Editors’ Note: This is the fifth blog post in our series featuring reflections on and learnings from the 2016 Carnegie Foundation Summit on Improvement in Education. Each post provides recaps of sessions you may have missed and further insight into presentations you may want to revisit.
Conversations in education have shifted in the last few years towards a new way to think about change efforts. For a long time, the focus has been on identifying and spreading “what works,” too often leading to the search for “silver bullet” programs or initiatives that rely heavily on implementation with fidelity for success.
Many involved in educational improvement have begun to realize that context (broadly conceived) matters, and are now looking for ways to learn how to shape programs to local conditions to enable success. These new approaches place a heavier emphasis on how good ideas get implemented — on the processes and mechanisms that make effective change possible across varied contexts.
At the recent Carnegie Foundation Summit on Improvement in Education, there was much discussion about what these processes and mechanisms look like in practice. One session in particular, Why Networked Improvement Communities (NICs)?, convened a panel, moderated by Paul LeMahieu of the Carnegie Foundation, that included Anthony E. Kelly from the National Science Foundation (NSF) as well as Art Seavey and Ash Vasudeva of the Bill & Melinda Gates Foundation. They shared their thoughts about how NICs, as scientific learning communities, integrate essential and promising mechanisms for change to accelerate educators’ efforts to improve.
The Carnegie Foundation, under the leadership of Tony Bryk, has been promoting the idea of networks in conjunction with the use of improvement science to help the field of education get better at getting better — learning faster how to get promising ideas to work effectively and reliably across contexts. NICs, in particular, have been treated in this space previously.
At the Summit session, Ash Vasudeva of Gates noted early on in the panel discussion that networks are just one part of a larger improvement strategy, but a very powerful way to help improvers learn what else needs to “wraparound” a promising idea to make it work and spread well. He went on to describe how funders have to start “getting out of the middle” (e.g., taking on the operational responsibilities for dissemination) by supporting new improvement strategies that promote coherent and deep investment in very promising solutions, while also providing a structure for these ideas to spread.
One example that Vasudeva offered concerned an effort to improve college readiness. While the work had shown some isolated progress, it became clear that it would take much more effort to get the few successful instances to move — and that this constituted a wholly different challenge than what had been done to realize success in one or two places only. “We have so much more to learn than we know now,” he explained.
Anthony Kelly (NSF) added to this example, noting that it took 50 years for the idea of Kindergarten to spread across the U.S. “Just because [something is] better doesn’t mean anything,” he explained. Every community has its own culture, incentives, practices, and beliefs. Therefore, it’s essential to first test ideas in new communities in order to facilitate the spread of innovations, as spread doesn’t just happen on the merits of an idea alone. Networks do a great job of creating the structures and conditions for diverse organizations and communities to engage with ideas on a small scale as well as observe and learn from them being tried in other places before trying to take it to scale.
Spread doesn’t just happen on the merits of an idea alone. Twitter
Kelly went on to assert that a NIC is not just a community that’s “smarter and more connected.” He instead emphasized that, when done well, these kinds of communities are “agents of change.” The panel discussion went on to highlight five main ways that NICs promote widespread change:
- They bring together a diverse community that is focused on both a common goal and a common approach (i.e., theory of improvement).
- They utilize a data infrastructure that informs the community early on whether something is or isn’t working.
- They break down the norm of improving in isolation.
- They add an element of discipline to powerful, but what can otherwise be incoherent collaborative efforts.
- They make varied contexts available for testing, in order to learn what it takes to get ideas to work well everywhere. These varied contexts also allow for the detection of interesting patterns, including both bright spots and trouble areas that can be learned from. Collectively, they offer the greatest potential for learning “what works, for whom, and under what conditions.”
Vasudeva added to this list of benefits provided by NICs by noting that this new approach might not only lead to deep change that spreads but also builds strong scientific learning communities that can be buffered from leadership churn and political shifts that may otherwise disrupt promising change efforts.
Improvement work should and does work best when negotiated from the bottom up and networks facilitate that well.Twitter
The panelists offered an important caveat as the discussion continued: no idea is perfect, and NICs are no different. They were quick to note that this model is not something they would ever dictate to a grantee, but it’s important to bring evidence to bear and allow groups to have access to these ideas to see if they agree it’s the right approach for them. Vasudeva and Seavey of Gates added that there is a lot of excitement about these ideas among their grantees and Kelly of the NSF explained that this is likely due to the energizing culture that networks build, allowing people to do their own work to contribute to the larger good. He shared powerful examples of how networks have enabled scientists to pick up and solve problems that would have otherwise been abandoned. The takeaway was that improvement work should and does work best when negotiated from the bottom up and networks facilitate that well.
In response to follow up questions from the audience, the panelists shared two challenges to this work: time and the necessity of building an effective network hub. Seavey raised what he called one of the biggest challenges of NICs and networks in general: they take time to build. This work requires patience, and funders tend to get caught in the middle between grantees and policymakers when it comes to wanting immediate solutions to urgent problems. His call was for “patient optimism” and Kelly added that open communication about the work happening at each level is essential to and effective at building trust among all stakeholders to sustain these kinds of efforts.
Time is also important to consider in building the hub of a network, but Vasudeva posed a question specifically concerning the kinds of organizations that are well positioned to do this work. He noted that few are especially designed to provide the “connective tissue” that is required to shift from the more typical one-to-one work to working in a coordinated and coherent way across a widely distributed network. It’s important to focus attention and resources on setting up the right communication and learning structures in place to prevent networks from stagnating at launch and staying small and ineffective. Organizations that have taken on this role as hubs, however, have been able to transform how they think about their work connecting and engaging people and organizations around important problems and scaling promising practices to solve them.
By transforming the way individuals and organizations work together, energizing communities, and establishing the sustainable learning systems and approaches necessary for successful change efforts, networked improvement communities really can accelerate educators’ efforts to improve. Bringing collaboration, innovation, and the discipline of improvement science together has started to build a proven path to make possible change in education that is not only effective, but reliable at scale.