The study was conducted at the refugee camp Kappler Knoten in Freiburg, Germany, involving temporary accommodation and integration programs for displaced families. The workshops were held in a quiet, empty room to minimize external distractions and allow children to fully engage in the activities. The study involved eight weekly workshops over 1 month, held twice a week, conducted during the summer of 2024, each lasting 1 h. The primary participants were six refugee children (3 boys, 3 girls), aged 7 to 12 years. Additional children participated occasionally based on availability and interest. The participants came from Kurdistan, Serbia, Syria, and Ukraine, representing diverse linguistic and cultural backgrounds. Hence, the sampling reflected the transient nature of refugee communities (Woodard, 2024), allowing for a dynamic and flexible group composition. A purposive sampling strategy was applied to maintain a balanced and inclusive group composition, ensuring representation of different experiences among refugee children. The facilitator, familiar with the children’s sociocultural contexts, conducted the research as a participant-observer, designing and leading activities, facilitating engagement, and documenting responses. This approach ensured that the children’s experiences were carefully observed and interpreted, integrating both educational and emotional dimensions (Clark and Moss, 2005). The workshop framework draws conceptual inspiration from White Ant Atelier, an international online atelier founded by M. Mohammadhosseini in response to the COVID-19 pandemic. It integrates Reggio Emilia and Philosophy for Children (P4C) methodologies to foster creative and reflective learning experiences (Table 1). ²

The data collection process employed a Mosaic Approach, a participatory research methodology designed to capture children’s diverse experiences through multimodal techniques. The approach emphasized children’s perspectives and co-construction of knowledge, aligning with the study’s sociocultural and embodied learning framework (Clark, 2005; Ryu, 2020). This method was chosen because of its flexibility in capturing non-verbal communication and children’s agency in the research process (Nicolaou et al., 2023). It allowed for a holistic representation of children’s embodied musical engagement, ensuring that their experiences were rendered through multiple data sources. To analyze non-verbal communication, video recordings were crucial to capture gestures, facial expressions, and movement-based interactions. Importantly, children were actively involved in the setup and dismantling of the cameras, fostering a sense of agency and participation in the research process. This involvement not only helped them become familiar with the documentation tools but also ensured a more natural and engaging environment during recordings. The data collection integrated structured observations, video recordings, verbal reflections, and embodied responses, ensuring a comprehensive and ethically responsible method of documenting children’s experiences (Clark and Moss, 2005). The recordings were complemented by facilitator notes, thematic reflections, and participant observations, which were systematically categorized using thematic and phenomenological analysis techniques (Nicolaou et al., 2023).

The video cameras were placed at fixed angles to ensure minimal disruption while capturing holistic body movements essential to the study’s focus on embodied musical engagement. Children were also given the opportunity to express their experiences through drawings, storytelling, and participatory feedback sessions, reinforcing the inclusive and participatory nature of the study. These structured yet flexible documentation strategies ensured that children’s voices and emotions were authentically represented, contributing to a rich, multidimensional dataset (Clark and Moss, 2005; Nicolaou et al., 2023).

The collected data were analyzed employing an adaptation of thematic analysis (TA), following Braun and Clarke’s (2006) structured framework. TA was chosen because of its flexibility in capturing both explicit and implicit meanings in qualitative data, making it particularly suitable for analyzing embodied experiences and non-verbal interactions (Braun and Clarke, 2023). Unlike other qualitative analysis methods, thematic analysis allows for an iterative process, enabling a balance between researcher-led coding and participant-driven insights (Liebenberg et al., 2020). This was crucial in ensuring that children’s embodied expressions were analyzed in a structured yet interpretative manner, aligning with the study’s focus on social connection and emotional engagement through movement. Using the MAXQDA, the analysis followed six key stages: (1) Familiarization with the data – reviewing video recordings, observational notes, and facilitator diaries to develop an in-depth understanding; (2) Generating initial codes – identifying meaningful features and recurring patterns in the dataset; (3) Identifying key themes – clustering codes into broader themes that represented children’s experiences; (4) Reviewing and refining themes – ensuring that the themes accurately reflected the dataset and aligned with embodied learning frameworks; (5) Defining and naming themes – finalizing thematic categories based on conceptual coherence; and (6) Producing the final report – synthesizing findings to answer the research questions.

To enhance methodological plasticity and in line with the mosaic strategy, a triangulation approach was employed, integrating three complementary data sources as “individual tiles” (1) Video data, capturing children’s embodied engagement, gestures, and non-verbal expressions, providing insights beyond verbal discourse. (2) Field notes, systematically documented after each session in accordance with P4C and Reggio Emilia documentation principles. These notes followed a structured format, encompassing (a) detailed observations of children’s interactions and responses, (b) facilitator reflections on pedagogical approaches, and (c) potential modifications for subsequent sessions. Session transcripts, capturing verbal interactions and the overall progression of discussions and activities to ensure a comprehensive textual representation of the learning environment. This triangulation methodology facilitated cross-validation between data sources, thereby enhancing interpretative depth and analytical reliability (Erickson, 2011). Furthermore, ethnomethodological videographic analysis was applied to examine bodily expressions, facial cues, and movement-based interactions, offering a nuanced understanding of non-verbal modes of communication (Bühler-Niederberger, 2020). Given the multimodal nature of the data, particular attention was devoted to the transcription process. A total of eight video was reviewed three times by researcher 1 to meticulously document both verbal and non-verbal cues, thereby ensuring interpretative thoroughness. All video data were included in the thematic and multimodal analysis, ensuring comprehensive coverage of children’s embodied engagement across the intervention. The video data were analyzed using a multimodal, interpretive framework grounded in sociocultural learning theory [Vygotsky, as discussed in Mahn (2012) and supported by established practices in multimodal transcription and analysis Bezemer and Mavers (2011)]. Drawing on these foundations, the analysis considered gestures, gaze, spatial orientation, and rhythm as meaning-making tools that reflect children’s embodied engagement and social interaction. To ensure analytical depth, video segments were transcribed multimodally, incorporating not only speech but also body movements, gaze direction, and affective expressions, following principles for multimodal transcription (Bezemer and Mavers, 2011; Cowan, 2018). Although no automatic emotion recognition software was used, the interpretation of non-verbal affective states (e.g., expressions of joy, frustration, or collaboration) was inspired by interdisciplinary approaches to multimodal emotion representation (Mocanu et al., 2023).

To enhance the richness of the analytical process, external expert consultation was sought not only for general validation but also for specific aspects of the coding process. Two researchers specializing in Caring Thinking (CT) and Philosophy for Children (P4C) were consulted to examine the alignment between observed behaviors and Lipman’s CT dimensions (Lipman, 2003). Their feedback focused particularly on the interpretation of non-verbal, embodied expressions such as empathic gestures, spatial proximity, and collaborative movement, which were treated as indicators of relational reasoning and emotional attentiveness. These expert insights contributed to refining the operationalization of CT in a multimodal, embodied context. Additionally, preliminary findings were discussed with peers and a co-researcher to ensure analytical transparency and critical scrutiny. The analysis was subjected to three iterative review.

In response to the challenges outlined in Section 3.3, the MSLE design emphasized non-verbal and embodied strategies. Consistent with Nicolaou et al. (2023), selected activities required minimal verbal interaction and supported symbolic expression through voice, gesture, and movement.

Specific multimodal strategies were incorporated, engaging the children through auditory, visual, and kinesthetic methods (Hogle, 2020), such as singing without lyrics, a specific mode of nonverbal group conversations. Concurrently, and in line with existing research (see 2.2), the sessions were designed based on principles of Dalcroze Eurhythmics, the Reggio Emilia approach, and other established methods of community music (e.g., body percussion), empowering the children in symbolic expression (Arseven, 2014; Kenny, 2022a; Kenny, 2022b), empowering children through embodied and symbolic forms of expression (Juntunen and Hyvönen, 2004).

Multimodality and participatory engagement, as core concepts of the MSLE, involved a variety of materials and methods. This aligns with the second and third principles emphasized by Nicolaou et al. (2023), namely, active participation and a non-linear approach (p. 9). Hence, the workshop sessions centered around vocal emotion, body percussion and vocal storytelling. Tasks like developing and presenting nonverbal vocal tales, which focus on illustrating narrative progressions and emotional experiences, aim to enhance children’s skills such as emotional modulation. Each session included three core components:

Thus, the integration of embodiment into the design was grounded in embodied music cognition (Nijs and Nicolaou, 2021), emphasizing the interconnectedness of cognition, emotion, and physical movement through musical and bodily interaction. This approach was operationalized through (1) physical engagement in music, where rhythmic activities helped children internalize musical structures through movement (2) culturally relevant vocal improvisation, which allowed children to express emotions through song (Mabus, 2023a; Mabus, 2023b); (3) and multimodal interaction, combining digital storytelling with physical expression to enhance both cognitive and emotional engagement (Woodard, 2024). These methods supported inclusion and resilience-building, aiming at fostering emotional and social wellbeing through caring thinking among refugee children (Crawford, 2017).

Aligned with the iterative DBR framework (see Section 3.1), the MSLE design was refined through four development phases. The iterative design process unfolded across 4 weeks of bi-weekly workshops, with refinements introduced after each cycle of observation, reflection, and thematic coding. In Week 1, exploratory prototyping revealed challenges with verbal participation due to linguistic diversity; in response, Week 2 sessions integrated more body-based and non-verbal storytelling activities, including gesture-driven vocal play and collective movement. Ethnomethodological observations in Week 2 highlighted social asymmetries during performance planning, prompting the introduction of structured turn-taking protocols and leadership rotation in Week 3 (e.g., Workshop 5, Transcript Section 9). Thematic refinement took place continuously through facilitator reflection and real-time feedback; for example, following a resource conflict in Workshop 3, collaborative decision-making tasks were adjusted to include fairness prompts and more group discussion time (see Transcript Section 12). Finally, in Week 4, appreciative thinking activities were adapted to emphasize collective celebration and recognition, as seen in expanded use of group applause, peer feedback, and reflective dialogues (Workshop 7, Transcript Section 18). These adaptive modifications were grounded in ongoing analysis using MAXQDA and guided by Lipman’s Caring Thinking framework, ensuring that the learning environment evolved responsively in both form and content (Liebenberg et al., 2020).

The research was conducted in accordance with ethical⁴ guidelines. Distinctly, the research must be trauma-sensitive, reflecting on power relations and prioritizing fully informed consent, especially given the vulnerabilities of the population involved. All procedures are designed to minimize the risk of distress or re-traumatization, while participatory approaches and ongoing researcher reflexivity address and attempt to mitigate power imbalances throughout the project. Informed consent is treated as an ongoing, interactive process, with information presented in an accessible manner and participants reminded of their rights at all stages. Overall, the research is grounded in equity-based learning and benefits from a researcher with close proximity to the issues, ensuring ethical engagement and reciprocity remains central to the study design. Comprehensive ethical oversight, including community consultation and institutional review, further safeguards participants’ wellbeing. Prior to the study, a background clearance letter from the municipality was obtained. Additionally, an agreement with the camp authorities was signed, ensuring that names, images, and videos of the children involved in the study would not be shared outside the research context.

Embodied actions were a primary means for children to express emotions, often serving as an outlet before verbal articulation.

Movements such as clapping, swaying, and often mimicking animals showed facilitation of emotional bonding gestures and expression of group cohesion. The theme tackles interactions that enabled the children to process emotions collaboratively (Sangiorgio, 2018). Thus, several incidents in the data were coded, where the children bodily expressed their emotion inspired by a story told during the session.

The theme comprises two observations:

Theme 1 illustrates how embodied interactions facilitated emotional attunement and group cohesion, supporting the development of emotional empathy. Beyond serving as a means of immediate emotional attunement, embodied activities also created a bridge between cognitive learning and emotional exploration. Animated storytelling paired with physical reenactments allowed children to engage non-verbally, facilitating participation despite linguistic barriers. This integration of movement, sound, and collective participation further reinforced emotional connection and empathy-building (Cowan, 2018; Mahn, 2012).

The workshops created a space for children to explore their cultural expression through movement and storytelling, promoting inclusivity and mutual respect.

This exchange highlighted how embodied practices foster cultural dialogue (Cowan, 2018).

The theme comprises two observations:

In this theme, interactions demonstrated the role of physical movement during the workshop and how it may have helped children manage emotions, showing self-regulation strategies.

Rhythmic breathing exercises and group activities were used at the start of sessions, resulting in visible relaxation and increased focus among participants (Maguire and Delahunt, 2017).

The theme comprises two observations:

Figure 2 presents a conceptual graph illustrating the interconnections between body–mind integration, emotional regulation, and the development of caring thinking. The relationships depicted in this network highlight how the data analysis brought to light how physical activities such as rhythmic breathing and synchronized movement contributed to practices that displayed practices of mindfulness, stress regulation, and empathy-building among refugee children. This visualization underscores the role of embodied practices in fostering emotional resilience and cooperative learning in theme 3.

Playful and improvisational activities encouraged creativity, collaboration, and empathy.

Improvised storytelling allowed children to use sound and motion to create collective narratives, resulting in spontaneous laughter and group bonding (Sangiorgio, 2018).

The theme comprises two observations:

Figure 4 presents a conceptual network illustrating the relationships between the identified themes, key elements, their impact on children, and practical examples from the study. Each theme (e.g., Physical Expression, Cultural Identity, Body–Mind Integration) is connected to its corresponding key elements (e.g., Emotional Connection, Cultural Storytelling), demonstrating how the data analysis revealed emergent traces of caring thinking especially in moments of playfulness and improvisatory elements. The visual representation highlights the interplay between these factors, emphasizing their collective role in promoting emotional resilience, social interaction, and cognitive development among refugee children.

Obviously, the workshops also offered opportunities to foster trust and belonging.

The safe space apparently allowed children to open up and form meaningful connections across cultural and linguistic divides.

This theme comprises two distinct observations, illustrating different ways in which a safe space fosters social connection. By enabling children to engage in cooperative play and develop trust in their peers, these interactions create a foundation for caring thinking, where empathy and mutual support become integral to their social experience. The first, Reduction in Isolation, reflects how previously withdrawn children gradually transitioned into active participation, underscoring the role of a supportive environment in encouraging social integration. The second, Group Cohesion, captures how collaborative activities, such as rhythmic games, gradually facilitated collective engagement and paved the way to establishing interpersonal bonds.

The theme comprises two observations:

While Theme 3 addresses the immediate role of movement in emotional self-regulation, Theme 6 explores the long-term impact of embodied learning in fostering cognitive and emotional resilience, emphasizing how children independently apply these strategies beyond the workshop environment. The workshops´ features apparently equipped children with tools to regulate emotions and navigate stress, thus potentially fostering cognitive and emotional resilience.

This example illustrates how embodied practices can effectively equip children with emotional regulation tools, supporting long-term stress management and emotional resilience. The theme comprises two distinct observations, each highlighting a different dimension of how embodied learning fosters cognitive and emotional resilience:

Figure 5 presents a conceptual network illustrating the relationship between emotional regulation and cognitive resilience in the context of embodied learning. The graph highlights how the data analysis indicated that mindfulness practices, such as rhythmic breathing and group synchronization, contribute to stress reduction, which in turn fosters cognitive resilience. Additionally, collaborative learning and increased focus play a significant role in developing empathy among refugee children. This interconnected framework underscores the long-term benefits of embodied learning strategies in supporting emotional wellbeing and adaptive coping mechanisms.