Launching ARPA-ED for Higher Ed

Inventing the Internet. Creating GPS. Testing bullets that change course in mid-flight, trucks that fly, and high energy lasers. The Defense Advanced Research Projects Agency (DARPA), was created in 1958 to cultivate breakthrough technologies for national security and in response to the Soviet Union’s launch of the world’s first satellite--Sputnik 1--into space. Now President Obama is calling for that same innovation in education learning technology. The President’s Council of Advisors on Science and Technology first called for the Advanced Research Projects Agency – Education (ARPA-ED) in 2010--to help technology play a transformative role in K-12 education and to address the under-investment in learning technology research and development (R&D).

Across the education spectrum, there's much buzz about how education technology is revolutionizing learning, in particular, the rise of personalized learning through adaptive learning, predictive analytics, and big data in education. In higher education, these new tools have enabled educators, administrators, and institutional researchers to move away from analyzing data for reporting and compliance purposes, and instead move toward analyzing data in real-time to support student success. This is possible because these tools rely heavily on student data, learning analytics produced in real-time, and algorithmic-based education technologies. As a result, they can help target content to the individual needs and interests of students, give faculty better insights on how students progress or don’t progress in their class, and even raise new questions about how curriculum and program design can best ensure student learning and engagement.

Despite (or maybe because of) their promise, some believe these new tools should go through a vetting process. Education technology vendors often partner with colleges and universities to develop and deploy these tools, and to be competitive they have to communicate why their particular solutions are better than the next vendor’s. However, education researchers and experts believe the next-generation learning technologies themselves should be vetted and peer reviewed in academia by researchers who can attest that they are grounded in the latest research and advances in teaching and learning.

Experts Call for Caution

Dr. Candace Thille, a pioneer in learning analytics, has raised concerns about how often the underlying algorithms that serve as the basis for education technology companies’ predictive tools are often black boxes--meaning it’s difficult to know or even understand how they work. Additionally, these black boxes, may not have been informed by or open to faculty and researchers focused on teaching and learning, and could have real implications for the success of higher education institutions and students that adopt them. And oftentimes, the algorithms, software, and tools are trade secrets--proprietary information that is of commercial value, making it unlikely for a vendor to make them transparent for research purposes. If a trade secret is misappropriated, a company can seek remedy through legal means.

In a recent interview, Dr. Thille proposed a way to prevent this from happening. She warned colleges to find ways to raise the money for R&D of learning software so that companies don’t end up owning the classroom delivery system of the future. Dr. Thille suggested colleges come together to build such systems--through a kind of crowdsourcing--or that federal government could step in.

Innovation Stalled

The federal government has tried to step in. The role and promise of education technology to transform the K-12 sector--similar to its role and promise in transforming higher education--was the impetus for ARPA-ED. ARPA-ED, President Obama believed, would be a key component of the country’s opportunity to out-innovate and out-educate the world, so that America can win the future.

After calling for its creation in 2010, in his FY 2012 budget, President Obama proposed the creation of ARPA-ED at the Department of Education, with funding of $90 million in its first year. ARPA-ED would allow the Department of Education to rapidly advance breakthrough innovations in education technology by creating interdisciplinary teams comprised of the nation’s top experts in education, technology, and other key disciplines. ARPA-ED would identify promising new approaches being pioneered in the private sector and other federal agencies in the areas of distance learning, intelligent tutoring systems, and real-time assessment. Winners would be selected based on their potential to create dramatic breakthroughs to empower learning and teaching. And, applying research and development strategies learned from DARPA, these technologies would be rapidly prototyped and transitioned to practice.

Although President Obama was unsuccessful in 2012 and 2013 in getting ARPA-ED funded, he requested up to $30 million for ARPA-ED in his 2017 budget. Despite these failed attempts, the proposal is still a promising idea and relevant for both K-12 and higher education. Why then hasn’t the case for ARPA-ED for higher education been more widely put forth?

One answer may lie in how little federal funding is awarded to universities for education R&D, especially when compared to other fields. For example, education R&D spending at universities financed by the federal government in 2014 was $661 million, down from roughly $664 million in 2013 and roughly $688 million in 2012. Compare this with over $7 billion in funding for financing R&D expenditures in biological sciences in 2014 and over $11 billion for R&D in medical sciences the same year.

Building for the Future

In addition to greater investment in education technology R&D, another way to turn the narrative around would be to include ARPA-ED for higher education in the forthcoming Higher Education Act (HEA) reauthorization. This wouldn’t be the first time this legislation created a new national center with leading educators, researchers, technology firms, and entrepreneurs tasked with advancing technologies that can transform teaching and learning. Digital Promise for example--an independent, bipartisan nonprofit--was launched in 2011 to accelerate innovation in education through technology and research. Digital Promise came to fruition as a result of being included the HEA reauthorization of 2008. What makes Digital Promise different from ARPA-ED and an ARPA-ED for higher education is that Digital Promise was designed for researchers and entrepreneurs to develop new approaches for rapidly evaluating new products, and not for traditional education R&D projects that can take many years. Furthermore, many of Digital Promises’ initiatives focus on learning technologies for K-12 aged-population and their environments, and not necessarily for college-going students in higher education institutions. Thus, R&D for education technology specifically designed for older populations has yet to be given a launching pad.

A potential drawback to advocating for greater investment in education technology R&D is that it might propel more collection of student data or further encourage the rise of an algorithmic-based education for all students. The idea is, adding investment means this is the direction higher education should go in. But higher education seems to already be heading in this direction, with hundreds of colleges already using predictive analytic tools to help enrolled students increase their chance at success. Only now, innovation is just more dispersed, happening on an institution-by-institution--or more approximately--an institution-by-vendor basis.

At the same time, greater attention and investment in education technology R&D at the national level through ARPA-ED for higher education could drive parallel efforts around the need for updated standards for student privacy, appropriate uses of student data, and increased training for staff who will use these tools. As a research and innovation hub, ARPA-ED for higher education could simultaneously add fuel to these parallel efforts, which should go in tandem with creating the learning technologies themselves. And at the very least, ARPA-ED for higher education would elevate the need for and investment in R&D on learning technologies to a national level, where it deserves to be. This is particularly true if the field is increasingly moving in this direction and federal investment in this area doesn’t keep up.

The United States prides itself on being a nation of innovators, entrepreneurs, builders, and pioneers. We’re gearing up in many ways for the future; but why aren’t we viewing education technology R&D in a similar light? How long will it take for ARPA-ED for higher education to finally get its launch date?"


Manuela Ekowo is a policy analyst with the Education Policy program at New America. She provides research and analysis on policies related to higher education including innovations in higher education delivery, the use of technology, open educational resources (OER), and ensuring equitable outcomes for underrepresented students.