From Collaboration to Integration: How a Community of Practice Supports Public School Teachers’ Understanding of Integrated STEAM Education

Abstract

Integrated science, technology, engineering, arts, and mathematics (i-STEAM) education has been recognized for its potential to promote interdisciplinary learning and connect scientific knowledge to socially relevant contexts. However, its implementation in school practices remains limited, often owing to conceptual ambiguities and a lack of sustained support for teachers. This study examines the role of participation in a community of practice (CoP) in facilitating the adoption of i-STEAM principles by public school teachers through collaborative lesson planning. Drawing on meeting transcripts, documents produced during the process, and interviews with participants, the analysis focused on the constitution of the CoP, the presence of i-STEAM elements in the teaching proposals, and the level of integration achieved. The results suggest that the CoP supported the development of more coherent and context-sensitive understandings of i-STEAM, while also promoting interdisciplinary design across diverse educational levels. These findings may also inform initiatives in other public education systems facing similar structural conditions, such as limited resources, disciplinary fragmentation, and restricted opportunities for collaborative curriculum development. The study highlights the value of CoPs as professional learning strategies and proposes a replicable analytical approach for evaluating how teachers engage with integrative pedagogies. Implications for teacher education and policy are discussed.

1. Introduction

Science, technology, engineering, arts, and mathematics (STEAM) education evolved from the STEM movement, created over three decades ago in the United States, in response to the growing demand for qualified professionals to meet the requirements of an increasingly technological job market. Initially, STEM focused on developing technical and scientific competencies; however, over time, this initiative evolved into an educational integration approach that bridges disciplines. One discussion of this model is that the delimitation of STEM disciplines by educational systems as priority policies can reinforce the hierarchization of these areas to the detriment of others and consequently shape the types of knowledge valued by society (Mejias et al., 2021).

Integrated STEAM education (i-STEAM) has emerged in this context. The inclusion of arts and humanities in STEM areas has been advocated for its potential to stimulate creativity and critical thinking, encourage positive attitudes towards science and mathematics, and insert science teaching into sociocultural environments, contributing to the formation of citizens’ commitment to social and environmental issues (Mejias et al., 2021). This holistic perspective aligns with Yakman’s (2008) definition of STEAM education as a dynamic system in which its constituent disciplines interact as mutually reinforcing domains, enabling learners to connect knowledge with personal, societal, and global challenges.

Unfortunately, given the expansion of the term’s use, a problem observed by different researchers is that many proposals identified as STEM/STEAM maintain traditional teaching practices, failing to provide genuine interdisciplinary integration. Some view the acronym as a trendy term used to attract attention and funding, without resulting in concrete transformation in the teaching-learning process (Akerson et al., 2018; García-Carmona, 2020; García-Carmona & Toma, 2024).

Although the term is widely known, and the implementation of these initiatives has recently increased, the problem is that there is no consensus on the criteria a didactic proposal must meet to be considered i-STEAM (Aguilera & Ortiz-Revilla, 2021). It is also unclear how practices following this approach should be planned and evaluated, which is reflected in inconsistent interpretations and difficulties in implementation by teachers in classrooms (Aguilera et al., 2022). Thus, providing teachers with i-STEAM-related learning experiences, accompanied by continuous theoretical and practical support, can increase their interest and pleasure (Boice et al., 2021).

This perspective correlates with the idea of a Community of Practice (CoP), a fundamental concept for understanding the social learning process, which can be defined as a set of relationships between people, activities, and the world over time, being an intrinsic condition for the existence of knowledge, whose social structure, power relations, and legitimacy conditions define the possibilities of learning (Lave & Wenger, 1991).

Considering that research in science education points to the potential of CoPs to support teaching practices through sharing knowledge, experiences, and resources that favor the implementation and maintenance of innovations, something that traditional improvement courses, which are disconnected from concrete practices, cannot achieve (Ortiz-Revilla et al., 2022; Weinberg et al., 2021), this article aims to analyze the impact of participation in a CoP on teachers’ understanding of i-STEAM.

To this end, the following research questions were sought: How are the CoP dimensions established in teacher interactions? What i-STEAM elements can be evidenced in the planning of didactic proposals? What level of integration is evidenced in the i-STEAM lesson plans developed by the teachers? The first question seeks to identify the elements that characterize the group as a community of practice. The second explores how teachers incorporate and adapt i-STEAM principles in their planning. The third aims to understand the level of integration present in the didactic proposals developed by participants.

In this study, “impact” refers specifically to teachers’ planned pedagogical practices, as documented in lesson plans and discussed in interviews, not to their enacted classroom teaching. We acknowledge the distinction between planned, enacted, and experienced curricula, as discussed by Matthews et al. (2016). However, planned practices may not enable us to understand how teachers articulate their pedagogical intentions and how their conceptions evolve within the CoP. From this perspective, planning becomes a relevant space for examining how teachers interpret and incorporate i-STEAM principles (Kim & Bolger, 2017).

2. Theoretical Framework

2.1. Integrated STEAM Education

Investigations into i-STEAM are an emerging field that seeks to integrate arts and humanities into STEM, with the potential to broaden epistemic perspectives and foster more connected learning. However, this approach remains under-theorized, with ambiguous applications and diverse implementation objectives (Mejias et al., 2021).

The growing expansion and popularization of this conception has generated multiple pedagogical models and interpretations, complicating the definition of disciplinary roles and directly affecting teaching and assessment practices (Mejias et al., 2021). Beyond theoretical models, the definition of “A” in the acronym also lacks consensus. According to Perignat and Katz-Buonincontro (2019), research alternates between arts as artistic education, as project/problem-based learning and technology, or as any non-STEM discipline, including humanities. Mejias et al. (2021) further propose including social sciences in dialogue with natural sciences, emphasizing their interconnected practices and mutual enrichment.

In this study, we adopt the perspective proposed by Ortiz-Revilla et al. (2022), who define i-STEAM as an integrated approach that articulates STEM with arts/humanities to address socially relevant problems, humanize knowledge, and promote equitable participation. We also acknowledge the tensions and ambiguities surrounding the term, assuming that arts and STEM are mutually instrumental rather than hierarchically arranged.

For Mejias et al. (2021), even in a normative context that values economic competitiveness, i-STEAM can advance equity by incorporating everyday, family, and community knowledge into science and engineering learning. This aligns with efforts to value cultural knowledge systems, such as traditional narratives, in disciplinary education.

2.2. Challenges in Understanding and Implementing i-STEAM Education

The implementation of i-STEAM education faces challenges, primarily due to teachers’ difficulties in conceptualizing and applying it in practice (Chu et al., 2019; Kim & Bolger, 2017). The multiplicity of conceptions and limited teacher training create a gap between the potential of i-STEAM and educators’ confidence in adopting it (Boice et al., 2021; Stevenson et al., 2025).

Lack of confidence is a significant factor in the success of i-STEAM implementation, and teachers who feel insecure about the content or methodologies tend to avoid activities related to this approach (Kim & Bolger, 2017; Wu, 2022). Among the difficulties faced by teachers, the following stand out: inadequate problem formulation, where many opt for simplified situations that do not stimulate critical thinking; difficulties in accurately identifying and connecting disciplinary content to the respective STEAM areas, often resulting in superficial associations that lack epistemological coherence; underutilization of knowledge from some disciplines, especially mathematics; and an isolated or multidisciplinary approach to content, which prevents effective integration between disciplines (Aguilera et al., 2022).

Such teacher-related challenges have direct consequences for students. When teachers lack confidence or clear strategies for integrative instruction, students encounter fragmented learning and fewer opportunities for authentic inquiry, resulting in a limited interdisciplinary understanding and problem-solving (Dare et al., 2019; Kim & Bolger, 2017). Recent studies confirm that limited professional development and low teacher familiarity with STEAM reduce the depth and coherence of implementation and, consequently, student engagement and learning quality (Correia et al., 2024; Fields & Kafai, 2023; Kessler et al., 2024; Martins & Baptista, 2024; Silva-Hormazábal & Alsina, 2023).

The challenges in promoting disciplinary integration are related mainly to the difficulty of visualizing how to integrate disciplines without one area prevailing over others or fragmenting knowledge. This process requires a more in-depth understanding of the relationships between disciplines and methodologies that facilitate this integration (Aguilera et al., 2022; Stevenson et al., 2025).

Despite the importance of disciplinary integration for student development, the propagation of knowledge among teachers has not received sufficient attention. Collaboration between educators often lacks time and experience in collaborative activities (Wu, 2022). Given that, even after participating in training courses, some difficulties persist in integrating and assessing specific disciplines (Ortiz-Revilla et al., 2023); it is essential to reassess these strategies and consider alternative practices to those commonly adopted by current educational systems, which can help bridge these gaps.

2.3. Learning in Communities of Practice

The development and implementation of i-STEAM lesson plans (here referred to as didactic proposals collaboratively designed by teachers) can be driven by collaborative work that allows for determining the alignment with i-STEAM and the CoP of professionals who share common objectives and interests (Weinberg et al., 2021). These communities promote joint knowledge construction based on their own needs and contexts, thereby rescuing teacher protagonism, with teachers acting as co-authors of innovation rather than mere recipients of externally formulated curricular directives. This perspective aligns with international findings that highlight the importance of collaborative structures in fostering integrated pedagogical practices and supporting the complexity of curriculum development in teacher education programs (Chu et al., 2019; Stevenson et al., 2025). In a CoP, learning opportunities arise from the movement of knowledge and practices among them (García et al., 2008).

From this understanding, learning the i-STEAM approach can be favored when developed in a CoP of teachers with distinct backgrounds and experiences (considering learning as an interaction process involving the mobilization of concepts, methodologies, and procedures from different disciplines; understanding the connections established between them; and mutual understanding of practice) (El-Hani & Greca, 2013).

CoPs are supported by social learning theory, which posits that knowledge is primarily acquired through active participation in these communities. Within this theory, three dimensions connect community and practice: mutual engagement, which involves the sustained interaction of people within the community and the roles and relationships that emerge from this interaction; joint enterprise, which refers to how members negotiate their responses to the community’s conditions and goals; and a shared repertoire, which consists of signs, symbols, tools, and languages used as resources that have specific meanings for the community (Wenger, 1998).

For the i-STEAM approach, this perspective emphasizes the importance of involving teachers in meaningful practices, offering resources that increase their participation, and promoting actions, discussions, and reflections that help them chart relevant learning trajectories (Weinberg et al., 2021). In mutual engagement, teachers develop reciprocal relationships, define roles, and collaborate with knowledge and practices from different areas. Joint enterprise shapes and is shaped by teachers’ participation, which can be represented by negotiation about what should be done, what is relevant, and when resources and practices need to be improved. Finally, a shared repertoire refers to the material and social resources that members develop or incorporate through participation, generating new ways of addressing recurring problems (Wenger, 1998).

Thus, the choice to work in a CoP derives from the knowledge that discontinuous teacher training meetings are insufficient to promote real integration because, in such cases, even when professionals from different fields can collaborate on projects, their interaction tends to be limited to sporadic exchanges of ideas or knowledge (Wu, 2022). Therefore, teachers should allow the theory to be implemented, sharing what works and what does not, thereby refining their approaches over time. This process of constant exchange is essential for teachers to adopt i-STEAM, adapting it to their specific realities effectively. In this direction, it is possible to observe that work in a CoP has the potential to ground the analysis of how learning occurs in a collaborative context and provide a favorable environment for sharing experiences and collective knowledge construction, which are essential aspects for developing i-STEAM.

3. Materials and Methods

This study uses a qualitative, single-case embedded design (Yin, 2014) to explore teachers’ collaborative learning in i-STEAM and understand the phenomenon within its real-life context, where the boundaries between the two are unclear.

3.1. Context and Participants

This study examines the CoP “Vozes Compartilhadas” (shared voices), formed in August 2023 in the context of an open innovation event—the “Mulheridades” Ideathon, promoted by Armazém da Criatividade—Porto Digital Caruaru—to identify and develop solutions for challenges faced by women in Pernambuco. The group was created to collectively develop pedagogical practices that promote positive attitudes towards science—specifically among their female students—seeking to inspire girls to engage with and envision themselves in scientific fields. This work was conducted through a combination of face-to-face and remote meetings for reflection, study, planning, and the sharing of best practices.

The community comprises twenty-five teachers working in public schools across ten municipalities in Pernambuco, northeast Brazil. These municipalities reflect diverse realities, including rural and urban areas, both in the state capital and the interior, traditional communities (such as Indigenous, Afro-Brazilian maroon communities, and Gypsy communities), peripheral and vulnerable populations, as well as communities with more favorable economic and social conditions.

Most participants had over ten years of experience, with their first exposure to i-STEAM occurring through the CoP. For this research, only the sixteen teachers who participated in all stages of the study were included.

Table 1. Participant Profiles.

Participants	Academic Background	Educational Level
Ada	Pedagogy	Early Childhood Education
Adriana	Biology	High School
Bertha	Biology	High School
Frida	Art	High School
Hedy	Biology	High School
Jaqueline	Biology	High School
Katherine	Biology	Adult Education (EJA)
Katemari	Mathematics	Middle School
Maria Laura	Mathematics	EJA
Marie	Chemistry	High School
Marta	Biology	High School
Mary	Mathematics	High School
Mileva	Physics	High School
Rosalind	Biology	High School
Sonia	Biology	High School
Valerie	Physics	High School

provides details on their academic background and educational level. The teachers chose pseudonyms based on historical female scientists for the study.

Although most participants hold STEM degrees, they are also certified and trained to teach general science in grades 6–9 (students aged approximately 11–14 years). Teachers maintained ongoing communication through a WhatsApp group (which also included the researcher) for organizational purposes, while an Instagram page documented training sessions. These platforms were used exclusively by teachers, with no student participation. All participating teachers provided informed consent for data collection, which included access to their WhatsApp interactions and the use of anonymized excerpts for research purposes. No student data were published, and both platforms were used in compliance with the approved ethical protocol (Code 7.453.253/2025).

Since participants often come from different cities, meetings are held both in person and virtually. After meetings, collaborative work continued in smaller, locally or remotely organized groups. The first meeting introduced the i-STEAM approach and its key methodologies. In the second part, teachers shared their initial school-based experiences of incorporating i-STEAM elements. The third meeting included an in-depth discussion and collaborative development of adaptable i-STEAM lesson plans in small groups. Afterwards, the teachers refined their proposals individually or in pairs, with all the versions recorded in a shared document.

3.2. Procedures and Data Collection

Data was collected from three sources: meeting session transcripts, interview transcripts, and planning documentation records. CoP meetings were recorded and later transcribed. The didactic proposals collaboratively developed by the teachers were also analyzed. These documents were completed using a standardized digital template and submitted via a shared Google Drive folder. Additionally, a semi-structured interview adapted from Ortiz-Revilla et al. (2023) was conducted to analyze the distinct levels of STEAM integration achieved and the perceptions and attitudes that influence teachers’ didactic proposal planning. The interview, recorded and transcribed for later analysis, consisted of ten questions organized into three dimensions, allowing evaluation of teachers’ cognitive, affective, and perceived control attitudes. The first, second, and third dimensions involved opinions on the importance and implementation difficulties of i-STEAM, as well as the importance of i-STEAM and the difficulties associated with its implementation. Additionally, they assessed anxiety and pleasure levels when implementing the approach, and self-perception of implementation capacity.

3.3. Data Analysis

We adapted the analysis tool for local science teaching practice communities proposed by Valois and Sasseron (2021) to understand the constitution of the CoP and how its dimensions are established. The tool encompasses the three dimensions of a CoP (Wenger, 1998), namely, Joint Enterprise, Mutual Engagement, and Shared Repertoire, and how these aspects relate to scientific knowledge domains, including conceptual, epistemic, social, and material (Stroupe, 2014) (

Table 2. Analysis Tool for the CoP Constitution ¹.

Question of Analysis	CoP Dimensions	Scientific CoP Domains	Description
What are the purposes of the meetings that teachers express?	Joint Enterprise	Social and Material	Identifies the purposes of the meetings and verifies the collective negotiation of objectives.
How do the teachers engage with the activity and the group, and what knowledge do they express, share, and recognize when planning and conducting activities?	Mutual Engagement	Social, Epistemic, Conceptual and Material	Identifies mutual help, tensions, shared knowledge, and public recognition of teachers’ participation in planning and developing activities.
What resources are considered by the teachers during the development of activities?	Shared Repertoire	Material and Social	It examines the use of resources and tools adapted by the community.

¹ Adapted from Valois and Sasseron (2021).

).

The framework of scientific communities of practice proposed by Stroupe (2014) enables a more nuanced investigation of how distinct types of knowledge and practices emerge and circulate within the group, particularly in relation to interdisciplinary integration. In applying Stroupe’s framework, we treated the social domain not as a general indicator of any interaction but rather as a specific marker of how relationships, group dynamics, and shared norms shaped collaboration. While social elements are inherent in any CoP activity, we coded data only under the social domain when the content explicitly addressed group coordination, mutual support, interpersonal recognition, or decision-making processes. For example, statements acknowledging a colleague’s contribution, negotiating leadership roles, or coordinating collective tasks were classified as social. In contrast, other types of statements are allocated to the conceptual or epistemic domains. The analysis of meeting transcripts was conducted independently by two of the study’s authors. After individual coding, the results were compared, and a consensus was reached through discussion and deliberation.

The evaluation of the lesson plans developed by the participating teachers was conducted via an analytical rubric composed of nineteen indicators adapted from Ortiz-Revilla et al. (2024) and Aguilera et al. (2022). These indicators are organized into three main dimensions: theoretical alignment, practical quality, and feasibility. The first dimension, theoretical alignment (indicators 1–3), assesses the educational purpose of the proposal, its alignment with the goals of i-STEAM, and the relevance and complexity of the problem being addressed. The second dimension, practical quality (indicators 4 to 18), examines how the proposal is structured and implemented, including aspects such as disciplinary integration, inquiry, modelling, engineering design, assessment strategies, collaboration, and social impact. The third dimension, feasibility (indicator 19), considers the available resources.

Each indicator is rated on a scale from 0 (emerging) to 3 (sophisticated), with specific descriptors guiding the scoring. The total score determines the overall level of practical integration, categorized as follows: 0 to 15 points indicates that the proposal is “in the process of acquisition”, 16 to 31 points is considered “basic”, 32 to 47 points is “advanced”, and 48 points represents a “sophisticated” level. In parallel, the theoretical dimension allows classification of the proposal as either “Pseudo-i-STEAM” (score between 0 and 4), indicating superficial or partial integration, or “i-STEAM” (score between 5 and 9), reflecting intentional and pedagogically coherent integration of the relevant domains. The feasibility dimension is assessed separately, using a scale from 0 (not feasible) to 3 (entirely feasible with basic, widely available materials).

The rubric allows the alignment with i-STEAM to be determined and the quality of the proposals to be evaluated. The first two authors of this article conducted the analysis independently, and the final categorization was obtained through discussion until a consensus was reached among all the authors.

Table 3. Planning evaluation system ².

Dimension	Values	Description	Level
What for	0–4	Pseudo-iSTEAM Proposal	Theoretical
	5–9	i-STEAM Proposal
What and how	0–15	In the process of acquisition	Practical
	16–31	Basic
	32–47	Advanced
	48	Sophisticated
Feasibility	0	Not feasible	Application
	1	Potentially feasible
	2	Partially feasible
	3	Totally feasible

² Adapted from Ortiz-Revilla et al. (2024).

presents the evaluation system.

To analyze participants’ conceptions of disciplinary integration, we adapted the integration complexity staircase proposed by Gresnigt et al. (2014), which organizes curriculum integration into six levels: fragmented, connected, nested, multidisciplinary, interdisciplinary, and transdisciplinary. Our adaptation translated these levels into the discursive indicators observed in the interviews. The fragmented level was characterized by conceptions that treat disciplines as isolated units; the connected level involved references to explicit links between subjects, usually established by the teacher. The nested level refers to instances where learning goals from one discipline are embedded in the context of another. The multidisciplinary level was identified when teachers organized content around a common theme or project, maintaining disciplinary boundaries. The interdisciplinary level involves resolving real-world problems that require integrating concepts and methods across disciplines. Finally, the transdisciplinary level was associated with problem-solving approaches that extended beyond school boundaries, including references to the involvement of external agents or community members. Two researchers independently developed and applied a coding guide based on this framework.

3.4. Data Triangulation

We employed a data triangulation approach that enabled us to connect the three sources of evidence (meeting transcripts, interviews, and lesson plans) to ensure the coherence of our findings. These materials were analyzed in an integrated manner to identify convergences, reinforce interpretations, and explore inconsistencies.

For example, we examined how teachers’ collaborative dialogue in planning sessions, including role assignment and tool sharing, mirrored interview patterns such as trust and peer learning. Additionally, we verified whether increased confidence in interdisciplinary work, reported in interviews, was reflected in their lesson plans through enhanced integration and comprehensible pedagogical goals.

This analytical approach enabled us to examine how the CoP supported shifts in i-STEAM planning and understanding. Divergences across data sources were not treated as contradictions but rather as indicators of cognitive tension and evidence of teachers’ ongoing learning processes.

4. Results and Discussion

4.1. How Are the CoP Dimensions Established in Teachers’ Interactions During Meetings?

The first dimension observed was the joint enterprise, expressed through the collective goal of fostering girls’ interest in science via i-STEAM. As Mary emphasized:

Mary: “The integrated STEAM approach, which is what we are going to try to use in the classroom for the girls in science […] I think this STEAM approach instills not only in the girls but also in the students a way to see everyday life in a different light, which is significant. After all, it inspires not only the girls but also the students to view everyday life differently, instilling in them a new perspective on life, which is significant. After all, it instigates not only the girls but also the students to see everyday life differently. It has a significant role in the classroom. After all, it instigates not only the girls but also the students to see everyday life differently.”

Teachers negotiated and adapted proposals collaboratively, regardless of immediate applicability. For instance, Ada, an early childhood teacher, contributed to a high school-oriented plan on cultural biodiversity by suggesting a local objective “to identify characteristics of cultural biodiversity within the community” and proposed producing a short video for Instagram to share students’ perspectives.

Ada: “[…] After class, we would make a video with the girls, them talking about what they already knew about this cultural biodiversity, you know? […] It would be a short video, to put on Instagram, to share among them and in the neighborhood, in the city where they live.”

Thus, the interaction between social and material aspects illustrates how the context in which activities occur influences the planning and execution of pedagogical proposals, requiring educators’ constant reflection and adjustment.

In the CoP, mutual engagement was expressed through collective work, collaborative relationships among teachers, and the challenges faced during planning and socialization. The social domain emerged when tasks were divided according to each participant’s skills, such as slide organization, group coordination, and presentation during socialization moments. The moments of planning in small groups evidence this fact. In group 3, Teacher Valerie highlighted, during socialization, how the collaborative dynamic was essential for the proposal to be developed: “[…] I congratulate Bertha, who typed and wrote with very nice words, [and] Marie, who was narrating […]”.

The epistemic domain was evident in specific knowledge contributions:

Jaqueline: “One of the things I was seeing and considering for the practices, and I think that greatly facilitates, and would like to suggest to my colleagues, is working with themes. For example, if I work on the theme of water or artificial intelligence, I can work on STEAM much more easily. This is because I had an experience at the museum. When we work with a theme, we inevitably address various areas. If we choose the themes, it becomes much easier to involve other teachers.”

The material domain appeared to use virtual environment tools, share Google documents for planning, and suggest resources for practical activities, such as Scratch, which Teacher Mileva indicated for programming activities. For instance, the mention of Scratch in the material domain goes beyond signaling a tool; it reflects an appropriation of technological resources aligned with computational thinking, one of the key components in i-STEAM. Its inclusion suggests that teachers are beginning to move beyond content-based integration and incorporate cross-cutting competencies, revealing progress toward authentic integration.

The conceptual domain emerged articulating through the articulation of area-specific theories and methods, such as Teacher Frida’s use of RPG character creation to explore illustration. This activity reflected the mobilization of artistic ways of thinking and forms of knowledge, such as narrative structure and symbolism, integrated into the problem-solving structure of the lesson plan, indicating the group’s ability to move beyond the aesthetic use of the arts towards deeper conceptual integration.

The shared repertoire refers to tools, documents, routines, and behaviors developed through member interactions, which serve as the primary outcome of CoP participation. The participants exchanged teaching materials and shared unplugged activities unfamiliar to many teachers. For example, Mileva shared her experience using Octostudio. Katemari shared her experience with field lessons at SERTA (Alternative Technology Service), where students learn how to develop gardens using sustainable practices.

The indicators of mutual engagement, responsibility for a joint enterprise, and a shared repertoire of resources—dimensions that distinguish a CoP from a project team (Wenger, 1998)—were all present among the group of teachers studied.

4.2. After the Meetings, What i-STEAM Elements Can Be Evidenced in Planning Didactic Proposals?

The evaluation rubric was applied to eight teaching proposals developed by the teachers. Three were created in groups of four during the meetings; others were not. In contrast, the others were later adapted to different contexts: two were developed by pairs working at the same school, and three were developed individually.

Table 4. Planning evaluation.

	Participants	Problem-Situation	Classification
1	Katemari, Mary, Mileva, and Rosalind	How can we promote sustainability in the school environment to minimise environmental issues?	Pseudo-i-STEAM
			In the process of acquisition
			Totally feasible
2	Ada, Frida, Hedy, and Katherine	How can cultural biodiversity contribute to building a more just and sustainable society?	Pseudo-i-STEAM
			In the process of acquisition
			Totally feasible
3	Adriana, Bertha, Marie, and Valerie	How do screens affect vision and cognitive development.	Pseudo-i-STEAM
			In the process of acquisition
			Totally feasible
4	Katherine and Maria Laura	How can we reduce the impacts of school absences caused by work-related issues?	i-STEAM
			Basic
			Totally feasible
5	Frida and Katemari	How can art be used to incorporate sustainability into our daily lives?	Pseudo-i-STEAM
			In the process of acquisition
			Totally feasible
6	Mileva	How can Pernambuco’s wind farms be adapted to reduce the impacts on the hearing and mental health of nearby residents?	Pseudo-i-STEAM
			In the process of acquisition
			Totally feasible
7	Ada	How can we be healthier?	i-STEAM
			Basic
			Totally feasible
8	Hedy	Where do expired or unused medicines go?	i-STEAM
			Basic
			Totally feasible

presents an overview of the group compositions, the issues addressed in the planning, and the evaluation results.

Of the eight proposals analyzed, only three simultaneously reached the theoretical i-STEAM and the basic practical level, the highest achieved in this study, and all were developed after the CoP meetings. This outcome reflects the group’s initial experience with i-STEAM planning and the limited time for written documentation during the sessions, which prioritized oral discussion. Nevertheless, proposals are shared among groups at the end of meetings within the school context. All proposals were also considered viable for implementation, highlighting a collective effort to adapt to the material conditions of the public schools where participants work. Figure 1 presents the development levels observed in the proposals based on the criteria of the analytical rubric used.

We present a set of illustrative cases organized by key i-STEAM indicators to clarify how the rubric levels were applied. These examples highlight the operational distinctions between levels based on actual lesson plan content.

• Problem or challenge of the proposal
  Level 0—In the process of acquisition: In Proposal 3, the question “How do screens affect vision and cognitive development?” is posed as a conceptual inquiry. However, it is structured and could be addressed through a theoretical explanation rather than an investigable problem aligned with the principles of project-based learning. Furthermore, it does not indicate that students are expected to explore the issue through inquiry, propose solutions, or take action.
  Level 1—Basic: Proposal 1 introduces the problem, “How can we work on sustainability in the school environment to minimize environmental problems?” This is a socially relevant and thematically engaging issue. Nevertheless, the problem remains broad and generalized, lacking situational context within the school or community. The proposed activities are short-term. They primarily focus on awareness, offering limited opportunities for in-depth inquiry or transdisciplinary exploration.
  Level 2—Advanced: In Proposal 7, developed for early childhood education, the question “How can we make our eating habits healthier?” is developmentally appropriate, context-specific, and framed in a way that supports sustained engagement. The challenge is explored through various reflective and hands-on activities, such as food classification, dramatisation, and garden creation. This promotes interdisciplinary learning across multiple STEAM domains over an extended period, which aligns with Level 2 expectations.
• Disciplinary Integration
  Level 2—Advanced: Proposal 8 demonstrates a coherent integration of science (environmental and health education), mathematics (data collection and statistical analysis), and visual arts (informational poster design). Students explore the real-world issue of improper medication disposal, analyze their findings, and design awareness materials to address this issue. The disciplines are connected through a shared pedagogical aim, promoting purposeful interdisciplinary work. Proposal 2 similarly attempts to integrate science, cultural studies, and the arts by proposing engagement with local communities and traditional knowledge. While the emphasis on cultural biodiversity and community dialogue reflects a strong commitment to contextualized learning, these elements were not fully embedded in the instructional methodology or assessment processes. As such, the proposal was not rated at the highest level of integration.
• Engineering Design
  Level 0—In the process of acquisition: In Proposal 5, students construct a “parreiral” system that combines fish farming and vegetable cultivation via recyclable materials. However, the activity is presented as a predetermined product rather than the result of a structured design process. No stages of problem definition, testing, or redesign are articulated. Similarly, in Proposal 4, students are expected to develop a communication tool. While this suggests a design intention, the process lacks the defining features of engineering design, such as iterative development, prototyping, and testing. In both cases, the absence of a systematic engineering approach justifies their classification at level 0.

Four indicators were at Level 0 across all the analyzed lesson plans: engineering design, process evaluation, result evaluation, and regulation of cooperative work. For example, in the indicator “Engineering Design”, a score of 0 is assigned when there is no mention of engineering processes; a score of 1 corresponds to the reproduction of teacher-defined steps; a score of 2 reflects student-driven testing without an autonomous definition of the problem; and a score of 3 applies when students independently define a problem and engage in a complete cycle of designing, prototyping, testing, and iterating.

With respect to engineering design, two plans did not reference the process, and three included activities with potential for development via methodology. However, they were not guided by this framework, and three demonstrated intentionality in implementing the approach, but neither followed its guidelines nor described its stages. These findings align with those of García-Carmona and Toma (2024), who assert that teachers exhibit low levels of pedagogical preparation and self-efficacy in incorporating engineering design into science lessons and often feel insecure about integrating such practices into their teaching (Radloff & Capobianco, 2021).

Despite discussions about the importance of teamwork, no strategies for regulating students’ cooperative work have been identified. Similarly, evaluation was not addressed during the sessions, which may explain its absence in the lesson plans. According to Ortiz-Revilla et al. (2024), although teachers recognize the importance of evaluation, many lack experience with process assessments that require more qualitative methods. Consequently, this limitation directly impacts the practice, leading to its underdevelopment in pedagogical activities.

Strengthening teacher preparation in engineering design requires structured opportunities to work through the entire design cycle—problem scoping, prototyping, testing, and iterative refinement—within real classroom settings. Research shows that such experiences build both pedagogical content knowledge and confidence in guiding students through open-ended engineering challenges (Capobianco et al., 2022; Radloff & Capobianco, 2021; Webb & LoFaro, 2020). Effective programs also integrate engineering concepts into disciplinary coursework—for example, chemical engineering in chemistry or environmental engineering in earth sciences—to help transfer design practices to school environments (Pleasants et al., 2019, 2020). Combining explicit instruction on engineering design processes with mentored field practice enables teachers to integrate these skills into science teaching that responds to specific contexts, creating coherent, inquiry-based lessons aligned with i-STEAM goals (Ryu et al., 2019).

A structured approach to teacher preparation in assessment is essential for i-STEAM implementation. Harris et al. (2023) emphasize that meaningful evaluation should accompany the entire learning process and document how students explore, test, and revise their ideas. Teacher education programs can address this need by engaging participants in the design and trial of assessment tools that capture intermediate stages of inquiry and design work. Key elements include the development of analytic rubrics aligned with scientific and engineering practices, the analysis of student artifacts such as notebooks and prototypes, and the collaborative discussion of evidence in professional workshops. Iterative cycles of planning, classroom implementation, and peer feedback strengthen teachers’ ability to integrate assessment with instruction and to use collected evidence to refine both teaching strategies and students’ learning trajectories.

While teacher preparation provides the foundation for integrating assessment and instruction, the practical enactment of these principles requires frameworks that support teachers in capturing and interpreting evidence of student learning during implementation. Building on Harris et al. (2023), formative assessment can be understood as an ongoing process that connects learning goals, instructional practices, and evidence of students’ reasoning. Teachers can apply this perspective by designing assessment tasks that capture how students engage in inquiry, problem-solving, and reflection throughout their projects. The authors highlight three key components that can guide these practices: learning models that clarify expected progressions of understanding, strategies for eliciting valid evidence of student thinking through observation and questioning, and tools for interpreting this evidence, such as analytic rubrics or learning journals. Applying these principles in the classroom enables teachers to identify intermediate stages of learning, discuss evidence collectively, and use these insights to refine instruction.

A concrete example that illustrates this approach is the Co-Measure rubric (Herro et al., 2017), developed to assess individual collaboration in STEAM activities. This assessment tool defines specific dimensions of collaborative work, including peer interaction, communication, inquiry depth, authenticity, and transdisciplinary thinking. When adapted within a CoP, the instrument can serve as a shared framework for teachers to observe and discuss how students collaborate and reason through problems, embedding assessment within instruction and fostering collective pedagogical reflection.

Building on this perspective, Activity Theory provides an analytical framework that can deepen the understanding of formative assessment by situating it within the broader system of teaching and learning. This theoretical lens conceptualizes learning as a socially mediated process embedded in activity systems and highlights the dialectical relationships among motives, actions, and operations. It enables evaluators to examine how pedagogical intentions are transformed into collective practices and how participants appropriate mediating tools in the process. Previous studies have shown that elements such as subject, object, tools, rules, division of labor, and outcomes can structure teachers’ understanding of their professional development (García et al., 2008). Drawing on this approach, future STEAM-oriented CoPs could design assessment protocols that integrate criteria related to collaboration, tool appropriation, and negotiation of meaning, expanding the understanding of formative processes without reducing assessment to individual results.

The indicators with the best evaluation in the lesson plans were disciplinary integration, authenticity, and feasibility/usability. The groups’ composition facilitated a transdisciplinary approach based on problem-solving. Additionally, all the proposals addressed real-world problems, enabling contextualized practices attainable with accessible and low-cost materials.

The contribution of the arts was reinforced by the fact that the group with representatives from these fields was the only one to achieve a sophisticated social impact. Proposal 3 was directed toward traditional communities and had the potential to generate long-term impacts. On the other hand, the two proposals that achieved a sophisticated level of investigation were developed by biology and mathematics teachers, suggesting that science and mathematics tend to emphasize methodological rigor. This suggests that the unique characteristics of each discipline influence the depth and focus of a project, reflecting their distinct contributions to the complexity and impact of a proposal.

Despite the strong disciplinary integration evident in the lesson plans, the methodologies lacked clear strategies for operationalizing this integration. Although most proposed activities were concrete, a consistent connection was absent. Another challenge identified was the written articulation of the proposal. While the oral discussions revealed a more apparent relationship between theoretical and practical dimensions and an effort toward effective integration, this coherence was not as clearly reflected in the written plans.

The integration evident in oral accounts reflects the dynamics between participation and reification (Wenger, 1998). The recorded discussions during CoP meetings show the fluidity of participation, where ideas are negotiated and emerging understandings of i-STEAM are expressed without the structural constraints of formal documents. Lesson plans, as reified artifacts, formalize knowledge within prescribed formats, which explains why they lag behind oral discourse. This disparity suggests that the temporal and structural limits of reification are more significant than any weakness in teachers’ conceptual appropriation, consistent with studies on the interplay between oral meaning-making and written formalization (Auby et al., 2024; Daele, 2010; Gómez-Blancarte & Miranda, 2021).

The predominance of oral interaction in conceptual integration suggests that teachers’ pedagogical reasoning unfolds through situated cognitive activity. Scholarship on teacher cognition demonstrates that teachers think and act within the social and discursive contexts of their work rather than apart from them (Borg, 2003, 2015; Contreras et al., 2020; Öztürk, 2021; Shavelson & Stern, 1981). Their judgments, decisions, and hypotheses take form through ongoing interaction with colleagues and through the use of artefacts that mediate professional reasoning. Speech is central to this process because it organizes and advances thought while simultaneously making it visible to others. Classic studies of lesson planning indicate that teachers rarely follow linear or prescriptive models; instead, planning develops as a cyclical process of formulating, testing, and revising ideas (Yinger, 1980; Wing-Mui So, 1997). These findings reveal that pedagogical reasoning is inherently dynamic and context-dependent, emerging most vividly in oral exchanges that allow teachers to negotiate meanings and explore multiple domains of thought before these are reified in written form.

Contemporary research further supports this view. Hutner and Markman (2016) argue that teachers’ cognition is socially distributed across individuals and departmental cultures, meaning that professional knowledge is shaped through shared representations rather than isolated reflection. Similarly, Adams and Krockover (1997) show that novice teachers rely heavily on dialogue with peers to articulate concerns and validate instructional decisions, emphasizing the reflective and collaborative nature of professional thinking. Through such interactions, oral communication functions as a collective cognitive space in which teachers co-construct, test, and refine pedagogical ideas. From a planning-as-practice perspective, this process illustrates the situated nature of teacher agency, as meaning is jointly constructed and only later stabilized in written documentation (Priestley et al., 2012). The difference between oral and written modes thus reflects a movement from open, generative reasoning to structured, finalized representation.

The feasibility of the lesson plans developed by the teachers can be attributed to their authorship, allowing them to tailor the proposals to the realities of the public schools where they work, considering the context and available resources. This is one of the advantages of a CoP, as it facilitates the creation of solutions. This advantage is evident in the presentation of the lesson plan developed by Group 2. The participants justified their choice of the sustainability theme, explaining that it is a broad topic with many possibilities for adaptation to local realities:

Katemari: “[…] There was a school where I made sustainable cookies with the students. We used banana biomass […]. We also made jam from the white part of the watermelon—the students ate the red part, leaving the white. […] This way, we reduced the waste generated from the school meals. The project’s idea, as discussed with the team, is to work closely with students, starting within the school and its surrounding neighborhood. If we aim for something too distant, it becomes harder to implement.”

The analysis of the lesson plans shows that collaborative work within the CoP supported the co-construction of more integrated and contextually grounded proposals, drawing on multiple knowledge domains. Despite these advances, challenges remain in areas such as engineering design, evaluation, and coordination of group work. The following section examines participants’ integration profiles based on interview data, exploring how distinct levels of i-STEAM understanding were internalized.

4.3. What Level of Integration Was Achieved by the Teachers Concerning i-STEAM?

Understanding the level of disciplinary integration achieved by teachers is essential for evaluating their appropriation of i-STEAM as a pedagogical approach. To this end, we applied an adapted version of the integration complexity staircase proposed by Gresnigt et al. (2014). This framework was used to analyze teachers’ discursive representations during interviews, allowing us to infer the conceptual maturity underlying their lesson-planning decisions.

As summarized in

Table 5. Integration profile evidenced in the teachers’ statements.

Profile	Number of Teachers	Statements
Connected	4	“You need to know mathematics, and it expands with various tools and the use of technology, whether it is simple and accessible to students or even more advanced technology.” (Adriana)
Nested	2	“We use things that we previously thought would distract students, like technology. However, we can transform this into a way that motivates them to engage with science and mathematics.” (Sônia)
Multidisciplinary	4	“I understand it as truly collaborative work, where teachers would sit down, come together, and create a project.” (Marie)
Interdisciplinary	4	“Instead of using mathematics in isolation, for example, I would combine mathematical knowledge with other fields to solve the problem, merging forces.” (Maria Laura)

, all the teachers expressed integrated conceptions of i-STEAM at some level, although none explicitly articulated a transdisciplinary perspective.

However, triangulation with planning data revealed that some participants incorporated transdisciplinary elements in practice, specifically through proposals involving families, traditional communities, or local institutions, despite not naming these features in theoretical terms. Teachers in the “connected profile” category understood i-STEAM as the coordination of separate disciplines, often enhanced by technological tools, but lacked articulation of shared epistemological purposes. In practice, this profile was consistent with group lesson plans 1 and 3, both of which were classified as “pseudo-i-STEAM and “in the process of acquisition” (see Table 4). These plans posed broad questions but lacked coherence across disciplines, suggesting that they were in the initial stages of integration.

Teachers in the “nested profile” view embedding goals from one discipline into the context of another. Lesson plans 5 and 6, developed by teachers in this profile, were rated as “pseudo-i-STEAM”, with limited integration strategies and no transparent design methodology. Teachers in the “multidisciplinary profile” emphasized collaborative planning and thematically organized content, but without integrated learning objectives. The planning developed by this group (Plan 2) addressed cultural biodiversity and involved contributions from the arts and humanities, but was also rated as “pseudo-i-STEAM” due to a lack of articulation among disciplines.

Teachers in the interdisciplinary profile explicitly articulate the integration of disciplinary knowledge to solve real-world problems. This conceptual clarity was reflected in lesson plans 4, 7, and 8, all of which achieved the “i-STEAM” classification and the “basic” level of practical quality. These plans addressed authentic, context-based challenges and combined scientific, mathematical, and artistic practices, demonstrating more mature integration.

During the interviews, no teacher explicitly referred to the involvement of external agents in solving the proposed problems. However, when analyzing the group-developed lesson plans, institutions. Evidence of the involvement of various agents, including families, other school staff, and establishments and institutions within the surrounding community, was observed. This difference between the verbal conceptualization of i-STEAM and its practical application in planning suggests that teachers can incorporate transdisciplinary elements into their proposals, but do not articulate this process in theoretical or conscious terms.

In line with several studies (Dare et al., 2019), there is limited understanding of the meaning of i-STEAM because a unified vision of what this approach entails has not been developed. Nonetheless, this research highlights progress toward this understanding, driven by participation in a CoP, which appears to serve as a social mediation environment. In this space, interactions have enabled advancements in the understanding and application of integrated practices, even though this process is still ongoing.

To further clarify how these conceptual profiles were reflected in practice, we triangulated interview data with the outcomes of the lesson plans developed by the participants.

Table 6. Connections between teachers’ integration conceptions, lesson plans, and corresponding rubric scores.

Teacher (Pseudonym)	Integration Profile (Conception)	Lesson Plan(s)	Lesson Plan Integration (Theoretical/Practical Classification)	Theoretical Score (Range: 0–9)	Practical Score (Range: 0–48)
Ada	Interdisciplinary	Plan 2 (Group)	Pseudo-i-STEAM/In the process of acquisition	2	14
		Plan 7 (Individual)	i-STEAM/Basic	5	20
Adriana	Connected	Plan 3 (Group)	Pseudo-i-STEAM/In the process of acquisition	1	12
Bertha	Multidisciplinary	Plan 3 (Group)	Pseudo-i-STEAM/In the process of acquisition	1	12
Frida	Interdisciplinary	Plan 2 (Group)	Pseudo-i-STEAM/In the process of acquisition	2	14
		Plan 5 (Pair)	Pseudo-i-STEAM/In the process of acquisition	4	11
Gladys	Connected	*	*	*	*
Hedy	Interdisciplinary	Plan 2 (Group)	Pseudo-i-STEAM/In the process of acquisition	2	14
		Plan 8 (individual)	i-STEAM/Basic	5	17
Katherine	Connected	Plan 2 (Group)	Pseudo-i-STEAM/In the process of acquisition	2	14
		Plan 4 (Pair)	i-STEAM/Basic	5	16
Katemari	Multidisciplinary	Plan 1 (Group)	Pseudo-i-STEAM/In the process of acquisition	2	12
		Plan 5 (Pair)	Pseudo-i-STEAM/In the process of acquisition	4	11
Maria Laura	Interdisciplinary	Plan 4 (Pair)	i-STEAM/Basic	5	16
Marie	Multidisciplinary	Plan 3 (Group)	Pseudo-i-STEAM/In the process of acquisition	1	12
Marta	Multidisciplinary	*	*	*	*
Mary	Nested	Plan 1 (Group)	Pseudo-i-STEAM/In the process of acquisition	2	12
Mileva	Interdisciplinary	Plan 1 (Group)	Pseudo-i-STEAM/In the process of acquisition	2	12
		Plan 6 (individual)	Pseudo-i-STEAM/In the process of acquisition	4	13
Rosalind	*	Plan 1 (Group)	Pseudo-i-STEAM/In the process of acquisition	2	12
Valerie	*	Plan 3 (Group)	Pseudo-i-STEAM/In the process of acquisition	1	12

Note. * The teacher did not participate in that activity.

presents a direct comparison of each participant’s integration profile, as identified in the interviews, with the corresponding classification of her lesson plan(s). This table offers readers an at-a-glance view of how conceptual profiles were—or were not—translated into practical i-STEAM designs.

In some cases, teachers first contributed to a collective lesson plan and later produced an individual or adapted version, resulting in a double classification (e.g., “Pseudo-iSTEAM → i-STEAM”). A dash indicates teachers who participated only in interviews.

This cross-analysis shows three relevant patterns: (i) coherence, when conceptual profiles aligned with practice (e.g., Maria Laura, Hedy); (ii) partial or inconsistent translation, when teachers expressed higher levels of integration conceptually but their plans remained at a pseudo-iSTEAM level (e.g., Mileva, Frida); and (iii) progression, when teachers advanced from pseudo-iSTEAM in collective planning to i-STEAM in later adapted plans (e.g., Ada, Katherine).

Although no formal baseline interviews were conducted, the transcripts from the first meeting and early lesson plan drafts provided indirect indicators of teachers’ initial perceptions of STEAM integration. For example, during the first meeting, Teacher Frida described a project using Scratch as a STEAM experience: “I worked on a volunteer project in rural schools. The aim was to teach children to develop small games on this platform (Scratch). It was gratifying; the group was mixed, and despite different skill levels, they managed to create great games.” This illustrates the initial perception that any technology-based project could be considered STEAM, reflecting a limited understanding of integrated STEAM practices.

Similarly, Teachers Ada and Hedy initially developed pseudo-iSTEAM plans in small groups. Loosely connected disciplinary approaches characterized these early plans. After engaging in collaborative discussions and group planning, both teachers refined their proposals into individual, context-sensitive i-STEAM plans, demonstrating a more coherent integration across different disciplines and an increased attention to real-world contexts.

These examples provide concrete evidence of participants’ starting points and allow tracing the evolution of their understanding and planning. Overall, participation in the CoP facilitated conceptual shifts and improvements in didactic planning, highlighting the value of collaborative reflection, discussion, and iterative design in supporting teachers’ development of integrated i-STEAM practices.

5. Conclusions, Limitations, and Future Perspectives

This study analyzed the impact of a CoP in supporting public-school teachers’ understanding of i-STEAM. Drawing on collaborative interactions, meeting records, and teaching proposals, the findings demonstrate that the CoP enabled the construction of shared meanings around interdisciplinary teaching and fostered pedagogical integration across disciplinary boundaries.

The group was characterized as a CoP according to three dimensions (Wenger, 1998)—mutual engagement, joint enterprise, and shared repertoire—evidenced by the collaborative relationships among teachers from different fields, the shared goal of developing strategies to engage girls in science, and the circulation of pedagogical materials and ideas, respectively.

Teachers’ lesson plans revealed how conceptual discussions within the CoP were translated into practice, with several proposals incorporating key elements of i-STEAM integration, such as problem-based learning, cross-disciplinary content alignment, and attention to socially relevant issues.

Notably, all the teaching proposals reached a level of feasible classroom application, and several advanced to the theoretical level of integration, even across diverse educational stages such as early childhood education, EJA, and high school. This contrasts with previous studies in which integrated proposals were typically confined to higher primary grades. The findings suggest that teacher authorship and contextual ownership, central features of the CoP, contributed to the relevance and applicability of the designs, in contrast to the top-down nature of many professional development programs.

Although the interviews were conducted at a single point in time, they took place at the conclusion of an active participation process in the CoP meetings. They followed the development of both collaborative and individual lesson plans. The analysis of interactions during the meetings and the comparison between collectively and individually adjusted plans indicate a broadening of understanding regarding disciplinary integration. In this sense, the interview served as a reflective synthesis of the formative trajectory, revealing signs of a more integrated appropriation of i-STEAM. Nevertheless, we acknowledge that confirming trajectories of conceptual change requires longitudinal studies.

This study contributes to the literature by addressing two underexplored dimensions: first, the empirical examination of CoPs as contexts for i-STEAM learning; second, the use of collaboratively developed lesson plans as evidence of teachers’ conceptual appropriation of interdisciplinary pedagogies. Drawing on an adapted analytical framework, the analysis of the planning process provides a structured basis for evaluating how teachers engage with integrative curriculum design in collaborative professional settings.

The results also highlight that supporting teachers in articulating their practices via shared theoretical language is as important as fostering the practices themselves. Many participants demonstrated advanced integration in their lesson plans but struggled to name or conceptually justify these practices. This highlights the need for formative spaces that combine hands-on co-design work with opportunities for reflection, meta-discussion, and theoretical grounding.

In addition, the inclusion of teachers from diverse educational levels and contexts helped challenge prevailing assumptions about where i-STEAM belongs in the school system. The study shows that when supported by collective, context-sensitive processes such as CoPs, i-STEAM can be meaningfully developed across a wide range of educational realities.

Two main limitations should be acknowledged. First, the analysis was based on a subset of CoP participants, which may have constrained the diversity of perspectives considered. Second, the short duration of the CoP limited the possibility of examining long-term processes such as the implementation, evaluation, and refinement of the proposals. These constraints also restrict the generalizability of the findings, which should be interpreted in light of the specific institutional and cultural context in which the study was conducted.

Despite these limitations, the results offer relevant implications for teacher education and educational policy beyond the national setting. The use of CoPs as a strategy for supporting interdisciplinary planning, situated professional learning, and teacher agency may inform the design of professional development programs in other public education systems facing similar structural challenges. Promoting collaborative, context-responsive models of teacher learning emerges as a necessary condition for advancing integrative pedagogies such as i-STEAM.

While the findings reflect the sociocultural realities of Brazilian public schools, several design features of this Community of Practice (CoP) are broadly transferable to other contexts. Central is the intentional gathering of teachers from diverse disciplines and school contexts, with lesson plans co-designed around their own priorities rather than imposed curricula. Other adaptable elements include sustained, low-cost digital communication (e.g., WhatsApp) and strong teacher authorship, which together enhance both relevance and feasibility.

Context-dependent aspects include the collective aim of inspiring girls in science—highly appropriate where gender gaps are significant—but are readily replaced by other equity goals in different settings, and the locally selected themes, which naturally vary across contexts. Clearly distinguishing these transferable structures from context-specific purposes can guide adaptation of the CoP model to a wide range of educational systems.