoming sections, we provide a detailed analysis of the results, offering insights that not only address the research questions but also reveal additional challenges and limitations that extend beyond the scope of the original inquiries. While the responses to RQ 2 and RQ 3 are discussed in an intertwined manner throughout the text, Table 3 presents the core findings specifically related to each research question, allowing for a focused overview of the key answers. The selection of the most relevant codes was based on their frequency and the information gained during the member checking procedure.

In the following, particular attention is given to the insights from team 1 and team 3 allowing for a deeper understanding of how joint reflections function in varied collaborative contexts. Whereas team 1 encountered significant challenges during their collaboration, team 3 experienced a smooth and successful working process with no major problems. By comparing these two contrasting cases, we provide insights into how teams with different experiences benefit from the debriefing process. Due to the breadth of the coding scheme, it is not possible to present all categories in detail (please see the Supplementary material for a complete overview of the coding scheme). Hence, we will focus on the most frequently occurring categories, as well as those identified as particularly important during the member checking process as these categories were highlighted by participants as central to their experiences.

This study transferred the well-established intervention of debriefings to the higher education context, while innovatively integrating BARS into the debriefing process. Using an embedded case study design with mixed methods, this study has a small sample size, which limits generalizability. However, it provides rich and nuanced insights into student collaboration. The analysis of contrasting cases has deepened our understanding of how teams with varying experiences benefit from the integration of regular reflections. Further strengths are data triangulation and methodological rigor (e.g., member checking) which enhanced reliability of the results by incorporating multiple perspectives and methods. Additionally, we combine insights of team research and educational research allowing for a nuanced understanding of the findings. In response to the call for incorporating qualitative data on team reflection (e.g., Gabelica et al., 2014), our article offers deep insights into the characteristics and quality of joint reflections in student teams. Moreover, our study complements previous research by employing a time-series design, which sheds light on the repeated support of student collaboration (cf. Eshuis et al., 2019), and by highlighting challenges when students are asked to engage with multiple perspectives during reflection (cf. Tan, 2021).

Despite these strengths, several limitations must be acknowledged. A primary limitation is the potential for self-selection bias, as participants who chose to engage in the interviews were mainly psychology students. This may have skewed the results and limit the transferability of the findings. Considering the formation of sub-groups, it would have been valuable to gather perspectives from both “sides” for a more balanced view. Another limitation is the heterogeneity of the teams, as differences in composition and experience, for example, may have influenced the findings. To ensure fairness, we decided that all teams should undergo the debriefing intervention. However, without a control group, the study lacks a comparative baseline. Due to the small sample size and the focus on a few specific cases, the results cannot be easily extrapolated to a wider population. At the same time, based on our experience working with a wide range of student teams over many years, we have no reason to believe that the insights gained here are not relevant to other similar contexts.

The outcomes of the current study offer valuable guidance for researchers and higher education practitioners. In the following, we outline and discuss the implications derived directly from our findings.

In summary, instructors are encouraged to integrate regular reflections as this approach shows potential for improving students’ collaborative experiences by prompting them to focus not only on task-related work but also on critically evaluating their own behavior within groups. Importantly, joint reflections should be integrated during course time. While holding four reflective sessions over the course of one semester was considered as too much by the students, we propose limiting this to two or three sessions of 30 min each. However, it is important to note that reducing the frequency of reflections may inadvertently shift the focus toward taskwork at the expense of teamwork. Without regular opportunities for students to reflect on team dynamics and improve their collaboration skills, the emphasis may fall primarily on completing tasks, sidelining the critical aspect of teamwork development. The optimal frequency and duration can vary considerably between teams, making a flexible strategy more effective than a uniform approach.

Previous research indicates that students require structure and guidance during joint reflections (Schürmann et al., 2025). To prevent superficial discussions and improve the depth of reflections, it is essential to provide clear and precise instructions. In this study, BARS emerged as a promising tool for enhancing student reflections. The students appreciated the clarity provided by the detailed descriptors of BARS, as it helped them better understand the components of effective collaboration. This suggests that BARS can serve as an effective framework for guiding students during self-led debriefings. Yet, students also perceived the BARS as too lengthy, which could discourage engagement during debriefings. A possible solution to this issue would be to shorten the BARS by not requiring students to answer all 10 dimensions in every debriefing. Instead, specific dimensions could be selected based on students’ interests, needs, or temporal aspects. For example, early in the collaboration process, the planning activities dimension may be more relevant, whereas other dimensions might take precedence later. This flexible approach could make BARS more manageable while still maintaining its effectiveness as a tool supporting student reflection. Further research is needed to evaluate the impact of using BARS in this modified format, particularly in comparison to reflections conducted without such structured scales.

Despite the usefulness of BARS and instructing one member of each team (i.e., multiplicator), the findings also suggest that students wished for additional support during the reflections. When teams encounter difficulties, they often struggle to articulate criticism constructively and manage differing viewpoints, indicating a need for stronger facilitation. On the other hand, when group work proceeds smoothly, students may find it difficult to appreciate the value of debriefings, perceiving them as unnecessary. These challenges suggest that educators should offer more guidance during debriefings, helping students navigate difficult conversations and reinforcing the benefits of reflection, even when collaboration appears to be successful. Yet, with regard to this, the question arises of how to do so in large classes with limited resources. Additionally, the student-teacher relationship itself may hinder students’ engagement with reflective practices (Chan and Lee, 2021). To address these challenges, it may be beneficial to have a more neutral facilitator support student teams during reflections. This could help mitigate any perceived power dynamics between students and instructors, fostering a more open and productive environment for reflection. A potential solution is to train student assistants in facilitating reflections. Given their close connection to the target group, they may be better equipped to understand students’ experiences and thereby foster a more open, non-hierarchical dialogue. With appropriate institutional support, another potential solution could involve interdisciplinary collaboration. For instance, students from disciplines that focus on team dynamics, such as psychology, could be trained as facilitators for student teams from other disciplines. This approach would support not only reflective practice but also provide valuable real-world experience for facilitators.

Taken together, this study provides deep insights into the role of reflection in fostering students’ collaborative experiences; however, challenging these observations in larger samples would provide an enhancing empirical basis for the findings. Future studies may also explore potential influencing factors, such as psychological safety or feeling heard, to better understand how these elements impact the quality and effectiveness of reflective practices. Gaining a deeper understanding of how reflection fosters collaboration and personal growth is crucial, particularly in preparing students for the demands of the 21st-century workplace. Expanding this research will help refine strategies that can better equip students for collaborative environments, addressing the pervasive gap in students’ preparation for group work across various global contexts (e.g., Le et al., 2018; Wilson et al., 2018; Xu, 2024).