Connecting the Work
In Doubling Down, I wrote about why we’re sharpening our focus in U.S. education on the moments when students lose momentum in their education journey.
We know from decades of research that certain milestones are highly predictive of long-term success: passing Algebra I by ninth grade, transitioning from high school to college, passing critical first-year college courses, and earning credentials of value that lead to economic opportunity.
In my recent piece on Donald Hertz Middle School in New York City, I shared what it can look like in practice when K–12 teachers, supported by technology, can better understand student learning and respond more effectively as students build the foundational skills needed to stay on track through key early milestones.
Here is another example of that same idea, but in a different setting: the first year of college.
For many learners pursuing careers in health care, engineering, and science, the first year of college can determine whether they stay on track or lose momentum altogether.
Each year, millions of students enroll in foundational science, technology, engineering, and math (STEM) courses, yet DFW rates, the share of students who earn a D or F grade or withdraw, average between 30 percent and 35 percent at four-year institutions.
Too often, students who are fully capable of succeeding fall behind because schools and institutions are inconsistent and uneven in providing timely instruction, guidance, and support when students struggle.
The consequences can be lasting: students change majors, lose confidence, or leave college altogether, closing off pathways to degrees and careers that once felt within reach.
Addressing this challenge requires redesigning first-year courses to enable more students to succeed while preserving rigor and high academic standards.
That’s why we’ve invested in REAL CHEM, an evidence-driven effort to redesign how college chemistry is taught and learned.
What REAL CHEM Represents
REAL CHEM is a tech-enabled general chemistry course developed by faculty, learning designers, and learning scientists at Arizona State University and Carnegie Mellon University. The course is part of a pilot within a broader effort to demonstrate how colleges and universities can better design first-year courses and student supports to help learners stay on track toward a credential.
Created by faculty for faculty, REAL CHEM integrates instruction, practice, assessment, and feedback into a single learning experience. Because AI is built directly into the platform, it can provide tailored guidance on specific concepts students are working on, while giving instructors clearer visibility into where students are struggling.
Part of what’s different about this tech-enabled tool compared to a traditional chemistry course is that REAL CHEM uses digital courseware to reimagine how students experience and engage with chemistry.
In many first-year chemistry courses, students spend significant time memorizing formulas and procedures without always understanding how the material connects to the world around them or why it matters.
Dr. Elaine Villanueva Bernal of California State University, Long Beach, explained:
“Chemistry is in the kitchen, the shoes you’re wearing, the birds outside. REAL CHEM brings these connections into the classroom, so students don’t just learn chemistry; they experience it.”
With REAL CHEM, students don’t watch lectures passively or work through static assignments. They interact with chemistry concepts in real time through guided activities, interactive video, timely feedback, and problems tied to real-world contexts such as energy systems, health care, and the environment.
Alexa G., a Psychology major at California State University Long Beach, described her experience this way:
“REAL CHEM has allowed me to connect chemistry to everyday experiences and the world around me. Through learning, I became inspired to take more action in caring for our environment.”
That kind of engagement matters, especially for learners balancing school, work, and life demands. When students experience material as relevant to their lives and receive timely support when they need it, they are more likely to persist through challenging first-year courses.
REAL CHEM’s AI-enabled tutor can help students work through problems without giving away the answer, allowing instructors to better target lessons. The goal is not to use technology to automate teaching or replace faculty, but to strengthen instruction and help more students stay engaged and persist.
Justin C., a Chemical Engineering major at the University of Connecticut, explained his experience:
“With REAL CHEM, I didn’t just memorize; I truly understood the concepts. Getting questions right on the first try showed me I was building real proficiency with the material.”
Although we’re excited about what we’re learning from this pilot, the foundation is also clear-eyed about the risks and responsibilities that come with using AI in education.
We believe tools like REAL CHEM should be grounded in evidence about how well they work and for whom, while also protecting student data, safeguarding student well-being, and supporting responsible implementation.
What the Evidence Shows
More than 34,000 students have now used REAL CHEM across hundreds of classes at more than 80 colleges and universities nationwide.
At the University of California, Riverside, REAL CHEM students scored 19 percentage points higher on midterms, equivalent to a two-letter grade increase, compared to peers not using the courseware.
Research from Digital Promise helps explain why: students in REAL CHEM courses experience more of the research-backed teaching practices known to drive learning, including active engagement, guided practice, and timely feedback. When those practices are sustained over time, students perform better, even when accounting for their starting point.
Another important insight comes from the University of Maryland, Baltimore County, showing that success is not driven by performance on a single test. It’s driven by whether students continue making steady progress throughout the course.
Those findings are important because first-year courses are often where institutions either accelerate student progress or see learners fall off track. For higher education leaders, these courses represent one of the clearest opportunities to improve persistence and completion outcomes at scale.
Why This Matters
REAL CHEM shows what’s possible when institutions use technology intentionally to strengthen teaching and provide timely intervention. But technology alone is not enough.
Lasting improvement depends on whether institutions support the educators and students using these tools and create the conditions for effective implementation. Real progress happens when institutions align teaching, advising, technology, and support systems around helping learners succeed.
Doubling Down Made Concrete
In Doubling Down, I wrote that we are not chasing novelty. We are focused on identifying what works and expanding those tools and approaches to reach more learners.
REAL CHEM offers a practical example of what that looks like: combining strong instruction, evidence-based course design, and thoughtfully implemented technology to better support students in one of the most critical first-year courses in higher education.
As colleges and universities work to improve student outcomes, the question is not whether technology belongs in higher education. It is how institutions use it to strengthen teaching, better support learners, and help them persist and complete credentials of value that lead to long-term opportunity.