The research team, led by Deepika Menon of the University of Nebraska-Lincoln together with colleagues at Southern Methodist University, Towson University, and Indiana University Southeast, used a mixed-methods design combining a validated self-efficacy survey with open-ended questionnaires administered at the beginning and end of science or STEM methods courses. The survey instrument, known as SETIS, measures three dimensions of confidence: personal capability, the ability to connect disciplines and engage students, and the capacity to manage materials and technology. Statistical analyses, including repeated-measures multivariate tests and paired comparisons, revealed significant gains across all three dimensions by semester’s end, with the largest improvement in teachers’ beliefs about their own abilities.

That growth matters because self-efficacy, a concept rooted in social cognitive theory, is strongly linked to whether teachers adopt innovative practices, persist through difficulties, and remain in the profession. Teachers with low confidence in STEM tend to avoid reform-based instruction and give up more easily when lessons falter. The finding that a single semester of hands-on, inquiry-based methods coursework, whether embedded in a science methods course or a dedicated STEM semester, produced medium-to-large gains suggests that teacher preparation programs can move the needle on the beliefs that ultimately shape classroom practice.

Yet the qualitative side of the study tells a more complicated story. Analyzing roughly 1,300 coded response segments, the researchers identified six broad categories of perceived challenges: teacher affect and experience, student-related concerns, content and curricular demands, pedagogical difficulties, support from colleagues and parents, and time and resources. By the end of the semester, participants reported significantly fewer overall challenges, particularly around science content knowledge and pedagogical confidence, but two concerns intensified rather than faded: worries about teaching engineering and doubts about integrating technology.

The researchers interpret this pattern through two complementary frameworks. Windschitl’s model of teacher dilemmas, spanning conceptual, pedagogical, cultural, and political tensions, helps explain how future teachers wrestle with what integrated STEM knowledge even means and how to enact it in real classrooms. Ertmer’s distinction between first-order barriers, such as limited time, materials, and institutional support, and second-order barriers rooted in personal beliefs and confidence, maps closely onto the challenges participants described. Content knowledge gaps and pedagogical uncertainty, the internal barriers, shrank after coursework. External constraints, especially time for planning and teaching STEM amid crowded elementary schedules, grew more salient as participants gained a clearer-eyed view of classroom realities.

Several findings stand out for their implications. Concerns about meeting state standards and assessment demands barely budged, with participants observing that elementary curricula leave little room for STEM when reading and mathematics dominate. One participant at a STEM integration school noted that despite the label, STEM instruction was rare. Meanwhile, structural equation modeling revealed that teachers who still felt unprepared in science or STEM content at semester’s end scored significantly lower on self-efficacy, confirming that content preparedness remains a critical lever. Intriguingly, identifying technology access as a challenge was associated with higher self-efficacy, suggesting that more confident teachers may simply be more aware of the resource constraints they will face.

The study also surfaced demographic patterns that warrant attention. Participants who had spent five or more years in college reported lower self-efficacy, possibly because their extended, full-year student teaching exposed them to a more sobering view of classroom challenges. Hispanic/Latino participants reported lower self-efficacy than their White peers, a result the authors connect to broader literature on inequities in STEM access and representation, and one they argue demands further investigation across more diverse populations.

The authors are careful about limits. Because data were self-reported at only two time points, and because most participants had not yet completed full-time student teaching, self-efficacy estimates may be somewhat inflated. The analysis also pooled data across five quite different programs, so it cannot isolate which specific course features drove the gains. Still, the breadth of the sample strengthens confidence that the overall pattern, rising efficacy alongside persistent and evolving concerns, reflects something real about how teacher preparation works.

The practical implications are pointed. The researchers recommend that preparation programs increase dedicated attention to engineering design and technology integration, precisely the areas where confidence lagged or declined. They suggest strategic school and informal-education placements so future teachers can witness successful integrated STEM instruction in action, something many mentor elementary classrooms rarely model. They also call for partnerships between universities and school administrators to address resource gaps, and for collaboration between STEM content faculty and education faculty to build better curricula. Longitudinal follow-up studies, the authors argue, should track whether the challenges these future teachers anticipate actually materialize once they enter elementary classrooms, closing the loop between preparation and practice.

Subject of Research: Preservice elementary teachers' integrated STEM teaching self-efficacy and perceived challenges

Article Title: “I learned a lot, but…”: preservice elementary teachers’ integrated STEM teaching self-efficacy and perceived challenges

Article References: Menon, D., Wieselmann, J. R., Haines, S., Asim, S., & Johnson, A. (2026). “I learned a lot, but…”: preservice elementary teachers’ integrated STEM teaching self-efficacy and perceived challenges. International Journal of STEM Education, 13(1), Article 56. https://doi.org/10.1186/s40594-026-00645-8

Image Credits: AI Generated

DOI: 10.1186/s40594-026-00645-8

Keywords: integrated STEM, self-efficacy, preservice teachers, elementary education, teacher preparation, STEM challenges, engineering education, methods coursework, mixed methods, structural equation modeling, educational psychology, K-12 STEM

The impetus behind this vision stems from the increasing realization that traditional educational frameworks, often siloed and static, fall short in catalyzing innovation and inclusiveness in STEM fields. The authors argue for a “field-initiated” approach—an infrastructure model that is organically informed by the needs and insights of researchers actively engaged in STEM education. This dynamic foundation ensures that the tools, datasets, and platforms developed are inherently relevant and adaptable to the fast-paced changes in both technology and pedagogy.

Central to their vision is the conception of infrastructure as multidimensional. Instead of a mere collection of physical assets such as laboratories or classrooms, infrastructure encompasses digital platforms, collaborative networks, methodological toolkits, and data ecosystems. These components collectively enable the continuous iteration of research hypotheses, pedagogical strategies, and policy frameworks, thereby accelerating the pace at which scientific knowledge translates into educational practice.

Technological advances underpinning this model are not merely supportive but transformative. The authors emphasize the role of big data analytics, machine learning, and cloud-based collaborative environments to synthesize vast, heterogeneous data streams—from student interactions and learning outcomes to socio-cultural variables influencing STEM engagement. This level of analytic sophistication promises to unravel complex causal links and emergent patterns that were previously obscured by traditional research methods.

The article also delves into the intricacies of governance within this infrastructure paradigm. It highlights the necessity of inclusive decision-making processes that draw from diverse stakeholders—including educators, students, policymakers, and technologists—to ensure equitable access and responsiveness to underrepresented groups in STEM disciplines. Such a participatory governance model is projected to dismantle longstanding barriers and foster a culture of openness and shared ownership.

Further, the authors propose strategic interoperability as a linchpin for the envisioned infrastructure. This principle mandates that various tools, datasets, and platforms adhere to standardized protocols, enabling seamless data integration and cross-disciplinary collaboration. Interoperability thereby magnifies the utility of each component, permitting cumulative insights and enhancing the scalability of successful educational interventions.

Emphasizing sustainability, the study urges the design of infrastructure that is resilient to technological obsolescence and adaptable to evolving research priorities. This involves modular architectures and open-source technologies that can be incrementally upgraded or reconfigured without wholesale system replacements—a crucial consideration in an era where both technology and educational standards evolve rapidly.

In considering the ethical dimensions, the authors take a proactive stance by embedding privacy-preserving mechanisms and bias mitigation strategies within the infrastructure design. Such safeguards are essential to maintain participant trust and uphold scientific rigor when engaging with sensitive educational datasets at scale.

Moreover, the envisioned infrastructure recognizes the integral role of teacher professional development. By providing educators with real-time access to research findings and analytical tools, it empowers them to tailor pedagogical approaches responsively, bridging the gap between research and classroom practice. This iterative feedback loop promises to generate a more adaptive and personalized learning experience for students.

Another transformative element is the incorporation of cross-sector partnerships, extending beyond academia to include industry, government agencies, and community organizations. These collaborations expand resource availability, diversify perspectives, and create pathways for applying STEM education research to real-world challenges, thereby enhancing societal relevance and impact.

An innovative aspect highlighted in the article is the use of virtual and augmented reality platforms within the infrastructure to simulate complex scientific concepts and environments. Such immersive technologies can revolutionize experiential learning, making abstract or inaccessible STEM phenomena tangible and interactive for diverse learner populations.

Importantly, the authors argue that this holistic infrastructure must be scalable and globally accessible, supporting both local context adaptations and international collaboration. By fostering a global STEM education research community, knowledge exchange is accelerated, and culturally responsive educational innovations proliferate.

Throughout the article, the researchers acknowledge the immense challenges associated with implementing such an ambitious infrastructure. These include securing sustainable funding models, navigating political and institutional inertia, and addressing the digital divide that may exacerbate educational inequities. Nevertheless, they present a strategic roadmap combining incremental steps and visionary goals to progressively realize this transformative agenda.

The potential implications of this new model extend far beyond academic research, promising to redefine STEM education paradigms worldwide. By uniting technology, collaboration, ethics, and pedagogy within a cohesive research infrastructure, the article posits a future where STEM education is more innovative, equitable, and impactful—equipping learners with the skills and mindset crucial for 21st-century challenges.

This visionary research underscores that the future of STEM education does not rest solely on isolated innovations but on cultivating an ecosystem deliberately designed to foster continuous learning, experimentation, and adaptation at scale. As STEM disciplines drive global innovation, the infrastructure described by Motz and colleagues may become the bedrock of educational evolution necessary to sustain this momentum.

In conclusion, the research articulated in “A field-initiated vision of research infrastructure for STEM education” heralds a paradigm shift. It calls researchers, educators, policymakers, and technologists to unite in constructing an infrastructure that is as flexible and multidimensional as the knowledge domains it serves. Through meticulous design and collaborative stewardship, this infrastructure has the potential to reshape how STEM education research is conceived, conducted, and translated into transformative educational experiences worldwide.

Subject of Research: Research infrastructure development to advance STEM education.

Article Title: A field-initiated vision of research infrastructure for STEM education.

Article References:
Motz, B., Diekman, A., Goldstone, R. et al. A field-initiated vision of research infrastructure for STEM education. IJ STEM Ed 12, 59 (2025). https://doi.org/10.1186/s40594-025-00581-z

Image Credits: AI Generated

DOI: https://doi.org/10.1186/s40594-025-00581-z