Towards Teaching Data Science
National Academies Workshop Starts Phase 2 of a Movement
The revolution of, the providence of, and the ubiquity of data — all big words to describe a simple reality: data is everywhere and we have to do something about it in education.
What if we taught the foundations of data science in high school? Middle school? Kindergarten?
Last week, I had the honor to help facilitate a national convening at National Academies of Sciences, Engineering, and Medicine (NASEM) to explore these questions in grades K-12. With nearly 100 in-person in Washington D.C. and 500 online, the workshop convened education researchers, policymakers, practitioners, and teachers to take stock early efforts in K-12.
This national convening marks a phase shift in the tide for bringing introductory data science to the U.S. public education system.
Back in 2019, a Freakonomics podcast put a spotlight on a number of ongoing efforts around the country to enliven K-12 mathematics curricula with data science programs; it quickly became one of the most popular episodes on the podcast. These projects complemented an earlier push in K-12 for Computer Science for All beginning in 2016 (lead by Medium writer Leigh Ann DeLyser), and a modernization of statistics education guidelines in 2020 — yet with an even greater focus on the unique aspects of data science.
The workshop also built upon a 2018 National Academies report at the college level, which recommended that “all U.S. undergraduate students should develop a basic understanding of data science to prepare them adequately for the workforce.” The same report also noted that middle and high school experiences may “better prepare students both for postsecondary curricula and for the data-driven workforce that awaits them,” and imagined a future in which “the vast majority of high school graduates would have a basic understanding of data science.” So how close are we to that future?
I was thrilled to help highlight the work of educators from around the country — and across school subjects — in one of the workshop’s six panel discussions. We heard from Suyen Machado (Introduction to Data Science / UCLA, CA), Stephanie Melville (San Diego Unified School District, CA), Paul Strode (Fairview High School, CO), and Katie Headrick Taylor (University of Washington, WA) — a mix of educators and program designers that were able to share what’s happening on the ground.
From both our panel and others, a few consistent themes emerged: student enthusiasm is through the roof; their motivation seems to drive accelerated learning through content; their option to choose topics of personal relevance increases engagement, and they can learn more about their own communities (and their challenges) through collecting and examining local data. Now imagine your own high school experience resembled that. Might it have changed your path?
Demand is also outpacing supply: students are asking for more courses, for opportunities in advanced topics like machine-learning and A.I., and even requesting information on internships and data-related job opportunities. In stakeholder interviews, one mathematics educator shared they “have never gotten those questions before in over 20 years of teaching.”
In preparation for the workshop, we also surfaced approximately 14 statewide pilot programs, 2,000 local schools and districts, and 1,500 teachers attempting to impart introductory data science to K-12 students — representing highly diverse geographical, social, and industry contexts. Collectively, these programs are impacting approximately 180,000 students across the country as of this school year. Moreover, these numbers exclude dozens of projects that are infusing technology-based data science experiences into science, social studies, and the humanities. While the numbers are significant, it still only comprises roughly 3% of high school students nationally. We have more work to do.
What does more work look like?
“Teachers need more help. Please help us.” While the panel emphasized that data science may help re-engage students in an otherwise digital world full of classroom distractions, teachers also need to feel confident and empowered to bring these experiences to the classroom. Both the panel on practice and a later panel on professional development (facilitated by the Friday Institute’s Hollylynne Lee) highlighted the need for greater investment in training: for teachers, for teacher trainers, and for the programs and policies that mediate teacher preparation and professional development. Creating space in state standards, in college admissions, in district support, and in licensure eligibility is needed to create the room for that training to then be utilized.
For research, we heard a related call to study and improve the methods through which teachers learn about data, statistical reasoning, and technology. We also heard calls for additional research into: learning progressions, tool and technology progressions, software accessible for students with disabilities, exploration of data literacies (e.g. datafication, data ethics, and data contexts) to be infused throughout curricula, and modernized assessments that give fidelity to data science processes rather than bullet-listed knowledge.
Want to learn more? The recordings and four commissioned papers will soon be available on the event page. The National Academies will also release a full summary report later this year, capturing the wide array of programs, early evidence, and recommendations throughout the two-day convening in greater detail. Many of these resources have now been organized by Data Science 4 Everyone, a clearinghouse organized to respond, in-part, to the momentum from the single podcast episode in 2019.
This workshop marked a phase shift in data science education: both from undergraduate to K-12, and from demonstration projects to some early scale.
Now we need to build more evidence, better resources, and greater access for all students — so that data literacy becomes as universal as reading or writing.
