STEM Education for the Next Generation

Over the past decade, precollege education in science, technology, engineering, and mathematics, or STEM, has expanded and developed dramatically. Instruction is now supported by a large and growing body of research, especially in science learning, as well as updated graduate degree programs for teachers and the adoption of guidelines such as the Next Generation of Science Standards. These are research-based, up-to-date standards developed by 26 American states in coordination with the National Research Council, the National Science Teachers Association, and the American Association for the Advancement of Science, and released in 2013 for adoption by states.

Additionally, new subjects such as coding and robotics are increasingly incorporated into the classroom, and artificial intelligence, or AI, has become a commonly used teaching tool. 

There is also a greater, more conscious focus on equity and inclusion in teaching strategies and curricula, especially related to women and girls, members of historically underrepresented groups, and learners with a range of disabilities. There is a greater awareness among educators that STEM education is not only about teaching science and helping students become technology literate, but also about encouraging their capacity to learn and integrate new information and facilitating their ability to work and cooperate with others to solve problems. 

Policy priorities

Science education has been a U.S. strategic and economic policy concern since the end of World War II. The Oct. 4, 1957, launch of Sputnik 1, the world’s first Earth-orbiting satellite, by the Soviet Union took the United States by surprise and revealed it had fallen behind in science and technology. The push to surpass the achievements of the Soviet space program led to the formation of NASA in 1958, the National Defense Education Act of 1958, and the Higher Education Act in 1965. 

These federal initiatives expanded science education and boosted public support for science. They also laid the groundwork for an innovation ecosystem and for many landmark achievements, such as the artificial heart, the cell phone, the space shuttle, and the personal computer.¹

“There’s a lot of research that shows if you know your brain evolves as you try new things, and if you know you will get better, you will try again and improve.” 

— Linda Darling-Hammond, Ed.D.

However, this high-tech shift did not include significant numbers of women and members of historically underrepresented groups, failing to tap into their potential to contribute to scientific advances.² The Science and Technology Equal Opportunities Act of 1980 aimed to increase participation by women and members of underrepresented groups by directing the National Science Foundation, or NSF, to address the lack of diversity in America’s STEM workforce. The America COMPETES Reauthorization Act of 2010 required the NSF to make increased participation by groups marginalized in STEM a criterion for funding.³ NSF policy adjustments resulted in greater diversity among graduate students seeking grants for science research projects, but these efforts have been hobbled by recent political opposition to affirmative action.⁴

In 2009, this time driven by tech competition from abroad and a shortage of STEM professionals, the U.S. made another push to improve its technical and scientific capacity. President Barack Obama launched the “Educate to Innovate” campaign to improve the global standing of U.S. students in STEM. The program, among other things, directed $4.35 billion to fund schools in states that committed to improving STEM education; established a national network of partnerships to raise money for teacher training; and initiated an annual White House science fair. In his Jan. 26, 2011, State of the Union address Obama stressed the crucial importance of STEM education, famously declaring this to be “our generation’s ‘Sputnik moment.’”⁵

STEM education does not appear to be a priority for the current U.S. administration. More than 1,400 NSF grants have been canceled since April 2025. Nearly half of these were for STEM education and included programs for Indigenous youth, students with autism, and girls in rural schools learning drone technology.⁶

The NSF cuts were a “gut punch,” said Julie Posselt, Ph.D. — an associate dean of the graduate school at the University of Souther California and the executive director of its Center for Enrollment Research, Policy, and Practice — in a New York Times article on May 22, 2025. “It’s about the association of educational research with interests and values that are at odds with the administration’s priorities,” she said.⁷ The decimation of the U.S. Department of Education and the administration’s threats to limit or cut off funding to such institutions as Harvard, Columbia, Cornell, Brown, and Northwestern have also set back STEM education, according to the article.

Bridge to the future

Recent developments in researchers’ understanding of how people learn and how they come to identify themselves — or not — as capable of STEM subjects are persuading experts to rethink how these subjects should be taught in the future. Given how many jobs now require some kind of STEM aptitude — two out of three, according to 2020 research by a group of STEM membership organizations — these adjustments are timely.⁸

Linda Darling-Hammond, Ed.D., is chief knowledge officer at the Learning Policy Institute, or LPI, a nonpartisan, nonprofit education policy and practice research center founded in 2015 in Palo Alto, California. She sets the direction of the organization’s research, which focuses on school redesign, civil rights and racial equality, deep learning, and the science of learning and development. Over the past 10 years, LPI’s research has informed education policy on the state and national level and its Educator Preparation Laboratory, or EdPrepLabs, programs and videos have expanded and improved teacher training.⁹

“It’s not that we want every kid to have a STEM career, we just don’t want them to shut the door to math and science.”

— Maya Israel, Ph.D.

Dr. Darling-Hammond contends that U.S. STEM and precollege education in general need fundamental change in four main areas to be future-ready: 

• New standards and assessment methods. “We have an antiquated system in the U.S. that does not reinforce deeper understanding, inquiry, and engagement with the subject matter,” she says. Indeed, in a recent blog she wrote, “At the high school level, the United States has clung to a math curriculum prescribed by a set of educators called the Committee of Ten, appointed by the National Education Association in 1892.”
• More professional development for STEM teachers. The U.S. also needs incentives to encourage educators to adopt new teaching methods. “Project-based learning is not unstructured learning,” Dr. Darling-Hammond asserts. “It’s highly structured and requires thoughtful preparation.”
• Better pay for teachers. Teaching needs to become a more desirable and respected profession. High-achieving countries (where students do well in math and science) pay teachers as much as other college-educated professionals. U.S. teachers are paid on average 25% less than other professionals, with differences from state to state. The pay gap is even greater for people with STEM degrees, who can earn 50% to 100% more in industry.  
• A culture of inclusion, especially in precollege mathematics. That begins with equal funding and equal access to qualified teachers. “A lot of inequality is baked into our system,” Dr. Darling-Hammond says. “Kids need to see science teachers who look like them and kids who look like them on classroom posters.” Teachers need to communicate to all kids that they have the capacity to learn, she says, “and we must stop perpetuating the idea that there is a ‘math gene.’” This is an important part of inclusive STEM teaching because, in the not-so-distant past, it was assumed that mostly white men and boys had this “math gene,” which reinforced the idea that women and historically underrepresented groups don’t belong in math and science. 

Students need to be aware of the growth mindset and encouraged to develop it. “There’s a lot of research that shows if you know your brain evolves as you try new things, and if you know you will get better, you will try again and improve.” Unfortunately, the assumption of fixed intelligence is deeply held by parents, educators, and even students themselves, she says. 

Although elements of the current political environment seem hostile to science and education, and the infrastructure for STEM fields is being threatened, “There are still a lot of people who really want evidence,” Dr. Darling-Hammond says. “It’s part of the federal role to stimulate, and we’re probably not going to have much of that stimulation in the near term. So, it’s going to be important for the states to take this up. Education is a state function, and I think there is an appreciation that we need to get much more serious about investing in STEM education.”

Reaching all students

Maya Israel, Ph.D., stands at the nexus of research, teacher training, inclusive student learning, and computer science. She is an associate professor of educational technology and computer science education in the School of Teaching and Learning at the University of Florida. She is also the founder and executive director of the Kenneth C. Griffen CS Education for All initiative, which focuses on teacher preparation and artificial intelligence, or AI, and ensuring that learners with disabilities thrive in computer science, robotics, and AI classes, among other goals. 

Over the past decade, she has seen many changes in STEM education. After 2016, there was a big push to extend computing to all precollege students. Previously, it had been confined to Advanced Placement, or AP, classes and older students. Other changes include integrating makerspaces and the arts into STEM. Additionally, Dr. Israel says, “You can’t talk about changes in the last 10 years without talking about the effects of the COVID-19 pandemic. Since 2020, there has been more widespread use of digital learning platforms. Simulation and virtual labs existed before the pandemic, but they have been integrated into schools and are more online-friendly.” 

Much more is known now about developmentally and age-appropriate STEM curricula, she says. “There used to be this idea of the average learner. We now know it’s much more complex than that.” Dr. Israel says it is now understood that the foundation for effective STEM instruction includes: 

• Universal design for learning. Instruction that is flexible and considers the range of differences in how students learn. 
• High-leverage practices. Teaching strategies that develop independent thinking and deeper understanding, such as scaffolding, modeling, and instruction prior to open-ended experiences.
• Accessibility. A simulation may not be accessible to a blind student, or the cognitive demand of a simulation may be too great for a student with reading issues. Adaptation is key.

To further improve STEM learning in young people, Dr. Israel advocates minimizing reliance on technology and computers in the early grades. “With young learners, it’s best to start with play and curiosity,” she says. “STEM learning should be about figuring out the world.” Then, as children age and develop more complex models of the world, more technologies, such as simulations and data processing, can be introduced. 

Another challenge that has recently been identified is students who simply feel STEM subjects are beyond their grasp. “There is a lot of literature now about STEM identity, and the fact that a lot of the way we look at our own abilities forms early in life,” Dr. Israel says. 

“When I first entered the field, equity and cultural issues were recognized, but were not central to the discourse. In the last decade, there has been a concerted push to expand the theoretical frame to include identity, culture, and the sociopolitical as central to the learning process.”

— Shuchi Grover, Ph.D.

“Honestly, by middle school, most kids have made up their minds about this. If a child has had early, sustained, positive experiences with math and science, they are more likely to believe that they can do math and science.” These ability beliefs often determine whether or not a student will engage and persist with scientific material. “Unfortunately, our culture has normalized people telling themselves, ‘I’m not a math person, I’m not good at science.’”

This understanding of early STEM identity can help learners who think they are not good at math and science see their abilities differently, Dr. Israel says. “For kids whose identity formation is still being shaped by their experiences, we want to center their voice, center their experiences. For example, if we can ask a question that is really relevant to them, they will be more likely to see a purpose in science.” Many of the problem-based approaches now used in STEM education are designed to engage those reluctant learners. Dr. Israel says, “It’s not that we want every kid to have a STEM career, we just don’t want them to shut the door to math and science.”  

Rapidly changing field

In her “long and winding journey” from computer science and physics to learning sciences and computer science education, Shuchi Grover, Ph.D., has been involved in many of the fundamental developments — both pedagogical and technological — in STEM education. 

She is currently director of AI and education research at Looking Glass Ventures/Edfinity, where she leads the conception, investigation, and development of learning environments and assessments in precollege STEM and computer science education. In addition to being the author of many scholarly articles, Dr. Grover is the creator and editor of Computer Science in K-12: An A to Z Handbook on Teaching Programming.

After earning a B.S. in computer science and physics at Birla Institute of Technology and Science in India, Dr. Grover moved to the U.S. for graduate studies in AI at Case Western Reserve University. She then joined the Technology, Innovation, and Education program at Harvard Graduate School of Education in 2000, where she was introduced to what she calls the “bible” of the learning sciences: How People Learn, published by the National Academies in 2001.¹⁰

“As I dug deeper into working with children, teachers, and schools,” Dr. Grover says, “I had more questions about how children learn, what it means to teach kids coding, how to imbue in them skills that we now call computational thinking,¹¹ and how to prepare teachers for creating these computationally rich learning experiences.”  

She points out, “One of the most fundamental ideas in the field of learning science is that learning is contextual.” The growing recognition that teachers need to better prepare precollege students for a computer-driven world set off a change in the learning sciences to include computer science, engineering, and data science, among other subjects. Similarly, the maker movement has fueled new research in precollege engineering education, and the rise of big data has expanded learning science into that area as well. 

Dr. Grover has witnessed other shifts in learning sciences. “When I first entered the field, equity and cultural issues were recognized, but were not central to the discourse,” she says. “In the last decade, there has been a concerted push to expand the theoretical frame to include identity, culture, and the sociopolitical as central to the learning process.” 

In addition, new interdisciplinary fields now intersect with the learning sciences. For example, the rise of MOOCs, or Massive Open Online Courses, other online and digital learning environments, and computational analytics have accelerated the development of educational data mining, learning analytics, AI in education, and what is called Learning@Scale.¹² Dr. Grover says, “Learning analytics research is now embedded in many learning environments to understand and foster collaboration; conceptual learning, motivation, and interest; and learning that is responsive to individual students’ needs.”

Another recent development is formalizing mechanisms for bridging research and practice. Design-based implementation research, for example, involves building research/practice partnerships throughout the research process. “A related area is participatory design and co-design, where educators are more intentionally involved in the design of learning environments and experiences,” says Dr. Grover. “There are increasing moves to include student voices and inputs as well.”

Robotics and hands-on learning

In 2010, with funds from the Amir Lopatin Fellowship at Stanford, Dr. Grover ran a robotics summer camp in India to examine the development of computational thinking. She used the GoGo board microcontroller, an open-source, programmable device for sensing and robotics. 

“Students could see the circuitry, input/output ports, and the resistors and capacitors,” she says. “They had to solder electronic components together to create sensors for their projects. This made it feel more like a computer science engineering experience for the young teens I worked with.”  

More precise than “hands-on” is “tangible or physical” computing, which has become increasingly popular in the last 20 years. Dr. Grover says, “I love using physical computing because, and this has been borne out by research, kids find it very engaging. The projects are tangible and goal-oriented and usually done in groups, so there’s an element of collaboration. The abstractions are both in physical space and in code. The projects are more interdisciplinary, bringing in science, math, engineering, data science, and computer science. In addition, being up and about affords embodied learning — a big part of the learning puzzle.”¹³

Robotics is known to increase interest in engineering and computer science among girls and students from underserved backgrounds, Dr. Grover says. “Because of the connection to real-world objects and artifacts, students can connect to their cultures, personal passions — such as sports and art — and pressing issues such as energy conservation. Robotics projects can teach cause-and-effect, conditional thinking, and basic turtle geometry to younger students,” she says. (Turtle geometry is a visual math and programming learning tool in which students move a turtle icon on a personal computer display.¹⁴) 

Dr. Grover says that in higher grades, students can engage in problem decomposition, or breaking large, complex problems into smaller, more manageable parts. They can also be introduced to collaboration and task division, as well as algorithmic thinking, systems thinking, and abstraction.  

Despite all the progress in understanding how students best learn these concepts, the teacher shortage and the lack of teacher training have been barriers to teaching abstraction, logic, and pattern recognition, Dr. Grover says. “It’s a lot to expect of teachers who have no background in programming and computer science to understand what computational thinking is about. The onus is on the community, the researchers and computer science folks, to help prepare teachers adequately.”

Policymakers, learning science experts, and STEM professionals agree that nurturing STEM literacy and making STEM education accessible to all learners and teachers is more important than ever. 

As National Academies of Science President Marcia McNutt, Ph.D., emphasized in her June 2025 “State of the Science” address,¹⁵ the critical, creative, and independent thinking that STEM subjects demand and develop from kindergarten to graduate school are crucial to U.S. economic and strategic security and to maintaining a democratic society open to new ideas. 

References

1. Dottie Rose Foundation. (Sept. 16, 2022). “A Brief History of STEM.”

2. Peterson, Charlotte. (June 30, 2023). “Physics, Patriotism, and Propaganda: American Education’s Continuity and Changes After Sputnik.” (Third place winner by a Stuyvesant high school senior in the David McCullough essay competition, hosted by the Gilder Lehrman Institute of American History.)

3. Miller, Katrina. (May 22, 2025). “Funding Cuts Are a ‘Gut Punch’ for STEM Education Researchers,” New York Times. (Subscription required.)

4. Muller-Parker, Gisele and Bourke, Jason. (Fall 2023). “Fifty Years of Strategies for Equal Access to Graduate Fellowships.” Issues in Science and Technology, vol. XL, No. 1.

5. Robelen, Erik W. (Jan. 26, 2011). “Obama Emphasizes STEM Education in State of the Union.” Education Week.

6.   Belsha, Kalyn. (June 10, 2025). “Why the Trump administration grounded these middle schoolers’ drones — and other STEM research.” Chalkbeat.

7. Blinder, Alan. (July 11, 2025). “Trump has Targeted These Universities. Why?” New York Times.

8. STEM and the American Workforce: An Inclusive Analysis of the Jobs, GDP, and Output Powered by Science and Engineering. (2020). Aerospace Industries Association, American Association for the Advancement of Science, American Chemical Society et al. 

9. Learning Policy Institute, Research. Action. Impact. 2023-2024 Annual Report.

10. National Research Council, et al. (2000). How People Learn: Brain, Mind, Experience, and School: Expanded Edition. Washington, D.C.: The National Academies Press. https://doi.org/10.17226/9853.

11. Computational thinking is a multistep, multilevel problem-solving approach essential to computer science, which has come to be recognized as applicable to all STEM disciplines and nonscience fields as well. In 2006, Jeannette M. Wing, Ph.D., executive vice president for research and professor of computer science at Columbia University, published an influential essay, “Computational Thinking” (Communications of the ACM — Association for Computing Machinery, March 1, 2006). In this essay, Dr. Wing asserts, “Computational thinking is a way that humans, not computers, think. Computational thinking is a way humans solve problems; it is not trying to get humans to think like computers.”

12. Learning@Scale involves formal learning environments, such as intelligent tutoring systems and remote university classes, that involve technology and very large and diverse groups of learners. Other examples of Learning@Scale include open courseware, learning games, and citizen science communities. 

13. Embodied learning recognizes the mind and body are not separate, that movement and bodily sensations are important to cognition, and that learning in a physically interactive environment heightens alertness and improves absorption and retention of new ideas. Dor Abrahamson, Ph.D., professor of learning sciences and human development at the University of California, Berkeley, runs the Embodied Design Research Lab and has published widely on the subject. For more information, see “The Future of Embodied Design for Mathematics Teaching and Learning,” Dor Abrahamson et al., Frontiers in Education, Aug. 24, 2020.

14. Turtle Geometry: The Computer as a Medium for Exploring Mathematics is a college-level text by Andrea diSessa, Ph.D., and Harold Abelson, Ph.D., published by MIT Press in 1981. Dr. Abelson is a professor of computer science and engineering at MIT and the founding director of Creative Commons and other efforts to democratize technology. In the introduction to Turtle Geometry, he describes it as “mathematics designed for exploration.”

15. “National Academy of Sciences Warns of a Decline in U.S. Leadership, Seeks Solutions,” SWE Magazine, this issue.