5

Navigating the Landscape of STEM Innovation and Implementation

As noted in previous chapters, one of the most distinctive features of the U.S. public education system is its decentralized structure and the resulting complexity and variability across national, regional, and local levels in how policies are made and how resources are invested in public education. In this chapter, we take a closer look at the current landscape of science, technology, engineering, and mathematics (STEM) education from Pre-K through high school with an eye toward how effective educational innovations arise, take root, and spread. Central to understanding this landscape is recognizing that there is an important distinction to be made between the typical configurations of actors, decision makers, and financial resources that are often involved in the development of evidence-based innovations as compared to the configurations that come into play as innovations are implemented, sustained, and spread across settings and populations. The lack of direct connections and the potential for misalignments across these two layers of the STEM education landscape provide context for why it is so challenging to get promising innovations to spread beyond their original implementation. The committee first takes up this implementation landscape—the dynamic configurations of actors and flow of financial resources—as it pertains to the creation and development of innovations and improvements, with particular attention to how federal investments are made. We then turn to the features of this landscape that are relevant to whether and in what ways evidence-based innovations are adopted, scaled, or sustained beyond the development stage, and whether and in what ways, too, the information is disseminated.

THE INNOVATION DEVELOPMENT LANDSCAPE

This section starts by examining the role of federal funding agencies in supporting research and development in STEM education. It first considers the knowledge that is created by these federal investments, and discusses challenges to mobilizing that knowledge in forms that are taken up by decision makers and enactors in educational settings. We then turn to instructional resources and interventions that arise from research grants and discuss some of the demands of aligning them with what schools and districts want and need; this includes examination of the difficulties of creating distribution systems with the capacity to disseminate them widely and support them as they take root in new settings. Finally, this section turns to innovations that arise from sources other than federal funding and considers some of the advantages and disadvantages of resources that come through alternative development pathways.

Federal Investments in the Creation, Development, and Evaluation of Innovations and Improvements in STEM Education

Most funding for research and development aimed at creating and studying innovations and improvements in STEM education and at evaluating their efficacy comes from the federal government, through award competitions run by federal agencies. As discussed in Chapter 2, the largest federal investments are channeled through the National Science Foundation’s (NSF’s) Directorate for STEM Education (formerly the Education and Human Resources Directorate), which is charged with providing national leadership to improve STEM education at all levels, and through the U.S. Department of Education’s (ED’s) Institute of Education Sciences (IES), which was created in 2002 as ED’s research arm and is intended to provide national leadership in expanding “fundamental knowledge and understanding of education from early childhood through postsecondary study” in all disciplines (Education Sciences Reform Act of 2002), including those related to STEM. Competitive awards made through these agencies fund both use-inspired research (Stokes, 1997), aimed at discovering new knowledge and building theoretical foundations to better understand processes and systems of learning, and applied research, which yields new evidence-based programs, practices, technologies, and models in STEM education.

ED has also introduced funding programs specifically focused on innovation. From 2010 to 2016, ED’s Investing in Innovation fund (i3), awarded grants with the goals of developing and testing promising new strategies or obtaining further evidence about expanded implementation of innovative strategies that were already supported by evidence. Grantees were expected to produce rigorous evidence about the strategy they were investigating and generally employed strong causal designs, such as randomized

controlled trials, designed to provide reliable estimates of the intervention’s or strategy’s average impacts. About one-third of the grants were validation or scale-up grants, which evaluated the effectiveness of strategies already supported by prior evidence at a larger scale and with new types of students or in new settings. A small set of these studies were focused on innovations related to STEM. The successor program to i3 is the Education Innovation and Research (EIR) program, administered by ED’s Office of Elementary and Secondary Education (OESE). Like the earlier i3 program, EIR funds a wide variety of projects at three levels (early-phase, mid-phase, and expansion). All EIR projects must serve high-need students, and mid-phase and expansion grants are expected to contribute to building a knowledge base about the contexts in which practices are effective and cost-effective.

Investigators who receive competitive awards from these agencies are typically academic researchers in universities and colleges or research professionals based in independent research and development organizations (e.g., WestEd, EDC, TERC, etc.) or institutions such as science centers and museums. State departments of education or local districts may also be awardees or may participate as partners on grants run by others for the duration of the funded project. For example, state and local educational agencies on their own or as part of consortiums can submit proposals to the EIR program mentioned above. Other federal agencies, including the National Aeronautics and Space Administration (NASA) and National Oceanic and Atmospheric Administration (NOAA), also fund some STEM-related education and engagement projects related to their agency missions using a somewhat different model, as discussed below. (See Table 5-1 for an overview of investments in STEM education by various federal agencies.)

One type of product that typically results from federal grant-funded endeavors is knowledge, such as theory and evidence related to the investigation of a set of research questions about learning and education, or evidence arising from controlled efficacy studies. Knowledge outcomes are typically distributed through publications, presentations, and repositories of various sorts. Funded projects may also result in prototypes or fully developed resources, such as curricula and related materials (e.g., science kits), technology-based learning resources, professional learning programs for teachers, educational programs for students, and real or virtual exhibits. In this overview of the innovation development landscape, we consider these different types of research products separately, since the affordances and challenges for spreading and scaling them are different.

Mobilizing Research Knowledge About STEM Education Innovations

For projects that result in new knowledge to have impacts beyond the direct participants in the funded project, the knowledge that is produced must be disseminated in ways that will reach relevant decision makers and

TABLE 5-1 STEM Investments across Federal Agencies

NOTE: These are the investments that are reported to National Science and Technology Council’s Committee on STEM Education (CoSTEM) and may not fully reflect all institutional programming related to Pre-K–12 STEM education. Column 2 is the total amount awarded. The distinction between columns 3 and 4 lies in the constituency of the primary award recipient. Column 3 includes awards to institutions who may indirectly benefit K–12 students or educators while Column 4 excludes such awardees.

SOURCE: Adapted from White House Office of Science and Technology Policy, https://www.whitehouse.gov/wp-content/uploads/2024/04/2023-CoSTEM-Progress-Report.pdf

practice communities. Individual researchers most often disseminate their findings through peer-reviewed publications and presentations within their own research networks, as is required for academic career advancement, though they may also occasionally reach out to practitioners and policymakers through other channels. Increasingly, research outcomes are also being shared in the form of accessible archives of audio and video data banks (e.g., Databrary.org) and annotated datasets, which can support additional analyses or replication efforts, as well as software code and protocols that can be the foundation for additional research or further development and application by other teams. However, these resources are most often accessed by other researchers, and many individual research teams rely on their own networks and local relationships to craft and implement dissemination plans. Wider project dissemination efforts are sometimes pushed to the end of the grant period, with little time, budget, or incentive to provide extended outreach or to follow up on whether dissemination efforts were successful. At the same time, it is important to note that federally funded research that does not lead directly to implementation in an education setting should not be viewed as a dead end. Weiss and others (e.g., Greene, 1988) have described what they term “enlightenment” or “conceptual use” of research and evaluation findings, a process by which knowledge “percolates into the decision arena in direct and indirect ways” (Weiss, Murphy-Graham, & Birkeland, 2005, p. 14). Many projects have yielded strong scientific evidence related to learning processes, the needs of different groups of learners, and strategies for improving instruction and engagement of students in STEM learning, and this evidence has informed theory and provided a more solid foundation for the design of curriculum, assessments, and instructional practices by other groups. However, the absence of systematic national infrastructure connecting researchers with consumers of the knowledge they produce or with reliable knowledge brokers beyond academic research communities likely means that research knowledge is significantly underutilized (National Academies of Sciences, Engineering, and Medicine [NASEM], 2022).

While many school decision makers and practitioners report that they are interested in learning more about education research and are increasingly required to use scientifically-based evidence to make certain kinds of decisions, studies suggest that how school leaders access and use research is not a straightforward linear process, but one that is influenced by organizational and political factors, available resources, and local dynamics as school leaders seek to justify decisions and persuade interested parties and participants within specific contexts (Asen et al., 2013; Coburn, 2006; Coburn et al., 2009a,b; Finnigan & Daley, 2014; Penuel et al., 2018). The sheer number and variety of specialized peer-reviewed research publications makes it challenging even for full-time academic researchers to keep abreast

of new developments and to synthesize research findings across STEM education. Efforts such as Institute of Education Science’s (IES’s) What Works Clearinghouse (WWC) and practice guides attempt to bridge the gap between education research and practice by consolidating information about successful interventions and the evidence that supports them for educators and decision makers. However, surveys and interviews of district leaders indicate that only a minority of leaders access information from the WWC “often” (13%) or “all the time” (4%; Penuel et al., 2017). Instead, school and district leaders report that they are more likely to access education research through books—especially those that present general frameworks or practical guidance—which are sometimes authored by academic researchers but may also be written by educational consultants (Penuel et al., 2018). These alternative publications are seen as more practitioner-friendly and actionable than peer-reviewed reports of empirical studies.

Why might this be? A recent National Academies consensus study report titled The Future of Research at IES (NASEM, 2022) noted several factors that may contribute to the difficulty of mobilizing knowledge that results from federally funded education research. First, there may be a mismatch between the types of problems that school decision makers are attempting to address and the structure of the research questions or interventions being investigated in grant-funded projects. In addition, researchers are often primarily concerned with finding reliable average differences between well-defined groups assigned to different conditions in a research design, and they may use local or regional samples of participants and settings that are circumscribed or non-representative in various ways (e.g., students with learning differences or limited English proficiency may be excluded). Emphasis on research designs that can support causal conclusions may overshadow the study of context variables and adaptation to local contexts (Means & Penuel, 2005). Education leaders, in contrast, are typically concerned with the constraints and complexities of their specific contexts and may be attempting to address the learning needs of a highly varied or particular student population; thus, they may be skeptical about the degree to which findings of individual research studies pertain to or could be realistically implemented in their situation. Although IES seeks to support research that informs not just “what works” on average but also “for whom” and “under what conditions” (Spybrook et al., 2020), such research is more difficult and expensive to conduct and hence less common.

Developing and Disseminating Promising Evidence-Based Resources for STEM Education

Research-based innovations in STEM education typically involve more than just new knowledge gained from studies: they also involve resources

in the form of new instructional materials, interventions, training programs, and tools. This kind of research can vary considerably in the degree to which the implementations being studied are carried out in normal educational settings or under more controlled conditions, but, most commonly, research teams running federally funded studies are heavily involved in the implementation, potentially conflicting factors are controlled, and adequate resources and support are provided using grant funding. Preparing to scale an innovation beyond the research phase—especially one that requires significant changes in practice, organization, or resources—is both very difficult and often overlooked or neglected. As discussed in Chapter 4, it is most likely to occur successfully if the designers of research-based tools work collaboratively and iteratively with practitioners to figure out how to best enact the intended innovation in realistic contexts. This may present a steep learning curve for many academic researchers, who may be unfamiliar with the practices, routines, and constraints of schools and classrooms. Research teams that have been successful in scaling up ambitious learning innovations have typically begun planning for scaling early in their work and have invested long periods of time working collaboratively in intended implementation settings and incorporating what they learn there (Roschelle, Mazziotti & Means, 2021). As McLaughlin and Mitra (2001) noted in their study of how theory-based educational change reforms are sustained and extended beyond their initial invention, “Moving from ‘vitro’ to ‘vivo’ requires the collaboration of reformers and practitioners and assumes that the reform itself is unfinished until that co-construction takes place” (p. 306).

For promising innovations to spread beyond the contexts in which they were first developed, there also have to be pathways for these types of products to be distributed in usable forms to schools. Physical materials (such as science kits) require a manufacturing and distribution system. Other innovations may rely on professional development for teachers to implement them effectively, which requires project staff to create and implement training programs on an ongoing basis. Although it is attractive to think that resources that were developed with federal funding might be widely disseminated at little or no cost, the reality is considerably more complex. Relatively few potent interventions can truly be given away with no additional support. Even online electronic resources have to be maintained and updated on secure websites, with features such as user registration systems, secure data transfer protocols, operating system updates, and interoperability across different platforms—all of which entail ongoing costs for professional services that persist after the original grant funding has expired. Unlike commercial developers, recipients of federal research grants who create new educational products and resources often lack the

expertise to market, distribute, and support products on their own. While innovation developers may wish to see their products disseminated and sustained, accruing the time and resources to do this themselves may be an especially uncomfortable fit with the constraints and incentives of academic jobs and may conflict with the demands of publishing research and pursuing new grant funding. Research labs in academic institutions are ill-suited to selling and distributing products, and academic designers may run into problems or resistance if they try to market what they have developed. They may face scrutiny from their own institution about conflicts of interest or time commitments, or they may be reluctant to publish their research with disclosures of a financial interest.

In recent decades, there has been a move toward open educational resources (OER) shared through various platforms that post materials that have been placed in the public domain or that have been released under licenses that permit various combinations of access, duplication, redistribution, adaptation, and repurposing. Creative Commons (creativecommons. org), for instance, is an international nonprofit organization that provides a variety of licenses that have been widely used to share resources at no cost for educational purposes. Materials shared in this way are typically digital resources that can be easily uploaded and downloaded through web browsers. Because there are few barriers to sharing resources in this way, OER can be an attractive mechanism for facilitating dissemination and spread. However, this model does not solve the problems of generating revenue to support resources more robustly, providing ways to track field-based revisions, or assuring that either original resources or modified versions of them are of high quality.

Another challenge in developing and disseminating evidence is that even innovations that have strong evidence demonstrating their potential to improve student learning and engagement may not attract the interest of school districts if their focus is narrow and doesn’t cover a sufficiently wide grade band or a large enough portion of the curriculum. School districts may prefer complete packages provided by commercial publishers that include elements such as teacher and student dashboards that integrate with learning management systems commonly used in schools, ready-made standardized assessments, and professional development for teachers.

Products developed with federal research funding can, of course, be transferred to for-profit publishers. While this might have the benefit of harnessing professional sales and marketing capacities, there is also a risk that innovations may be converted into products that are not faithful to what was originally developed and studied; another risk is that they may be acquired but not distributed. Formerly accessible materials may also

disappear behind paywalls. While fields such as medicine and engineering have a history of commercializing products that result from research, the norms and processes are less familiar in education research, and standardized business models are lacking (Kathy Perkins, personal communication, December 7, 2023). Nevertheless, as we will see later in this chapter, some innovation developers have successfully found ways to disseminate, scale, and sustain their products.

Some federal agencies, such as NASA and NOAA, also provide funding for STEM education, but their purview is more specific to their agencies’ mission, in contrast to the broad mandates carried out by NSF and the U.S. Department of Education. This allows these agencies to both develop educational programs and resources and also to implement and disseminate them widely in a more integrated fashion. The Smithsonian Institution has also received funding via an IMD grant to successfully disseminate, scale, and sustain products. In these cases, the same program, under a single large umbrella, can develop educational resources and also form partnerships to train and support large networks of practitioners using those resources, with opportunities for collaborative and iterative development over long-time spans. Two examples of such initiatives, NASA’s Science Activation program (NASEM, 2020) and the GLOBE Program are described in more detail below. This structure circumvents some of the gaps and misalignments that arise when the development and implementation landscapes for educational innovations are not connected.

Finally, it is important to note that federal investments in STEM education have often extended beyond the formal Pre-K–12 education system. Innovation and impact extension in STEM that takes place during out-of-school time (OST) settings have also been supported in a number of ways over the years. Government investment by many federal agencies including NSF, NASA, NOAA, and ED has supported the development and expansion of projects in science media, informal science institutions, and community serving organizations. Investments by ED and NSF in public television programming with a math and science focus, including early childhood programs, have provided broad access to learners of all ages across the nation. Focused NSF support has also provided the fuel for the expansion of science centers and experiences with new science, such as the nanotechnology-based Nanoscale Informal Science Education Network. NOAA’s Environmental Literacy Program provides grants and in-kind support to the informal science education (ISE) world to improve ecosystem stewardship. NOAA has also developed Science On a Sphere, a global display system to visualize planetary data in support of earth science education. NASA, through its Science Activation program (described in more

detail below), develops partnerships between the agency and members of the ISE field to support youth participation in their missions of discovery and to build a connection with authentic science through citizen science activities. Collaborations between the formal and informal science systems have also been supported at the federal level, including museum-based standards-aligned curriculum like Smithsonian Science for the Classroom and accompanying professional learning for teachers, kit-based school curriculum like FOSS, and family support for getting ready for school. These collaborative efforts among school educators, informal science institutions, community organizations, and the federal science agencies demonstrate the impact of collective action.

Innovations Arising from Sources Other Than Federal Funding

Innovative STEM programs and resources also arise from sources other than federally funded initiatives. Private corporations, museums and science centers, professional organizations, foundation and corporate philanthropy, groups of teachers, and local community groups have all contributed to enriching and enlivening opportunities for engaging children and youth in STEM and providing supports and resources for STEM education in both the formal Pre-K–12 system and in the OST sector. For example, in recent years, private funders have been active in creating new programs and resources for computing education, both in and out of school, as described in Box 5-1.

In some cases, private funding supports expanded delivery of a particular program through sponsorship and marketing. In other cases, this funding directly supports innovation efforts through investments in research, evaluation, and scale-up activities. Compared to innovations that are developed through competitive federally funded awards, these alternative development pathways can have both advantages and disadvantages. To the extent that their origin is more “bottom up,” these innovations may be closer to the intended implementation environment from the outset and may have local or regional champions who gather resources and build collaborations to sustain them. They may also be more specifically tuned to the needs and goals of a particular community. Another significant benefit is that corporate or philanthropic funding can often be secured quickly, in contrast to federal funding, where a solicitation may only accept applications once per year and the timeline from submission to the start of a funded project may take a year or more. However, if the funding does not include a commitment of resources for research and evaluation, the innovation may lack evidence of educational effectiveness beyond simple counts of numbers of events and participants.

THE INNOVATION IMPLEMENTATION LANDSCAPE

In contrast to the innovation development landscape, where federal policy and funding play a predominant role in concert with designers situated primarily outside of the Pre-K–12 public education system, the innovation implementation landscape for STEM education in the United States involves actors, financial resources, and decision-making processes that are located within the public education system, mostly at the state and local levels. There are relatively few direct routes guiding the flow of innovations arising from federally funded projects to schools and districts throughout the country. This situation is somewhat akin to the state of travel in the United States before the interstate road system was established: some schools and districts—such as large city districts, districts that have connections with research universities, or districts in states that develop their own initiatives—sit on well-traveled thoroughfares with ready access to new ideas and resources, while others are only reached by circuitous, lightly traveled routes and have little capacity to investigate or invest in education innovations.

Implementing Innovation: Key Considerations

As discussed in Chapters 2 and 3, the United States has a long history of state and local control in education policy and decision making. Leaders at state, regional, and local levels who identify, vet, and implement innovations in educational settings have distinct sets of concerns that may be quite different from the concerns of innovation designers. As they consider a potential innovation, they may ask whether it addresses an important need or issue; whether it is responsive to the local population and context; what resources are required to implement it and whether those resources can be easily acquired; whether additional staff development will be required; whether it will be welcomed or resisted by staff, students, and the local community; whether it is compatible with or potentially in competition with or disruptive to existing curriculum and routines; what timeline is required for implementation; and how they will know whether it is achieving intended outcomes.

The degree to which a particular innovation is likely to spread and endure depends not just on research data of effectiveness, but also on the extent to which it can provide answers and solutions to these questions of practical implementation. Some innovations are narrowly targeted and can be implemented fairly readily with existing resources under standard conditions in the educational setting. For example, a classroom teacher may decide to incorporate an online tool for math homework assignments and formative assessments that is compatible with curriculum already in use

and can be easily integrated with the school’s existing learning management system (e.g., ASSISTments).¹ A more ambitious or unconventional innovation, in contrast, might require a lengthy and carefully planned campaign to gather resources, develop a team with appropriate skills, build community support, and address challenges such as dissent or resistance to change.

NEGOTIATING THE TERRAIN BETWEEN THE LANDSCAPES OF INNOVATION DEVELOPMENT AND IMPLEMENTATION

As noted previously, many funded projects are not spread beyond their initial instantiation, and if they are spread, are not sustained. Although they may generate important research evidence of impacts on STEM learning or other desired outcomes in their original context, they lack a functional dissemination model that reaches directly into classrooms or other educational settings. Since 2002, IES has funded types of projects (originally referred to as “goals”) that suggest an implicit linear progression from initial development of innovations to focused efficacy studies and then effectiveness or replication studies at larger scales, with the latter goals encouraging studies with new and larger populations under typical rather than ideal circumstances. However, evidence suggests that relatively few innovations progress through these project types in a connected way, and only a minority of grants are associated with future grants (NASEM, 2022). As a result, many research-based innovations may be very promising and have evidence of effectiveness, but they have not completed the process to be fully developed into forms that can be readily implemented in realistic educational settings with diverse populations of learners. In recent years, as discussed in Chapter 4, there has been increasing interest in STEM education in a contrasting nonlinear perspective, in which decision makers and practitioners are directly involved with researchers in designing, adapting, and scaling innovations (Roschelle, Mazziotti & Means, 2021).

Where innovations have gone through a complete development process and are ready to be implemented more widely, the question of how they are promoted or discovered comes into play. One potent route for evidence-based instructional resources is to become part of state adoption cycles. Historically, local school district leaders have had the authority to select and purchase materials, in line with the U.S. tradition of local control, but a growing number of states have developed policies to evaluate and adopt or approve materials at the state level, with the goal of ensuring consistent access to high-quality instructional materials (HQIM) throughout the state (Doan & Kaufman, 2024). These efforts are also in line with the federal

___________________

¹ See https://new.assistments.org/

Every Student Succeeds Act, which encourages districts and schools to choose interventions that have been rigorously studied and shown to improve student outcomes. Boards in states with adoption policies periodically review materials and evidence of their effectiveness in various subject areas and publish lists of approved materials, which may be either mandated, such that districts must adopt them or apply for a waiver or recommended but not mandated. States with adoption policies may further incentivize the use of selected materials by providing aligned professional development or negotiating master contracts with publishers to make it easier and more affordable for districts to acquire them.² State education agencies also receive both federal and private grants, some of which may then be awarded to LEAs to support HQIM and teacher learning. Whether mandated and incentivized or not, adoption lists are a direct way for states to call attention to high-quality materials and save districts the time and money required to carry out their own reviews or figure out procurement strategies. In 2017, the Council of Chief State School Officers formed the High-Quality Instructional Materials and Professional Development (IMPD) Network,³ which currently includes 13 states and helps states develop policy, communications strategies, and professional learning opportunities aligned with HQIM. Some adoption states also partner with EdReports, which is an independent organization that conducts regular reviews of curriculum materials with national teams of educators (Doan & Kaufman, 2024).

Evidence suggests that state adoption policies increase use of standards-aligned HQIM to some degree. For example, the American Instructional Resources Survey Project, conducted independently by the RAND Corporations, has found that adoption and usage of HQIM was significantly higher for teachers in states in the IMPD network (Doan et al., 2022). State adoption policies can be viewed as an example of what Weiss, Murphy-Graham, and Birkeland (2005) have described as “imposed use” of research and evaluation evidence, where a higher level of government mandates or incentivizes specific actions based on evidence. Local districts may vary in whether they are responding to the evidence itself or to the mandate or incentives, but this is a pathway of influence that appears to be effective and can allow federal, state, and local levels to negotiate varying degrees of control and decision making.

Some adoption policies are limited to particular subjects (especially English language arts and mathematics) or grade levels (e.g., reading materials for early elementary grades), and so state adoption may not be an

___________________

² The federal government has also provided funding support: partially in response to the shocks of pandemic schooling, emergency funding through Elementary and Secondary School Emergency Relief included funding for states to advance the adoption of mathematics and English language arts HQIM.

³ See https://learning.ccsso.org/high-quality-instructional-materials

option for many STEM innovations. However, professional organizations and other non-profit organizations are another route by which high quality, standards-aligned STEM materials and programs can be highlighted and promoted. Organizations such as the National Science Teachers Association, the National Council of Teachers of Mathematics, and the Council of State Science Supervisors routinely call attention to instructional materials and programs through a variety of channels that reach educators, including their websites, regional and national conferences, journals, reports, and online seminars.

Though state approval is one route to widespread dissemination of innovations, many innovation designers are left to develop their own pathways to spread and sustain the resources they have created. Although we have highlighted many of the challenges of trying to move innovations beyond the environments in which they are initially developed, the committee also documented a number of cases in which designers have been successful in spreading or sustaining their own innovations, or other parties have sought out effective innovations and worked to bring them to more schools and students. A notable finding is that the pathways by which they have done so are almost as varied and numerous as the innovations themselves. In this section, we feature a series of examples, drawn from both programs in the commissioned compendium and from others documented in published research, to illustrate this variety. At the same time, we note that they tend to share some common features, such as ensuring that their interventions and programs address important needs in educational settings, mobilizing and supporting both enactors and facilitators, and partnering to build necessary capacity and infrastructure (Cohen & Mehta, 2017; McLaughlin & Mitra, 2001).

Innovations Starting with Research & Development

We first consider several innovations that started with funding for research and development, accumulated evidence of their effectiveness, and—either simultaneously or successively—have engaged in significant efforts to extend or sustain their innovation. These tend to fall along a continuum of scaffolding for sustainability. Some innovations are able to develop their own distribution model that enables them to stay active over time and to successfully spread the innovation to more users or further develop it in new contexts or with new applications; however, they continue to struggle with achieving a sustainable financial model and are at risk of not remaining viable over a longer time. Some also thrive under the direction and leadership of a founder or key advocate but struggle to create a succession plan for maintaining the innovation when that leader is

no longer involved. PhET Interactive Simulations is an instructive example of a program that has been active for more than 20 years and has been impressively successful in spreading worldwide but that finds it challenging to maintain a sustainable business model that can support the full potential of its innovative resources.

PhET Interactive Simulations

PhET has created 169 open-source simulations to support sustained inquiry learning in math and science across multiple topics and grade levels from elementary through higher education. These simulations have been used more than 1.4 billion times around the world, with about half of that usage in the United States and with a notable increase during and following the COVID-19 pandemic. Development has been ongoing since the early 2000’s, with support from numerous NSF grants as well as private foundation money. Multiple published research studies indicate that PhET simulations lead to significantly higher learning gains than traditional instruction (Antonio & Castro, 2023; Banda & Nzabahimana, 2021) with substantial effect sizes. However, federal grant funding in the United States has only supported new research and development, with no funding for scaling or sustainability. In recent years, PhET has built a scalable teacher professional development model and a teacher-leader network, but this was done with funding available in Latin America and Africa, not the United States PhET has followed several strategies to make the simulations more available, including building them in such a way that they can be embedded in other education products (e.g., those produced by commercial publishers). Although the PhET website is free, the program’s developers have also tried to develop some revenue sources to build a more sustainable financial model, with modest revenues coming from individual donations from teachers, from a low-cost direct to consumer app, and from business-to-business partnerships. These, however, account for less than 15 percent of annual expenses. Despite their notable success in scaling and sustaining PhET simulations, the developers call attention to several intractable issues. Their open education resources—which they believe are too expensive and require too much specialized knowledge to have been produced by the private sector—do not have a compatible business model that generates sustainable revenue. They lack the funding to maintain and update simulations as technology changes (most notably the disruptive transition from Java and Flash to HTML, which has resulted in mothballing dozens of simulations with proven efficacy). Finally, while they have completed research on how to effectively implement inclusive design features in simulations to serve students with special needs and learning differences, they do not have funding to scale these features.

Innovations with a Robust Distribution System

A bit further along this continuum are evidence-based innovations that have developed a robust distribution system that is financially sustainable over a long term. These typically involve having a home in or partnership with a stable organization that can handle sales, marketing, distribution, and support on an ongoing basis. The first case highlights Youth Engineering Solutions (YES) and Engineering is Elementary (EiE), engineering curricula for grades Pre-K through 8, which were originally developed by academic researchers with funding from NSF and are now distributed through the Museum of Science, Boston, with additional support from industry and foundation partners. The second case is the Building Blocks Pre-K Math curriculum, which was developed by Douglas Clements and Julie Sarama with funding from NSF and has been extensively studied during scale up using the TRIAD scale up model, also developed by the same research team.

Youth Engineering Solutions (YES) and Engineering Is Elementary (EiE)

YES and EiE are innovative curricular resources designed to introduce engineering in Pre-K–8 classrooms and out-of-school settings, such as afterschool and camp programs. Initially supported with grant funding, including several NSF grants, the designers of YES/EiE work closely with teachers in design-based development and research cycles to create materials that can be used effectively in typical classroom environments. The YES/EiE approach situates problems in societal contexts and invites students to think critically about the impact of their engineered solutions. It also uses flexible, multimodal activities to encourage all students to generate original solutions to problems, to manipulate physical objects to deepen their conceptual knowledge, and to persist and learn from failure. The programs are accompanied by professional learning resources attuned to teachers’ needs, with the understanding that engineering is an unfamiliar discipline for most preschool and elementary teachers.

Because engineering is a new discipline in Pre-K–12 education, it often faces barriers to finding a regular place in school curricula. Clearing instructional time for engineering during the school day, making time for teachers to participate in professional learning opportunities, and developing a system to provide and replenish physical materials have been persistent challenges. To facilitate uptake, the YES/EiE development team has aligned classroom-based materials with state standards and Next Generation Science Standards and also worked with researchers to demonstrate impacts on desirable student outcomes, including engineering and science knowledge, interest, and attitudes. The curricula are also used in out-of-school programs, where there is often more freedom and flexibility to focus on engineering. There have been over 60 published conference papers and

journal articles that the team has written about YES/EiE and over 40 others written by scholars outside the team. These includes measures of fidelity of implementation (Lachapelle & Cunningham, 2019), a randomized controlled trial to examine student outcomes (Cunningham et al., 2020), and studies to examine the influence on different populations of learners including English language learners (Cunningham, Kelly, & Meyer, 2021), and low-income students (Robinson et al., 2018).

YES/EiE resources are used nationwide and reach over five million youth a year. The Museum of Science, Boston is the home organization for the resources and has provided some financial support as well as organizational stability over the years. Digital resources, which were originally provided for sale, are now available for free download, and kits and print materials are available for purchase. External grants from corporate, federal, and individual sources fund the development of new resources and the purchase of materials and professional learning for schools and districts. This combination of financial resources has been stable and is expected to be sustainable.

Building Blocks™ Pre-K Math and the TRIAD Scale-up Model

The work of Julie Sarama and Douglas Clements is notable in having devoted attention to (a) the development of theory and evidence about children’s learning trajectories in mathematics (Clements & Sarama, 2004), (b) the creation of an extended curriculum for preschool children aligned with progressions of how children learn mathematics, known as the Building Blocks™ curriculum (Clements et al., 2011), and (c) development of a comprehensive, generalizable model for scaling up evidence-based practices and programs known as Technology Enhanced Research-based Instruction, Assessment and professional Development (TRIAD; Sarama et al., 2008; Sarama & Clements, 2013). Their work on learning trajectories involves theorizing and testing descriptions of the progression of children’s thinking and learning within a domain (mathematics, in this case) and then developing a set of instructional tasks that facilitate pathways through the trajectory (Clements & Sarama, 2004). With support from NSF through multiple grants in the last two decades, Sarama and Clements co-evolved the curriculum and the TRIAD scale-up model and conducted multiple research studies, including large randomized controlled trials with thousands of children in hundreds of preschool classrooms across several states.

Sarama and Clements define scale-up “as the instantiation of an educational intervention in varied settings with diverse populations, addressing the needs of multiple sociopolitical stakeholders, so as to achieve (1) satisfactory fidelity of implementation and, as a result, (2) the intervention’s goals for over 90% of the children who could benefit from the intervention and ultimately

(3) eventual transfer of the intervention to local ownership, sustainability, persistence of effects, and continuing diffusion” (2013, p. 174). The TRIAD scale-up model includes a set of research-based guidelines for working with teachers and administrators, children, families, and communities to achieve scale-up that successfully fulfills this definition. These guidelines include planning for the long term and preparing for increasing complexity as the intervention scales; continuously communicating and building commitment to a shared vision among administrators, teachers, and families; promoting equity among all participants, including in the allocation of resources; and providing in-depth, ongoing professional learning opportunities and coaching situated in the classroom with a focus on understanding developmental progressions in children’s mathematical thinking and their pedagogical application in the curricular activities and assessments in the Building Blocks™ curriculum. Their approach anticipates the need for schools and teachers to make adaptations as they increasingly take ownership of the program but supports them in maintaining the integrity of the intervention by helping them learn to distinguish productive adaptations from those that would dilute or undermine it.

The full Building Blocks™ PreK Math, including manipulative kits, digital activities, and assessment and resource guides for teachers (e.g., printable masters and family letters) is now available for purchase through McGraw Hill. Professional learning resources for teachers are offered in the form of an online course, on-demand video and other resources, and live training sessions.

Standing Program Provides Larger Umbrella

As mentioned above, federal funding, particularly through NASA and NOAA, has provided support for STEM learning using a somewhat different model, in which a standing program provides a larger umbrella with internal infrastructure and longer funding horizons to enable the iterative development of a set of shared resources and to develop robust collaborations to disseminate and implement those resources. The examples below present two cases that exemplify this approach: the Science Activation (SciAct) Program and the GLOBE Program, both of which are based at NASA.

NASA Science Activation Program

The NASA Science Mission Directorate’s (SMD) SciAct “connects diverse learners of all ages with science in ways that activate minds and promote a deeper understanding of our world and beyond.” Started in 2015, the Science Activation program is a cooperative network of competitively selected teams from across the Nation working with NASA infrastructure activities to share NASA science with learners of all ages” (NASEM, 2020).

The program has its own office and program team embedded within NASA’s SMD, and the infrastructure provided by NASA connects these teams to scientific expertise and a large library of NASA resources. SciAct uses a “collective impact” model, with NASA’s SciAct team serving as a “backbone” to coordinate organizations and activities that share a common interest in space science and an agenda for promoting accurate and compelling learning experiences.

In 2023, SciAct facilitated almost 76 million learner interactions in the United States and across the globe (National Aeronautics and Space Administration, 2023). The program also supports data collection and publication of peer-reviewed reports. Key to SciAct is the integration of NASA science mission experts and internal education production resources, such as scientific visualization laboratories, NASATV and NASA.gov, with the development and distribution capabilities of the external education experience providers and networks, such as museums, libraries, and media providers. The successful integration effort requires monthly and annual meetings attended by members of all of the teams, annual identification of cross collaboration opportunities, and whole group education opportunities on issues such as diversity, citizen science, and mission science. SciAct features cooperative agreements of up to five years with the potential for renewal, which provides the time necessary to create a thriving community of collaboration. The program bridges the landscape of innovation and implementation through the careful cultivation of a portfolio of diverse projects, partnerships and dissemination opportunities (NASEM, 2020).

GLOBE Program

First announced by the U.S. government on Earth Day in 1994, GLOBE is an international inquiry-based education program focused on environmental science and earth system science. GLOBE, which stands for Global Learning and Observation to Benefit the Environment, aims to engage students in scientific activities, improve science education, and increase environmental awareness and understanding of the Earth as a system. GLOBE operates in 127 countries, with around 41,000 schools and 50,000 educators involved historically. There are about 278,000 GLOBE observers who engage with the program.

GLOBE aims to increase STEM literacy, particularly in earth system science, environmental science, and climate and resilience science, through purposeful active learning, community engagement, and building a community of practice. GLOBE’s theory of change is engaging people in taking environmental observations and measurements, analyzing the data, and optionally putting it into GLOBE’s open database. The backbone of the GLOBE Program is a set of 40+ measurement protocols. Initially,

the program was scientist-driven, with scientists designing protocols for students to collect data. Over time, it has evolved to enable more student-driven investigations using GLOBE tools, with scientists providing context and motivation.

GLOBE operates in coordination with U.S. federal agencies like NASA (which is now the lead agency), NOAA (which was the former lead), NSF, and the Department of State, all of whom have goals aligned with GLOBE’s mission around STEM education, environmental literacy, and international collaboration. The core GLOBE infrastructure is primarily funded through federal appropriations to NASA. This supports the GLOBE Implementation Office for training, materials, and community engagement, as well as a data management system. Although securing funding commitments and maintaining interagency agreements has been bureaucratically challenging, the longstanding support for GLOBE’s core infrastructure helps explain its longevity and widespread expansion.

Partners obtain additional funding from various sources such as federal agencies, school districts, philanthropies, and international contributions based on bilateral agreements. In U.S. schools, factors such as school administration support, alignment with local standards, and availability of partner support have been persistent challenges, as detailed in program evaluations. Program leaders continue to explore opportunities to secure additional funding commitments from federal agencies beyond in-kind support, as well as other sources, to support their scaling and sustainability efforts.

Independent yearly evaluations of the GLOBE Program were conducted during the first ten years of the program. These evaluations included evidence of GLOBE participants achieving STEM learning goals and skills, and many other aspects of GLOBE implementation and impact, including how implementation varied in different contexts. Since the initial evaluation, many projects based on GLOBE have also included an evaluation component. Going forward, the project would like to enhance program evaluations to further understand factors impacting domestic scale and sustainability, such as teacher turnover and instability in school administration support (Amy Chen, personal communication, April 26, 2024).

Building Durable Systems

The examples discussed so far represent cases in which a distribution system was purposely built to drive a given innovation or a collection of related innovations from development through widespread implementation. While the system allows the targeted resources to evolve and expand in various ways over time, the system is not intended to serve other unrelated innovations or to support STEM education more generally. A different

approach is to build “durable systems” with the capacity to identify, implement, and support multiple programs and resources so that unrelated STEM education innovations can successfully reach and serve the intended constituents and communities. These systems will be discussed further in Chapter 8, but here, we provide a brief introduction by way of several examples, which vary considerably in how they came into being, who the main partners are, and how they operate to connect participants, supporters, and resources in the service of STEM education.

Regional Educational Laboratories

The first example is Regional Educational Laboratories (RELs), which are federally funded through ED to disseminate research evidence and build resources to support the application of high-quality research in educational settings. Organized around geographic regions, RELs are one of the few examples of a federally funded pathway that directly connects national research investments with their application and scaling across all levels of the education system.

IES funds a network of ten RELs (see Box 2-4 in Chapter 2) covering American states and territories. RELs partner with educators and policymakers in their regions to support local, regional, or state decisions about educational policies, programs, and practices, with an emphasis on the use of evidence and data to improve student outcomes. Their purview includes but is not limited to STEM education. The authorizing legislation for the REL Program stipulates three main activities:

• Applied Research and Development: RELs partner with districts and states to identify high-priority needs and to develop peer-reviewed research products intended to be actionable for teachers and educational leaders in schools. They also develop other products to support scaling up of best practices, such as toolkits, with accompanying professional development programs.
• Training, Coaching and Technical Support for Use of Research: RELs provide a variety of professional learning opportunities, including intensive professional development focused on important topics; coaching to apply research evidence to inform critical decisions; and technical support to build capacity in collecting, analyzing, and utilizing data.
• Dissemination: RELs are charged with synthesizing and communicating research and evidence so that high-quality, scientifically valid information is readily available and provided to educators and policymakers in ways that support action and application in local contexts.

These activities are intentionally designed to complement other federal investments in research and development in education and to create functional bridges between research and practice, with the goal of translating research for decision makers in schools and building tools and professional capacity to leverage evidence-based improvements in education.

RELs are involved in a number of specific projects related to STEM teaching and learning, but their approach is illustrated by two current mathematics projects. The Teaching Fractions Toolkit Partnership, run by REL Midwest with the Illinois State Board of Education and school districts in Illinois, focuses on strengthening 6th grade students’ conceptual understanding of fractions and narrowing gaps in math achievement among student groups. The Toolkit includes intensive professional development and supports for teachers to understand and apply evidence-based practices for teaching fractions and related concepts. It also works with math leaders and school and district administrators to develop wraparound supports to reinforce the broader developmental trajectory of fraction understanding across grades Pre-K–8 in complex school systems.

REL Northeast is partnering with district leadership teams from a group of districts in Connecticut to improve mathematics outcomes for multilingual learners, who typically score low on state assessments. One part of this project is building capacity among district leaders to support English learners and multilingual learners through a workshop series in which participants create district logic models, analyze district data, learn about evidence-based practices, and create action plans. A second component is a research study on the impacts of a related teacher professional development program using an intervention called Visual Access to Mathematics, which helps teachers explore how visual representations can support problem solving and communication for both English learners and other students learning critical middle grades content.

As they work to develop resources to help scale research-based practices and resources, REL projects typically undergo usability and feasibility testing as their components are iteratively developed and then later test them in formal studies to examine their summative impact. RELs also engage in cross-REL activities to further disseminate and scale high-quality programs and products. The REL program has had notable longevity since the first RELs were authorized and established in the mid-1960s as a network of large-scale labs. Although the number of RELs and their priority areas have changed periodically through the program’s history, their focus on connecting high-quality research with practice and on creating strong, long-term partnerships and networks to support scaling and impact has been continuous.

State STEM Ecosystems

A second type of durable system designed to support the scaling of multiple STEM innovations is state STEM ecosystems (see Chapter 2 for a brief discussion of STEM learning ecosystems in relation to regional actors). Many states (or regions within states) now have some form of a STEM ecosystem.⁴ The committee met with a panel of representatives from four such state-level STEM organizations: Jeff Weld from the Iowa Governor’s STEM Advisory Council, Lee Meadows from the Alabama STEM Council, Michael Vargas from the Arizona STEM Acceleration Project, and Jeremy Babendure from the Arizona-based SciTech Institute. Although these groups vary in the specifics of how they are organized, where their funding comes from, and what particular initiatives they support, they share an ecosystems approach to building long-term regional partnerships to call attention to STEM education and to aggregate and channel resources to support it.

Many state STEM ecosystem initiatives have been motivated by local businesses and industries concerned with STEM workforce development, including the development of competencies such as critical thinking, problem solving, communication, and facility with technology, and they have typically been launched with support from the governor’s office or the state legislature. In Iowa, for instance, a group of people from industry, higher education, and the K–12 system gathered on their own over several weekends and drafted a STEM education roadmap with a short set of recommendations to policy leaders about how to cultivate a future workforce in the state with strong STEM capabilities. Their plan gained the support of the governor, who created the Iowa STEM Advisory Council by executive order in 2011, and then garnered an initial appropriation of $4.7 million from the legislature in 2012. In Alabama, advocates similarly developed a STEM education strategic plan rooted in workforce development that was presented to the governor, who formed the Alabama STEM Council by executive order in 2020 with financial support from the legislature. State STEM ecosystems leverage these public investments by securing contributions from business and industry partners and grants from foundations and funding agencies. Their ability to create long-lasting partnerships and to raise flexible funds on a consistent basis allows them to grow and maintain efforts over a longer time horizon than efforts that depend on one-time project funding. However, leaders also note that they bear a heavy and continuous burden of fund-raising, and funding streams are not consistently reliable.

___________________

⁴ See https://stemecosystems.org/. Note that this is not an exhaustive list. Other empirically grounded ecosystems exist, such as the Chicago City of Learning initiative (Pinkard, 2019; Quigley et al., 2016) and the Synergies project (Falk et al., 2015, 2016a,b).

Central to the notion of a state STEM ecosystem is the development of regional networks and partnerships that engage state and local organizations in the support of STEM education activities, events, and infrastructure development, both in and out of schools. They also connect teachers and educational leaders with each other, which helps ideas and resources flow to districts throughout the state, including those that are chronically underresourced and understaffed in STEM-related areas. For example, Arizona’s SciTech Institute uses a dual strategy of creating regional hubs, which connect local resources and foster collaboration, and also supporting statewide working groups, which can share expertise and develop collaborative projects focused on high-priority needs and challenges across the state.

State STEM ecosystems have supported a great variety of programs and initiatives, including professional learning opportunities and externships for teachers; small competitive grants to teachers and districts to develop specific projects; out-of-school events and programs for youth; STEM ambassador programs for students; and STEM festivals for the general public. As just one example, the Arizona STEM Acceleration Project has supported more than 450 teachers from every corner of the state as STEM Fellows, who receive a stipend of $4,500 plus $2,000 to purchase STEM supplies. Fellows complete at least 30 hours of STEM professional development and also submit four lesson plans, which are reviewed and then added to a repository that is available to other teachers statewide. State ecosystems have also targeted specific critical needs in their respective states. For example, in order to address the state’s critical shortage of STEM teachers, the Alabama STEM Council has created partnerships and obtained annual funding from the legislature to launch new UTeach STEM teacher preparation programs at six universities in Alabama, with the goal of preparing 250 new STEM teachers every year.

A noteworthy feature common to many state STEM ecosystems is that they have embraced the need to develop sophisticated communication strategies about the importance of STEM and the value of providing engaging and effective STEM learning opportunities to children and youth across the state. They produce blogs, podcasts, and radio and television ads that showcase students actively pursuing their STEM-related aspirations to raise awareness among the general public and to drive support for STEM education.

State STEM ecosystems vary in their approaches to measuring and documenting their outcomes. Most regularly compile statistics on participation in their programs and also conduct analyses of return on public investments. They may also look for gains in students’ interest in STEM and persistence in STEM education and career pathways, general gains in

academic performance on state-wide math and science tests, and public support for STEM education.

STEM NOLA

A third example, STEM NOLA, illustrates a markedly different approach to building a durable system to support STEM education. In contrast to many efforts that attempt to import STEM learning opportunities from more resourced to less resourced communities, STEM NOLA has grown a thriving STEM ecosystem from the ground up, using local leadership, partnerships, and resources to create a dynamic suite of programs in neighborhoods in New Orleans that have traditionally struggled with weak schools and lack of opportunity. In an unusual turn of events, the quality and popularity of these programs now draw participants from more affluent communities and have prompted families and educators to work with STEM NOLA to upgrade STEM education in the public schools.

Launched in New Orleans in 2013 under the leadership of Calvin Mackie, STEM NOLA is a community-based STEM learning ecosystem that designs and delivers a variety of hands-on learning activities, events, and programs with an emphasis on engaging and serving youth in underserved neighborhoods. Its history stands in marked contrast to initiatives that originated with grant funding to STEM education researchers and professional developers. Dr. Mackie, who had previously been a tenured professor of mechanical engineering at Tulane University, and his wife Tracy, who was a trained pharmacist, saw that STEM education in local schools was inadequate and uninspiring, with the result that students were being turned off rather than engaged by STEM, risking a tremendous loss of talent. Initially working out of their garage, the Mackies built a robust learning ecosystem from the ground up, engaging local STEM professionals and college students as volunteers and interns to provide a full menu of hands-on learning opportunities in local neighborhoods during out-of-school time. STEM NOLA focuses on children and youth, with the goals of laying an early foundation of STEM exposure and inspiration, building trust-based intergenerational relationships with local STEM mentors and role models, and providing connections to help young people make sense of and advance through education and career pathways in STEM. Their flagship program has been STEM Saturdays, where familiar local facilities are transformed into temporary laboratories with a series of hands-on activity stations staffed by at least one college student and STEM professional volunteer and a maximum of four students. Activities, which are divided into four grade bands, are related to the theme or topic of the day and selected to be engaging and culturally and environmentally relevant. Each activity culminates with students building or

making something, such as functional catapults, rocket cars, circuits, batteries, levees, and motorboats. STEM NOLA has also branched into other formats, including summer camps, larger-scale STEM Fests, and, more recently, providing kits, curriculum, and teacher professional learning opportunities for school-based STEM.

The program’s community engagement model has generated a remarkable response: since its inception, STEM NOLA has engaged over 130,000 Pre-K–12 students and over 25,000 families, 87 percent of whom qualify for free and reduced lunch programs, as well as more than 3,500 STEM professionals and trained student volunteers. It has attracted more than $20 million of financial and partnership support from numerous businesses and industries, universities, foundations, the state of Louisiana, and the Department of Defense, and has paid more than $4 million to college student interns. STEM NOLA has expanded into other cities in Louisiana and across the Gulf Coast and beyond using a versatile scaling model. Onetime pop-up events can be managed by a central team and delivered either in-person or virtually to any location with a local organization supporting recruitment. If support from a funding sponsor or industry partner is available, full-time STEM NOLA staff members can work with local partners to help accelerate the development of a new learning ecosystem with local ownership.

To date, STEM NOLA has primarily documented its impact in terms of numbers of participants and events and their geographic distribution and spread. They have also conducted pre/post surveys of students, which show notable gains in their level of interest in STEM. These measures are in line with their stated goals of increasing exposure, participation, and engagement, especially in underserved communities. STEM NOLA has not created formal curriculum frameworks for materials they develop and distribute, though they do aim for topical alignment with state STEM standards. Nor have they measured other outcomes and impacts, such as student learning or depth and sustainability of program implementation and ecosystem development in new locations. While they would welcome participation by researchers, they note that they do not have internal capacity to conduct research beyond participation metrics.

The Importance of Systemic Capacity

Stepping back from individual initiatives to take a broader view, a theme that emerged repeatedly from the committee’s analysis of innovations that have achieved some success in growing and evolving along one or more of the dimensions of scaling, as identified in Chapter 4 is that networks of functional partnerships among actors in multiple roles at multiple levels often enabled and facilitated the progress of those innovation. The more

ambitious an innovation is, the more likely it is to require not just immediate strategic partnerships but also durable systemic infrastructure in order to take root and be sustained in new settings. The committee believes that building interconnected systemic capacities across the education landscape would provide more robust and hospitable environments for effective innovations to scale and spread in equitable ways and be productively adapted to meet the needs different groups of learners.

Systemic infrastructure weaves together elements such as the interplay between designers, facilitators, and enactors; the aggregation of human, social, and material resources; the need for collective learning opportunities for actors in different roles; and the dynamics of leadership, agency, and ownership over time. Building systemic infrastructure involves connecting people and organizations who occupy different roles and who have responsibility and agency for different spheres of influence in ways that promote alignment and integration, so that goals and activities are complementary and mutually reinforcing rather than divergent and in tension (Hopkins & Woulfin, 2015). Building such collective capacity takes considerable time and consistent effort. Actors and participants at boundaries and intersections—whether across the development and implementation landscapes, at different levels of the formal education system, or at any of the many junctures where non-system actors engage with educational settings—need to develop mutual understandings of each other’s positions and concerns and of how to align what each party can bring in the service of STEM teaching and learning for all students. What’s more, these systemic structures need to be dynamically sustainable. Many initiatives in education tend to devolve over time unless commitment and communication are maintained and continually renewed, in order to survive turnover of personnel, changes in priorities and accountability structures, and new demands on budgets and other resources. These issues are considered further in Chapter 8.

SUMMARY

The configurations of actors, decision makers, and resources that are active when evidence-based STEM education innovations are being developed are typically very different from the configurations that populate the implementation landscape in the public education system. This is particularly true for projects that have been funded with grants from federal agencies to researchers and designers who generally work outside local schools and districts or state education agencies. Many promising research-based innovations never scale because they are not set up to navigate the transition from their original research and development environment to implementation in real-world settings. The innovations may not integrate easily with existing curricula or practices, they may not be readily adaptable to the

needs of diverse student populations, or they may be ineffective in the absence of professional learning opportunities for teachers and school leaders to learn how to implement them independently. Innovations may also fail to spread because the original designers lack the knowledge and resources to engage in professional marketing or to provide reliable distribution systems. These barriers to successful transitions from initial development to widespread implementation are primarily structural rather than intentional, arising from divergent incentives, knowledge bases, resources, and needs across the development and implementation landscapes. In general, most designers would like to see their innovations take root and spread and school leaders would like to access high-quality research-based innovations if they are responsive to school needs and can be tailored to local circumstances.

Where programs and resources have been able to spread and be sustained over time, they have done so in many different ways and with varying degrees of success. This variety is due, in part, to the sparseness of standard pathways across national, regional, state, and local levels for bringing innovations to scale, with the result that many individual innovations create their own unique systems and networks. Interventions and programs that have scaled successfully also tend to share some general features, such as ensuring that they address important needs in educational settings, mobilizing and supporting both enactors and facilitators, and partnering to build necessary capacity and infrastructure. They also tend to start the process of planning for scaling early in their lifespan and to find ways to sustain their efforts over multiple years. There are also some models of durable systems that have created networks of relationships, partnerships, and resources with general capacity to spread and support multiple STEM programs and resources. These include RELs, funded by ED, state policies to support the adoption of HQIM, and state and community-led STEM ecosystems. Actors and supporters outside the formal education system, such as professional organizations, informal science organizations, families and community members, and industry partners, can be influential in advocating and leveraging additional resources for new and improved STEM opportunities and shaping them to be responsive to the needs of local participants and settings.

REFERENCES