NextGen Evidence

Applying a Next Generation Evidence Approach to DARPA for Education

By Kelly Fitzsimmons
February 24, 2023

This is a repost of an op-ed originally published in The74. Click here to view the original article.

Recent headlines on the state of the nation’s education system have been grim. Math and reading proficiency have fallen. Schools face staffing shortages, increasing mental health challenges and classroom politicization. While the federal government has poured money into schools in the wake of the COVID-19 pandemic, funding has been insufficient and too poorly targeted to address learning loss, retain competitiveness in the global marketplace and meet the needs of today’s students. 

The recent $1.7 trillion federal omnibus package provides hope for improvement, with $70 million allocated to the Institute for Education Sciences to pilot a program akin to DARPA — the Pentagon’s hub for research and development that have produced innovative technologies like GPS, the internet and speech recognition software. A DARPA for education could help practitioners test and implement new approaches to teaching and learning, and understand what interventions are most successful.

For this program to succeed, the government should take what Project Evident calls a Next Generation Evidence approach. Informed by five years of work with practitioners, funders, researchers and policymakers, and signed by over 60 field leaders, the Principles of Next Generation Evidence can provide a helpful framework for design and implementation. Specifically, this approach would mean doing the following:

  • DARPA for Ed must work to understand the needs of education practitioners — teachers, principals, superintendents, and staff at state and local education agencies and nonprofits. Too often, funding programs have failed to take practitioners’ challenges, times and limited staff capacity into account, leading to unintended consequences like schools struggling to spend pandemic relief dollars on time or not making hires for fear of a fiscal cliff. Furthermore, funders have been quick to jump from one approach to another in search of a magic bullet, while ignoring the less glamorous but critical work required for thoughtful implementation. This new program should support practitioners’ ability to effectively absorb funding and implement plans, and be realistic about the time it often takes to see measurable improvement.
  • The pandemic has widened the opportunity gap and left vulnerable students further behind. New innovations must leverage data and evidence to ensure all students can reach their full potential. This requires considering equity both in how data and evidence are generated and in the outcomes sought. This includes understanding participation and outcomes by different groups; engaging practitioners and families in identifying challenges and designing solutions, and considering how racial and economic inequities impact learning.
  • Students and families must participate in developing new solutions to ensure they are of real value (in a way that empowers them but does not cause undue burden). Rather than simply doubling down on preparation for standardized testing in response to NAEP results — which many consider an outdated and narrow way to understand overall student success — DARPA for Ed should encourage practitioners to work with communities to define what success looks like for students, and to develop strategies for how it is achieved and measured.
  • The education sector has long underinvested in research and development. DARPA for Ed is a step in the right direction and should support practitioners in developing processes for learning, testing and improving, and promote the exploration of new technologies such as ethical artificial intelligence.
  • Funders and policymakers often equate evidence with research results alone and discount other types of useful data, including administrative (such as absenteeism rates) and qualitative (such as student voice). DARPA for Ed should encourage practitioners to leverage the wide variety of quantitative and qualitative data at their disposal to strengthen decisionmaking and outcomes.

Across the country, there are promising instances of practitioners taking a Next Generation Evidence approach. A few examples:

At the early learning platform Noggin, R&D is a key component of the content production pipeline. A cross-disciplinary team of content developers, instructional design experts and research scientists have developed a process for rapid-cycle research that allows them to continuously improve content as it is being developed and to quickly and economically test for evidence of learning impact. 

Baltimore City Public Schools and Project Evident partnered to strengthen career and technical education programming by blending available data sets to conduct root cause analyses and facilitating over 100 stakeholder meetings. As a result, Baltimore was able to make rapid, scalable improvements to its CTE curriculum, resulting in stronger and more equitable programming. 

The education nonprofit Digital Promise is building an AI-powered recommendation engine that uses school and student data to connect teachers with professional learning opportunities tailored to their students’ greatest needs. Digital Promise initially approached AI with skepticism, unsure about the resources required and concerned about potential bias. However, it has come to view AI as a powerful tool that will allow for the expansion of high-quality supports to more schools.

The capacity to do the type of work highlighted above should be the norm for nonprofits and school districts, rather than the exception. DARPA for Ed has the potential to help make that happen. Adopting a Next Generation Evidence approach can help ensure the program’s successful and sustainable implementation and lead to stronger and more equitable outcomes for students.