ARPA-Ed and the Future of Education Innovation
Proposed Federal Program Would Put Technology to Use for Education
Amid the media’s focus on the debt crisis and congressional gridlock, policymakers have had scant airtime to discuss one of the core challenges facing our nation in the 21st century: making sure our education system continues to prepare the world’s best and brightest.
Backstage from the media circus, the Obama administration and Senate Democrats have been quietly drumming up support for a new $90 million program to accelerate innovation in educational technology, or ed tech. Called the Advanced Research Projects Agency for Education, or ARPA-Ed, in the president’s 2012 budget, this lean, mean little research agency would create a platform for ed tech innovators to compete to develop cutting-edge learning tools.
On Wednesday, as the Senate Health, Education, Labor, and Pensions, or HELP, committee moves to overhaul No Child Left Behind, Senator Michael Bennet (D-CO) is expected to introduce an amendment that would bring ARPA-Ed to life. If enacted, ARPA-Ed would put competitive grants and contracts in the hands of innovators seeking to develop technologies to revolutionize the way students learn. Because it kindles competition, ARPA-Ed is an interesting Obama initiative that has the potential to reverse the sputtering trajectory of the U.S. education system.
As it stands, the situation in public education is precarious. In December 2010 the Program for International Student Assessment, or PISA, announced that the United States ranked 14th in reading, 17th in science, and 25th in math among 63 world nations in academic performance. Comparatively, Chinese students in Shanghai ranked first across the board, with Singapore, Korea, and Finland following close behind. This news has fueled concerns on both sides of the political spectrum that the United States is losing its competitive edge amid a crippling educational and economic malaise.
While our students are subject to global competition, U.S. schools are facing mounting pressure at home to conform to the standards set by the No Child Left Behind Act, or NCLB. Schools are scrambling to achieve proficiency in math and reading by 2014, lest they face fiscal penalties. The looming deadline for improved performance—and the fear that it will not be met—has prompted President Obama and his secretary of education, Arne Duncan, to look into restructuring NCLB in a way that will be more consistent with his vision for education. Central to this vision is ARPA-Ed. ARPA-Ed is built on the same model as DARPA, the defense research program responsible for technologies such as the Internet and GPS.
Proponents of ARPA-Ed aim to exploit cutting-edge technology and developmental psychology in education to create a more efficient, individualized, and engaging learning regimen for today’s students. ARPA-Ed would follow its predecessors in adopting a bottom-up, rather than top-down approach to nurturing innovation. Instead of attempting to dictate what the most effective means of instruction might be, ARPA-Ed will reward learning about learning, researching research, and thinking about thinking.
Until now, the strategy for using technology in education has been to bombard schools with gadgets, sometimes at the expense of teachers’ jobs and to the detriment of art and music programs. The Kyrene school district in Arizona, for instance, touts state-of-the-art classrooms, spending $33 million on student laptops, display screens, and accompanying software. Yet despite all its high-tech bells and whistles, Kyrene’s math and reading scores have stagnated since 2005, even as scores in the rest of the state have risen. Further, it is unclear whether some modest improvements can be attributed to better incorporating technology into classrooms or better teachers using said technology. Technology for technology’s sake, it seems, is not a successful approach.
In light of disappointing results, such as in the case of the Kyrene school system, it would appear that a shift in the way we equip our classrooms is in order. A quick survey of the current landscape of public education, with its socioeconomic performance gaps that could be better characterized as canyons, would suggest that our current system isn’t working.
Is technology part of the solution?
Yet there is hope in ARPA-Ed, which holds promise of funding many exciting innovations that could not only produce innovative educational technologies but also transform the way technology is used. Areas of research include:
- Interactive digital tutors that can guide students through curricula and provide constant feedback.
- Adaptive technology that adjusts to students’ learning styles and rates, and improves as the student uses it more.
- Making learning software function more like video games do.
ARPA-Ed would expand upon efforts initiated by the “Digital Promise Center,” which Congress authorized but didn’t fund in 2008. Education Secretary Arne Duncan unveiled the Center in September, acknowledging that the institution represented an “incredibly important turning point.” A major aspect of the so-called “digital promise” is research into digital tutors.
Digital tutors seek to marry the virtual age with psychological principles of learning. One such tenet of learning is that immediate feedback and reinforcement strengthens the association between stimuli. For instance, when an amateur is learning the violin and plays a wrong note, for best results, an instructor should correct the mistake immediately after it is made in order to give the musician the best odds for not repeating the mistake. Hence the revised maxim, “Practice doesn’t make perfect—perfect practice makes perfect.” Yet musical instruction usually only occurs once or twice a week, leaving novices to correct their own mistakes—a difficult feat when you don’t know what you’re doing wrong in the first place. This was exactly the problem researchers at the National University of Singapore confronted when they developed a virtual violin tutor to help kids practice more effectively.
The Navy has also implemented digital tutors to teach recruits to become IT specialist administrators in just seven weeks. A healthy, all-American suspicion of technology—complete with fantasies of a dystopian future where humans are oppressed by Skynet—may inform skepticism as to whether such a program really works, and whether machines are even capable of educating. But Navy researchers found that these computer-trained specialists outperformed their traditionally educated counterparts with up to three years of experience.
New tutoring software will capitalize not only on advantages in immediate feedback but also in the realm of individualizing education. Analytics are already widely used in business to customize the content you see on the Internet. Analyzing online habits is how Amazon manages to recommend merchandise that you really want or how Netflix seems to have a handle on your taste in movies. Companies like Amazon and Netflix collect data from users’ personal choices and incorporate them into algorithms that filter their databases to formulate suggestions for your next rental or purchase.
Research into education software will employ similar tactics to gather information about how a student learns. One test prep company, Knewton, is already leading the pack. In just a few hours, Knewton’s software can collect 150,000 data points on a student—data points like how long it takes a student to answer a question or to learn a lesson, which kinds of modules are mastered more quickly, and where the student is lagging. The highly sophisticated software aggregates this information, determines a student’s unique pattern of learning, and adjusts accordingly. This technique allows for remarkably personalized instruction. Four major universities have already purchased the product for catching up incoming students on college-level math.
Medical schools have also implemented programs for adapted learning, as reported by the NIH. These programs allow for student control of pace and media, though students are quick to maintain that they view the system as a complement to, rather than a replacement for, instructor-based learning.
If ARPA-Ed led to developments in personalized software for students in grade school, the benefits could be substantial, especially due to the plasticity of the young mind. Yet children, as opposed to GRE test takers and medical students, are slightly less inclined to sit in front of a boring slideshow and answer questions about it. That’s where the next ARPA-Ed research area would come in. The National STEM Video Game Challenge aims to make learning more like something kids actually want to do.
Much like sneaking broccoli into the mac n’ cheese, the National STEM Video Game Challenge seeks to trick kids into learning. The 2010 winners of the Developer Prize, Dan Morton and Dan White of Filament Games, created a game to teach kids about the structure and function of pathogens. The game “You Make Me Sick!” compels players to build a virus or bacteria and infect a host. You can play the prototype here.
Engaging and challenging educational games, coupled with personalized curricula, could afford all kids the opportunity to learn skills to prepare them for higher education. Furthermore, integrating computers and other gadgets into classrooms will free up teachers to have more one-on-one time with students and better control disciplinary problems when they arise. A school in South Carolina that implemented iPad technology for all its students reported a noticeable decline in class disruptions.
But the advantages extend beyond the scope of monitoring class behavior and individual children’s performance. Currently, the mainstream model of education—boasting short class periods delineated by bells that signal it is time to herd students to their next room for structured learning—seems ill-equipped to absorb change. Due to images evoked of children on a conveyor belt, scholars have deemed this educational structure the “factory model.” The factory model sets teachers at the center of children’s education, relying on educators to produce material for students to consume, rather than focusing on the needs of the individual learning.
Making the most of technology
Some, like Forbes’s Clayton Christensen, hold that technology in classrooms will be the vehicle by which America’s schools transition from what Christensen termed “teacher-centric” to “child-centric” education. Christensen wrote in 2008 that students suffer from rigidity in the curriculum, which is influenced by state mandates, district regulations, and the expectation that all students will master the same skills to the same proficiency level by the time teachers conclude covering the material. These problems, he argued, must be addressed by shifting the education paradigm to a modular one where learning can be self-directed and tailored to students’ learning styles and paces. In this sense, the new educational technology promised by ARPA-Ed will not simply be an accessory to learning but rather will be crucial to changing the system itself.
Consider, for instance, one contemporary school of thought: the “flip model” of education. Inspired by Salman Khan, founder of the Khan Academy, the central concept behind the flip model is that students listen to lectures for homework and come to class prepared to engage in deeper discussion and analytical activities. The Khan Academy provides clear, concise lectures on a wide breadth of subjects, complete with exercises to test students’ understanding—and most of it is free. All of it is online.
In Lawrenceville, Georgia, some teachers at the Gwinnett School of Mathematics, Science, and Technology have traded homework for lectures and use precious class time for dissecting difficult problems and engaging students in group activities like acting out the processes of a cell in a 10th-grade biology class. This line of thought is consistent with emerging research that suggests that even children who prefer to work alone do better when collaborating in groups. Critics of the system, however, are quick to point out that it is rendered useless in the absence of a basic technological framework for online instruction.
Even if ARPA-Ed fosters competition for the best, most exciting educational tools available, implementation in classrooms is another hurdle entirely. Said Bryan Goodwin, spokesman for Mid-continent Research for Education and Learning, “Good teachers can make good use of computers, while bad teachers won’t, and they and their students could wind up becoming distracted by the technology.”
The reality that those who don’t teach well won’t use technology well, either, exacerbates an underlying problem: Integrating technology into schools in such a way that decentralizes the classroom is incongruous with current pedagogical thinking. In fact, some studies suggest that even teachers who are well-versed in new technology tend not to use it in ways that encourage the aforementioned “student-centric” learning.
Like technology, however, the landscape of the teaching profession is evolving. Programs like Teach for America are plucking talented college graduates from diverse backgrounds and honing their instructional skills in underserved communities. Notably, the Teacher Incentive Fund holds promise for attracting and retaining motivated young teachers in high-poverty schools by compensating better for better results—a concept not novel to most other professions in the United States. Having grown up, rather than having had to keep up, with technology endows this new generation of educators with an advantage.
Efforts on the Hill to close the comparability loophole, by which schools obtain federal funding by over-representing teachers’ salaries in less affluent districts, should help alleviate the plague of low teachers’ salaries in schools that need good teachers the most. Finally, the age of teachers’ unions eschewing serious discussions about tenure and performance-based pay has come to a close, as leaders meaningfully engage opposition in the shared goal to protect teachers and secure the best opportunities for more students.
The outlook for ARPA-Ed
But ARPA-Ed still faces opposition from the right. Conservative pundits have called the initiative wasteful and “duplicative,” claiming that $90 million is too steep a price to pay, and that the Department of Education already spent $4 billion on teacher retraining. Advocates like Tom Schatz, president of Citizens Against Government Waste, would rather see that money funneled to the states than appropriated at the federal level. The assertion that ARPA-Ed is just another teacher retraining piggy bank, however, grossly misrepresents the multifaceted, innovative program.
ARPA-Ed breaks with existing programs in that it focuses on research into learning, not teaching. Moreover, $90 million seems like a steal for educational investments designed to yield breakthroughs that could have positive social returns for decades. To put the figure in perspective, the federal government spent $1.34 billion—15 times the price tag of ARPA-Ed—in 2010 on its “Race to the Top” program, in which states competed for funding by implementing the best reforms, like adopting performance-based pay and promoting charter schools. For six times the cost of ARPA-Ed, we fund adult education; still more money is doled out for federal student aid. As it is, the federal government only spends a fraction of its budget on education, and ARPA-Ed’s cost would be a tiny fraction of that.
America’s children deserve the best education science can bring them. Sen. Bennet’s amendment to create ARPA-Ed would be a step toward escaping our educational malaise and helping our students climb back to the top of the ranks. It is through implementing a program that emphasizes core American values like independent learning, creativity, and the individual work ethic, that we can regain our footing in the international sphere and return to doing what we do best: pushing the boundaries of innovation.
Lauren Simenauer is an intern with Science Progress and a senior at the University of Virginia. She is finishing her degrees in biology and psychology. Sean Pool is the Assistant Editor of Science Progress.
Comments on this article