Introduction

Integrated STEM Education involves combining knowledge from multiple disciplines, including science, technology, engineering, and mathematics (STEM). This approach enables students to develop a range of skills, such as problem-solving and innovative thinking (Watermeyer and Montgomery, 2018). STEM education is highly regarded worldwide as a way for countries to showcase their commitment to educational reform. However, a significant challenge in implementing STEM education is the shortage of suitable highly qualified STEM teachers (Wright et al. 2019).

Encouraging teachers of STEM-related subjects to transition into Integrated STEM education is a crucial strategy for increasing the number of STEM teachers (Ketelhut et al. 2020). Considerable research has been conducted on this topic, primarily by providing professional development programs for STEM educators (Nadelson et al. 2012; Laffey et al. 2013). However, it remains a point of contention whether short-term programmes for training teachers can successfully provide long-term support for their development. Amrein-Beardsley et al. (2013) highlighted that the complexity of teachers’ professional abilities, including educational beliefs, professional knowledge and teaching skills, among other factors, is the main reason for this issue. This poses a challenge for short-term training programmes, which may struggle to address all aspects adequately.

From the perspective of teachers’ identity transformation, exploring the characteristic changes in teachers during the transformation and development stage presents a novel theoretical framework. El Nagdi et al. (2018) highlights that the professional identity of teachers is a dynamic and progressive process, created through the interaction of personal and professional characteristics with new educational experiences. Teachers Professional Development Theory suggests that teachers go through distinct stages of development, ranging from novice to expert, with each exhibiting unique characteristics (Berliner, 1988). However, for STEM teachers, how did they transition from subject teaching to integrated STEM education? What changes occur in teachers’ understanding and practice of STEM education during the process of identity transformation? Currently, there are few studies that conduct in-depth investigations on these issues.

Literature review

Teacher Identity Transformation in integrated STEM Education

Integrated STEM education is “the teaching and learning of the content and practices of disciplinary knowledge which include science and/or mathematics through the integration of the practices of engineering and engineering design of relevant technologies” (Bryan et al. 2015). Integrated STEM education provides students with a comprehensive situation of using multidisciplinary knowledge, supporting repeated testing and refining solutions. In STEM education activities, children learn how to think and invent, because these skills and experiences are what they value. (Galanti and Holincheck, 2022).

Integrated STEM education was introduced in China in 2014 and has since become a focal point in the reform of K-12 education. Despite its growing prominence, the long-standing subject-centered pedagogical approach and the lack of a comprehensive technology and engineering curriculum present considerable obstacles to the implementation of STEM education in Chinese schools. The current stage of STEM teacher training, curriculum development, and talent cultivation is in its infancy, necessitating a shift in educators’ perceptions, attitudes, and self-efficacy to embrace integrated STEM education (Zhong et al. 2022). Financial constraints and deficiencies in instructional management practices serve to further complicate this transition (Wang et al. 2020). Nevertheless, as STEM education develops in China, an increasing number of teachers are transcending the conventional boundaries of their respective subjects, gradually adopting integrated STEM pedagogical approaches.

The process of transitioning teachers to integrated STEM education entails a shift in teachers’ identities. Teachers’ identity refers to an individual’s core beliefs about teaching and becoming a teacher (Grier and Johnston, 2009). According to Grier and Johnston (2009), teachers’ identity is a dynamic process that continuously evolving and developing itself through personal and professional experiences. During this process, teachers develop or transform their beliefs and values about the meaning of being a teacher and the type of teacher to strive for as they interact with different policies, school and classroom settings, and colleagues. As a result, there are three features of teacher identity, namely multiplicity, discontinuity, and sociality (Akkerman and Meijer, 2011). As a general theory of teacher identity transformation, a considerable number of studies have examined the factors affecting teacher identity transformation from different aspects (Fouad and Bynner 2008; Xu and Huang 2021). However, in the field of STEM teacher education, few studies have revealed the characteristics and patterns of subject teachers’ transition to STEM.

The challenges to be a STEM teacher

The creation of a new identity for STEM teachers occurs during the process of STEM instructional design and practice, necessitating teachers to assume a variety of roles and duties associated with this identity. According to Slavit et al. (2016), STEM teachers’ role encompasses multiple facets, including being learners, explorers, investigators, curriculum developers, negotiators, and collaborators. Yang’s et al. (2021) research classifies STEM teachers into three groups: designers, implementers, and disseminators. As designers, teachers autonomously or collaboratively create innovative curricula and activities for personal or group use. Implementers enact interdisciplinary lessons in class. Disseminators facilitate or lead STEM implementation expansion.

The multiplicity of STEM teachers’ roles means that teachers will experience complex tracks and face more challenges, conflicts and even crises through their careers. Carrier et al. (2017) investigated the development trajectory of pre-service teachers and found that they were initially on the edge of the teacher community, and moved to the center with the accumulation of experience. The journey is characterised by moving from a student role, to a teacher candidate, and eventually to a fully qualified teacher. Galanti and Holincheck (2022) identified a significant tension that exists between the various roles that STEM teachers must fulfil as learners, professional teachers, and innovators. The formation of teachers’ identities as learners is shaped through their participation in both formal and non-formal educational experiences. STEM teachers gain their expertise by accumulating classroom experience and completing teacher education courses. To become innovators in STEM education, they must continue to overcome challenges, engage in problem-solving activities, and gradually acquire the skills to design and implement effective STEM initiatives. Zhang and Jiang (2023) identified three phases of teacher identity formation concerning their emotional development: ‘an interested yet baffled learner’, ‘an enthusiastic but apprehensive explorer’, and ‘an exhilarated yet discontented instructor’. Throughout this progression, teachers experience both affirmative and negative emotions that are entwined, facilitating the shaping and establishment of their professional identities.

Factors affecting STEM teachers’ identity

An array of factors, including personal, social, cultural, and situational influences, have intertwined to significantly impact the formation of a teacher’s identity (Jiang et al. 2021). In terms of gender differences, it has been observed that female teachers tend to display more enthusiasm toward STEM teaching and actively engage in collaborative activities, whereas male teachers may express more pessimistic views (Al Salami et al. 2017; Wegemer and Eccles, 2019). Wang et al. (2011) investigated teachers’ perceptions from different professional backgrounds on the integration of STEM subjects into their practices. The findings indicate that teachers typically begin integrating STEM concepts in ways that suit their comfort level, leading to diverse understandings of STEM education in practice. Similarly, Grier and Johnston’s (2009) analysis of the transformation of mathematics and chemistry teachers into STEM education implies that instructors commonly depend on their pre-existing skill sets when transitioning to a new career.

The school’s location and type also have an impact on the development of STEM teachers. Yoon and Kim (2022) indicated that teachers working in urban schools were more likely to access diverse professional development opportunities, including the integration of ICT into their teaching practice. However, rural schools provide more opportunities for teachers, such as teaching science subjects that stimulate intelligence, linking scientific topics with rural life, and having a sense of satisfaction and job security (Goodpaster et al. 2012). Lesseig’s et al. (2019) study revealed that primary school teachers exhibit more autonomy and flexibility in aspects such as curriculum arrangement, teaching resource usage, and content selection. On the other hand, secondary schools offer more career awareness activities, such as organized field trips to science research and engineering facilities, and provide opportunities for students to participate in regional and state-level robotics competitions.

Based on the aforementioned investigation, it can be concluded that the move of subject teachers to STEM education leads to a transformation of teacher identity. This transformation is determined by internal aspects, such as gender and professional background, as well as external factors, including school location and school type. However, a systematic analysis of the dynamic development process from subject teachers to STEM teachers is currently lacking. Such an analysis could uncover changes in cognitive, practical, and other aspects during the transformation of a teacher’s identity. Exploring this issue can enhance our understanding of the characteristics and main challenges encountered by teachers during different stages of development, thereby providing more specific recommendations for the professional development of STEM teachers. Hence, this study’s central focus is to explore the following research questions (RQs):

RQ1: From subject teaching to STEM education, what are the stages of transformation that teacher identity has undergone?

RQ2: From subject teaching to STEM education, does the transformation of teacher identity follow a similar path?

RQ3: From subject teaching to STEM education, how do individual internal and external factors affect the transformation of teacher identity?

Method

Participants

The study identifies suitable research participants through the Zhejiang Province Master Teachers Network Workshop, an online training community supported by the Department of Education in Zhejiang Province. This platform aims to enhance teacher excellence by creating high-quality digital educational resources, facilitating online training, and offering teaching guidance. It is structured around provincial master teachers and includes a community of regional subject leaders and key school instructors.

With the help of provincial teaching master teachers, we established specific criteria for the selection of research participants, which included: (1) having more than 10 years of teaching experience, (2) currently involved in integrated STEM teaching, and having more than 3 years of STEM teaching experience, (3) to be able to find public press information on the internet about this teacher who is leading STEM activities. Ultimately, 10 STEM teachers from 6 different regions in Zhejiang Province were interviewed. Basic information about these teachers is shown in Table 1.

Table 1 The background information of interviewees.
Full size table

Data collection

Relevant data were collected using personal in-depth interviews. The question design drew upon research by Grier and Johnston (2009) and Zhang and Jiang (2023, and encompassing four main areas: comprehension of STEM education among teachers, teacher self-development, STEM teaching practices, and teaching reflection. The semi-structured interview outline focused on five questions:

  1. 1.

    What inspired your shift from subject-specific teaching to STEM education?

  2. 2.

    While developing your identity as a STEM teacher, what transformations have you undergone regarding your perception of STEM education?

  3. 3.

    In what methods have you enhanced your capability to implement STEM teaching activities?

  4. 4.

    Please provide examples of STEM project activities you have carried out. What difficulties or challenges have you encountered, and how did you surmount them?

  5. 5.

    If you were to offer advice to a new STEM teacher, what recommendations would you suggest?

Interviews were primarily conducted via online video or telephone from the commencement of October 2022 until the conclusion of December, with an average duration of 50 min. Further, each individual who agreed to participate in the research as an interviewee was asked to sign a consent form confirming that they had understood the implications of their involvement and that they were willing to participate.

Procedure and data analysis

Interview Text Encoding Scheme

This study employed grounded theory to code and analyse the interview transcripts. Grounded theory is a qualitative research method developed by Glaser and Strauss (1992). It helps us to understand patterns of human behavior in social contexts through the analysis of interview or field survey data. Grounded Theory employs iterative analysis through three levels of coding: open, axial, and selective coding, to establish connections between concepts and form explanatory theories. NVivo 11 was employed to conduct qualitative analysis of the interview transcripts.

Firstly, open coding was independently executed by two experienced researchers, attaining an acceptable consistency level (Cohen’s Kappa = 0.712). In instances where any concepts experienced discrepancies in labelling, the researchers conferred with each other based on the contextual information provided in the text until an agreement was made. The process of open coding adheres to the saturation principle of grounded theory. This principle posits that data meets research requirements when key concepts are reiterated until no new ones emerge. Eventually, the study identified 155 concept labels. These labels were grouped based on their similarities, and the most vivid and representative statements capturing the essential characteristics were selected to form 24 open codes. Secondly, the study categorises based on the content attributes of open coding, creates axial coding and identifies eight categories: understanding of stem concept, knowledge learning, curriculum design, project development, teaching practice, teacher role, teaching reflection and developmental reflection. Finally, the process involved selective coding, which led to the formation of four core categories: teacher cognition, instructional design, instructional practice, and instructional reflection. The final coding results are shown in Table 2.

Table 2 Coding schema.
Full size table

Epistemic Network Analysis

Epistemic Network Analysis (ENA) is a quantitative ethnographic technique used for modeling the structure of connections in data. ENA was originally developed to model cognitive networks— the patterns of association between knowledge, skills, values, habits of mind, and other elements that characterize complex thinking (Shaffer et al. 2009). In recent years, ENA has been utilised in various forms of research pertaining to teacher education, including examination of teachers’ TPACK, beliefs, and collaborative problem-solving (Oner, 2020; Benna and Reynolds, 2021). Moreover, the ENA serves as a valuable complement to the Grounded Theory. Grounded theory aids researchers in constructing theoretical models and offers abstract explanations of latent variable relationships pertaining to a given topic buried within the data (Glaser and Strauss, 1992). ENA, on the other hand, facilitates the analysis of the interplay between various factors within the theoretical model, while also monitoring changes in the dynamic relationships between these variables over time (Swiecki et al. 2020).

In ENA, three crucial parameters prevail: unit, section, and coding. For this study, we are analyzing teaching groups with diverse genders, professional backgrounds, school categories, and school regions as the units. Our primary objective is to identify the variations in cognitive networks among these distinct teacher groups. Each sentence in the interview responses constitutes a section. Elements that co-occur within the same section are correlated, while elements in different sections remain unrelated. The coding process involved coding the understanding of STEM concepts, STEM pedagogical practices, and reflections on teaching. Elements that were mentioned were marked as “1” and those that were not mentioned were coded as “0”. For this study, ENA 1.5.2 Web Tool software was selected to analyze the data. This software enables researchers to present multidimensional data analysis results in a two-dimensional space, facilitating the observation of cognitive network changes during the teacher identity transformation process from a holistic perspective.

Results

The stages of teachers’ identity transformation

The coding analysis of the interviews revealed that the changes in teachers’ identities were reflected in four key areas: the understanding of STEM education, the designing of STEM activities, teaching practice and reflective teaching. During the interviews, researchers prompted participants to reflect on their understanding of STEM education at various stages by asking follow-up questions. This approach revealed temporal cues in the interview content. For example, participants might state, “In the beginning, I learned through imitation,” “Subsequently, I commenced the design of my own curriculum,” and “Presently, I lead a team.” These temporal indicators, along with the coded content, allowed the study to map the transformation of teachers’ identities into three stages: imitation, exploration, and innovation, referencing the frameworks proposed by Slavit et al. (2016) and Yang et al. (2021). To delineate teachers’ identities across these stages, the study selected representative words from the open coding and categorized them into relevant dimensions, as depicted in Table 3. Table 3 outlines the characteristics of the teacher community at various stages of identity transformation within STEM education.

Table 3 Stages and characteristics in the transformation of teachers’ identity.
Full size table

Imitation stage:

① Teachers often start as apprentices, participating in STEM activities led by experienced educators: “I participated in STEM activities organized by experienced teachers as an apprentice.” (T3).

② They improve their understanding by mimicking successful lessons: “Imitating successful lessons or examples in order to improve understanding of STEM education.” (T4).

③ Initially, they perceive STEM education as merely creating an object: “The initial perception of STEM education was an activity of ‘making an object’ “ (T7).

④ There is a desire to go beyond traditional subject teaching: “I wanted to step out of my comfort zone (of subject teaching) and try to incorporate my own subject or experience in STEM activity design.” (T1).

Exploration stage:

① Teachers begin to see the broader scope of STEM, recognizing its cross-disciplinary applications: “I realized that the most significant distinction between STEM and discipline-based teaching is that STEM learning activities will give students more opportunities to apply their knowledge across disciplines and engage in independent investigations.” (T1).

② They identify Project-Based Learning (PBL) as a promising approach for curriculum design: “the PBL model was identified as a promising approach for curriculum design. “ (T9).

③ Collaboration becomes crucial, often requiring teamwork with colleagues: “ Some STEM activities require collaboration with colleagues to complete them.” (T2).

④ Teachers acknowledge the need for continuous learning to effectively implement STEM education: “As I progress in my practice of STEM education, I find that I must learn more and more, such as how to organize more effective STEM instruction, how to promote interdisciplinary collaboration among teachers, and how to assess student learning.” (T10).

Innovation stage:

① Teachers integrate local cultural resources into STEM curricula: “I began to recognize the value of local intangible cultural heritage resources and incorporated them into the design of STEM curriculum activities.” (T1).

② They emphasize the importance of equity in STEM education: “if only some part of children can participate in integrated STEM activities, I think it is not enough.” (T6).

③ Innovative practices include leveraging external partnerships: “we realize that the capacity of schools and individual teachers is limited, so we try to incorporate the power of enterprises and families, and innovate through school-enterprise interaction and home-school collaboration in the STEM teaching model.” (T5).

④ Experienced teachers become mentors and advocates for STEM education: “I have also taken a few apprentices now, all of them come from my school and neighboring schools.” (T8).

The path of teachers’ identity transformation

Through examining question 1, we have gained insight into the stages and characteristics of the transformation of STEM teachers’ identities. Utilizing this information, we have devised a theoretical model of STEM teachers’ transformation and progression, which is presented in Fig. 1. Before transitioning to STEM education, these teachers had extensive experience in their respective subject areas. This research aims to determine if they experience similar developmental stages during their transformation into STEM educators. Comparative analysis of teacher interviews revealed three distinct pathways in the transformation of teacher identity. These pathways are visually represented in Fig. 1.

Fig. 1: The theoretical model of STEM teacher transformation and development.
figure 1

The model comprised three distinct stages: imitation, exploration and innovation. Each stage was characterised by a cyclical process of cognition, design thinking, practice and reflection. The model also identified three pathways of transformation and development, which were associated with the three stages.

Full size image

Path 1: From Imitation to Exploration

Two factors are pivotal in the transition to STEM education. Firstly, the introduction of new educational concepts and exemplary practices can prompt teachers to reconsider the essence of education. As one teacher noted, “The most significant insight that STEM provided was the realization that education is not merely about studying a specific course or preparing for examinations, but about cultivating students’ problem-solving abilities in authentic settings” (T10). Secondly, the mentorship of experienced educators accelerates the shift from the imitation to the exploration phase. For instance, one teacher expressed gratitude for their mentor’s support: “I would like to thank Mr. Zhao, an expert teacher, for his guidance over the past year. This included observing STEM activities, designing STEM teaching and learning tasks, and receiving invaluable feedback. I am now able to design STEM activities independently…” (T3).

Path 2: From Imitation and Exploration to Innovation

Teachers navigating the transition to STEM education often exert considerable effort. One teacher recounted the challenges faced when designing their own STEM curriculum: “I encountered numerous difficulties, including determining content, organizing activities, fostering teacher collaboration, and evaluating learning outcomes…” (T2). This process also requires a significant degree of creativity. For example, in designing a localized STEM curriculum, a teacher incorporated the cultural heritage of the ancient Taihu Lake Water Conservancy Project, collaborating on projects like “Little Farmers of Taihu Lake” and “Bridges over Taihu Lake.” Moreover, these teachers often assume leadership roles, as one expressed: “I hope that through my work, I can attract more teachers to participate in this process” (T5).

Path 3: Leaping from Imitation to the Innovation Stage

Among the ten teachers interviewed, only two, T6 and T8, followed this particular pathway. Upon reviewing their interview responses, both teachers had over 15 years of experience teaching physics. One teacher reflected, “Much of my previous teaching aligns with interdisciplinary principles… The STEM concept has provided a framework to reevaluate and consolidate my past teaching practices” (T6). The other teacher remarked, “STEM has offered a theoretical framework that helped me reorganize my integrated activities from the past 20 years in mechanics, optics, and electricity, providing clarity on their significance” (T8).

The internal and external factors affect teachers’ identity transformation

For a deeper understanding of how individual internal and external factors affect the transformation of teachers’ identities, this study employs eight dimensions created through axial coding as its analytical framework. Additionally, a comparative analysis of STEM teachers from different genders, professional backgrounds, school types, and locations is conducted. Throughout the research process, ENA was used to describe changes in the group cognitive network over time in various stages. Each graph in the study displays the overlapping of the cognitive network structures of teachers across various grouping conditions.

Genders differences in teacher identity transformation

Figure 2 shows a comparison of ENA across teachers of different genders at the three stages. The rectangular dashed boxes indicate the confidence intervals at the 95% level for the center of mass position. Coloured line illustrates that the associated groups exhibit greater significance in this dimension than the control group. The width of the line reflects the strength of the link between them.

Fig. 2: Comparison plot of STEM teachers in different gender.
figure 2

The red line in the figure represents the data for male participants, while the blue line represents the data for female participants. The interpretation of each code is provided in Table 3.

Full size image

During the imitation stage, there was a notable difference between male and female teachers along the x-axis (male group M = −0.21; female group M = 0.12; u = 1.00; p = 0.03 < 0.05; R = 0.91). The disparity was emphasized by the male teachers’ stronger connections in the factors of teacher role (TRo), knowledge learning (KL), and teaching reflection (TRe). These results suggest male teachers are more inclined to reinforce their STEM teaching identity through teaching reflection and knowledge learning, which is an essential component of their teaching role.

During the exploratory stage, there was also a significant difference between male and female teachers (male group M = −0.34; female group M = 0.12; u = 1.00; p = 0.03 < 0.05; R = 0.92). Male teachers primarily focused on teaching practice (TP), emphasizing on project design (PD) and developmental reflection (DR). In contrast, female teachers placed greater value on teaching reflection (TRe), which was closely associated with teaching practice (TP), teachers role (TRo), and curriculum design (CD).

The contrasting concerns of male and female teachers during the innovation stage are evident (male group M = −0.33; female group M = 0.34 ; u = 1.00; p = 0.01 < 0.05; R = 0.95). Male teachers concentrate on two key areas: the correlation between curriculum design (CD) and project development (PD), and the connection between teacher roles (TRo) and developmental reflection (DR). Conversely, female educators give more weightage to reflecting on teaching practices.

Overall, male educators prioritised project development (PD), whereas female educators exhibited greater emotional engagement and were more willing to participate in diverse teaching reflection (TRe) activities.

Professional background difference in teacher identity transformation

To comprehend variances in the identity transition of STEM teachers from diverse professional backgrounds, the research categorised physics and mathematics teachers into Group 1, the Chinese, English, and History teachers into Group 2. Figure 3 presents a comparison of the ENA of teachers from varying professional backgrounds in the three stages.

Fig. 3: Comparison plot of STEM teachers in different professional backgrounds.
figure 3

Group 1 is for physics and maths teachers, while group 2 is for Chinese, English and history teachers. The interpretation of each code is provided in Table 3.

Full size image

During the imitation stage, there was a significant difference observed among teachers from the two groups along the x-axis (Group 1M = 0.20; Group 2M = −0.21; u = 1.00; p = 0.02 < 0.05; R = 0.92). Primarily, Group 1 teachers demonstrated a stronger connection between STEM teaching reflection (TRe) and teachers roles (TRo). Group 2 teachers, on the other hand, showed more emphasis on comprehension of STEM concepts (UC) and teaching reflection (TRe).

During the exploration stage, Group 1 had a mean score of 0.25, while Group 2 had a mean score of −0.34. The Mann-Whitney U test showed a statistically significant difference (u = 24.00; p = 0.01 < 0.05; R = −1.00). Mathematics and Physics teachers demonstrated a higher level of commitment towards teaching practices (TP) and project design(PD), and paid greater attention to the developmental reflection (DR), which enhance student learning, the quality of schools, and personal growth. Group 2 teachers, however, primarily emphasised on teaching practices (TP) and teaching reflection (TRe).

During the innovation stage, significant differences were observed between Group 1 (M = 0.25) and Group 2 (M = −0.34) (u = 24.00; p = 0.01 < 0.05; R = −1.00). Group 1 developed a stronger connection between curriculum design (CD) and project development (PD), and enhanced their understanding of STEM education (UC). In contrast, Group 2 focused on teaching practices (TP) and forged stronger connections with teaching reflection (TRe), project design (PD), and an understanding of STEM education. Overall, teachers with backgrounds in mathematics and physics appeared to have a comparative advantage in implementing STEM education, as they could provide more examples to peers through STEM project development. In contrast, teachers with subjects background like Chinese and English concentrated on STEM teaching methodologies, displaying a more active engagement in teaching reflection within the context of STEM education.

School type difference in teacher identity transformation

Due to the different stages of students’ cognitive development, primary and secondary schools present distinct requirements for STEM education activities. Accordingly, the demand for STEM teachers varies across these categories. This study explored the cognitive differences between primary and secondary STEM teachers in the process of identity change, as shown in Fig. 4.

Fig. 4: Comparison plot of STEM teachers in middle school and primary school.
figure 4

The red line represents the participants from middle school, while the blue line represents those from primary school. The interpretation of each code is provided in Table 3.

Full size image

During the imitation stage, no significant difference was found between the two groups of teachers (Primary group M = −0.30; Secondary group M = 0.22; u = 33.00; p = 0.06 > 0.05; R = 0.76). However, the teachers in the primary group exhibited a much stronger linkage between their teaching practice (TP) and knowledge learning (KL). Secondary school teachers, in contrast, prioritised reflecting on STEM education, encompassing developmental reflection (DR) and teaching reflection (TRe), as well as knowledge learning (KL) and the role of teachers (TRo).

During the exploration phase, substantial discrepancies between the two cohorts were observed (Group 1M = −0.25; Group 2M = 0.20; u = 1.00; p = 0.02 < 0.05; R = 0.97). These discrepancies were primarily manifested through the fact that the primary school teachers evinced a stronger association between curriculum design (CD) and the teacher’s role (TRo), while the secondary school teachers group displayed a more robust link between curriculum design (CD) and teaching practice (TP).

During the innovation stage, a significant difference between the primary and secondary groups was observed (Primary group M = −0.27; Secondary group M = 0.33; u = 0.00; p = 0.01 < 0.05; R = 1.00). Notably, primary school teachers demonstrated a stronger correlation between teaching practices (TP) and developmental reflection (DR) in comparison to secondary school teachers, who exhibited stronger correlations between teaching practices (TP), curriculum design (CD), and teaching reflection (TRe). Primary school teachers consistently prioritised STEM teaching practices (TP), gradually developing an orientation towards their role as educators (TRo). In contrast, secondary school teachers focused earlier on changing their identity as a teacher and, at a later stage, improved STEM teaching practices through project design (PD) and curriculum design (CD). This could be attributed to the elevated complexity in the design of STEM education activities, knowledge requirements, and teaching organization in secondary schools.

Regional difference in teacher identity transformation

The cultural features of the environment where the school is situated, the available resources, and the opportunities for teacher training might affect teacher identity transformation. Consequently, this study contrasts and examines STEM teachers’ cognitive network configuration in rural and urban schools during their identity transformation process, as presented in Fig. 5.

Fig. 5: Comparison plot of STEM teachers in Urban and Rural.
figure 5

The red line shows participants from urban areas, while the blue line shows those from rural areas. The interpretation of each code is provided in Table 3.

Full size image

During the imitation stage, a notable difference was observed between the two groups of teachers (Urban group M = −0.48; Rural group M = 0.17; u = 20.00; p = 0.03 < 0.05; R = −0.90). Teachers belonging to the urban group exhibited a more pronounced correlation between knowledge acquisition (KL) and their role as teachers (TRo). Conversely, teachers from the rural group had a stronger emphasis on understanding STEM concepts (UC), and a greater association between knowledge learning (KL), teaching practices (TP), and teaching reflection (TRe).

Significant differences emerged between the two groups during the exploration stage (Urban group M = −0.38; Rural group M = 0.19; u = 1.00; p = 0.03 < 0.05; R = 0.90). Urban group teachers were primarily focused on curriculum design (CD), while their rural counterparts emphasized teaching practices (TP).

During the innovation stage, notable variations were found between the two groups (Urban group M = −0.44; Rural group M = 0.07; u = 0.00; p = 0.02 < 0.05; R = 1.00). Both sets of teachers prioritised teaching reflection (TRe), yet urban group teachers emphasised the correlation between teaching reflection (TRe) and developmental reflection (DR), while rural group teachers concentrated on the association between TRe and curriculum design (CD). This distinction in perspectives indicates the contrasting approaches of urban and rural teachers when observing and reflecting on STEM education. While urban teachers utilise reflection to enhance their curriculum design, rural teachers use it to solidify their knowledge of STEM concepts and its relevance to individuals, students, and learning development through teaching practices.

Discussion

The transition of subject teachers to STEM education is a unique experience in teacher professional development

The study found that the transformation of STEM teachers still follows the laws of the theory of teacher professional development stages, which is a development process from novice to experts (Berliner, 1988). “Novice” subject teachers need to understand the fundamental principles of STEM education, methods of curriculum design, and implementation through learning, observation, and imitation. Nevertheless, they are not complete “novices.” They possess a profound comprehension of their students, exhibit expertise in subject teaching, and possess widespread teaching experience. These characteristics allow teachers to readily venture beyond their subject-teaching comfort zone and transition into STEM education. During the shift, subject teachers take on varied teacher identities such as learners, team collaborators, curriculum designers, project developers, and teacher leaders. These findings align with the perceptions of STEM teachers’ multiple identities, as described by Slavit et al. (2016) and Yang et al. (2021).

The STEM teacher transformation model depicts multiple pathways of teacher identity transformation

The transition from subject matter teachers to STEM educators represents a new phase in teacher professional development. STEM teachers undergo stages of novice, competent, mature, and expert, as delineated by Berliner (1988). This study identified three distinct stages in the transformation of teachers: imitation, exploration, and innovation, constructing a model of STEM teacher development. The model illustrates varied pathways teachers may follow during their transformation process. While all teachers experience the imitation stage, a minority progress to exploration. A few advance through all stages, and a select few bypass exploration directly to innovation. These findings contribute to understanding the non-linear nature of teacher professional growth (Akkerman and Meijer, 2011).

These insights offer valuable implications for optimizing STEM teacher training. Novice subject teachers benefit from observing and modeling STEM teaching activities, which aids in grasping STEM concepts, alleviating concerns, and fostering willingness to depart from traditional subject teaching. Teachers with initial STEM experience benefit from resources on integrated STEM curriculum design and enhanced support for organizing and collaborating on STEM initiatives, accelerating their development as STEM educators. For subject experts, training in STEM education theory and curriculum design facilitates reflection on teaching practices and expedites transformation.

The impact of multiple internal and external factors has increased the complexity of teacher identity transformation

The analysis based on the Teacher Epistemic Network provides in-depth insights into teacher identity transformation. This study showed that female educators focused on STEM teaching practices and introspectively examined their roles. Furthermore, they demonstrated a high level of proactiveness, expressing their ambition to excel as educators and impact a wider audience. These findings closely align with the research conducted by Al Salami et al. (2017) and Wegemer and Eccles (2019). However, this study revealed that male teachers placed greater emphasis on interdisciplinary curriculum integration and project design during the transition process, which is not consistent with the results of Wegemer and Eccles’s (2019) study. It should be highlighted that Wegemer and Eccles’s study primarily examined teachers in the exploratory stage, while our research encompassed the entire trajectory of STEM teacher development. Moreover, our investigation underscored those teachers hailing from professional backgrounds such as physics and mathematics possessed a natural proclivity for transitioning into integrated STEM education. Conversely, teachers from non-STEM professional backgrounds necessitated additional efforts to attain competency in integrated STEM education. Hence, it is essential to account for the differences in teachers’ professional backgrounds to overcome resistance to the change in teacher identities.

The ENA analysis, which is based on teachers’ geographical regions and school types, has provided profound insights into the transformation of STEM teachers. The study revealed that primary school teachers place a strong emphasis on STEM teaching practices and reflection. Their contemplations encompass various aspects of STEM education as well as student, school, and personal development. In contrast, secondary school teachers tend to direct their focus towards curriculum design. The reasons for behind these differences can be attributed to the higher demands placed on STEM programs in secondary schools, including curriculum structure, content knowledge, and assessment criteria (Lesseig et al. 2019). In terms of schools’ geographical regions, urban teachers consistently focused on the core topic of curriculum design. This is because they benefit from greater access to information, resource support and developmental opportunities. In comparison, teachers in rural schools primarily rely on their own efforts and those of their colleagues to enhance their comprehension of STEM education. However, they excel at mobilizing community resources to facilitate the integration of STEM activities into their curriculum. For instance, “I will encourage parents with professional knowledge in engineering or technology to participate in STEM project activities, in order to establish cooperation between schools and enterprises.”.

Conclusion

Encouraging subject teachers transition to integrated STEM education is an important measure to enhance their professional abilities. Many studies have proposed that teacher training can transform teachers’ understanding of STEM education and enhance their ability to design and manage STEM activities. However, if the developmental stage of STEM teachers is not taken into account, as well as individual and school differences, relevant training activities may become ineffective.

These research findings can provide valuable insights for enhancing STEM teacher training. Firstly, targeted support and improved training effectiveness can be achieved by developing training programs that consider the characteristics of STEM teacher development stages. Secondly, measures should be formulated to motivate teachers to fully utilize their unique characteristics and strengths in relation to gender and professional background, to facilitate the transformation of subject teachers into STEM education. Finally, the unique advantages of schools and regions should be fully leveraged to more effectively utilise existing resources. This will help establish a regional STEM education community and achieve localisation of STEM education.

This study has constraints. Firstly, the sample may not be entirely representative as it primarily consisted of teachers with physics backgrounds, and fewer from other fields. Moreover, limitations are present in the data collected through interviews. To enhance the validity of our findings, future research should consider increasing the sample size and incorporating additional data sources. For instance, it may be beneficial to examine the course design and teaching videos of STEM educators’ professional development to gain a comprehensive understanding of the trends and characteristics of STEM teacher transformation.

Availability of data and materials

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.