Great points. One quick response that relates to how NSF and the academic culture do business. You are right that researchers are often incented to (re)invent rather than reuse. There is no way to change that culture over night, but you note a few ways that NSDL is already helping to solve this problem. IMO there are also a number of mechanisms funders, including NSF, can use to shift incentives. To begin, the Foundation and its reviewers need to know when proposals are really reinventing, not inventing, and decline (or redirect) them. Of course this assumes improvements in managing the knowledge the field is accumulating, so that everyone has easy access to the current state of the field (IMO NSDL could play a role here). Then I think NSF also needs to actively encourage reuse, telling a prospective PI, for example, not to build yet-another-strand-map tool, but rather to build a new service on it, or apply it in a particular area. For many programs I think “building on” or “applying” grants should be at least on a par, in terms of funding dollars, with “building new” grants. There is a precedent for this in the old CCLI adaptation and integration track that only funded work that built on previously developed educational materials.

You also touch on the community-based governance in large projects at NSF. I’m all for input from the community on key decisions that might affect, for example, the overall direction of the initiative or the design of the technical infrastructure. But not if the process unwinds without end; at some point the cost in terms of time exceeds the benefits in terms of buy-in. Perhaps that’s a good topic for a future blog posting!

Hallelujah! I couldn’t agree more that NSDL deserves a single, robust Strand Map Service that can support teachers and learners to locate and comprehend educational resources, and support cognitive scientists and other researchers to better understand the benefits of semantic-spatial displays on users’ cognitive processes. I think there are a couple of reasons why services like this appear to be adopted slowly relative to cultural or business phenomena like YouTube.

First, within the NSDL and NSF university-based research communities, there is a very strong self-help mentality. We are rewarded to invent new things, not to reuse things created by others. I think the NSDL program has made tremendous inroads in building a culture of collaboration and sharing and getting projects ready for the idea of building portals and learning environments from shared services, like the Strand Map Service.

Second, there are no standard ‘business’ mechanisms to reach, or pay for, service level agreements within NSF university-based research communities. Until recently, it was a leap of faith for another institution to build significant services on top of the Strand Map Service because there were no assurances that we would be able to sustain the operations of this web service beyond the life of the grant. Likewise, trying to put this Service on a more commercial footing, like a purchased service, would have been very difficult (impossible) within a grant-driven culture, as groups have to be able to predict at the time of writing their proposals which services they will use and how much they will cost. That is, there is no economic flexibility in the system - groups are committed for years in advance to a specific set of deliverables on a fixed (and fully accounted for) budget. Again NSDL has done the community a significant service by providing for a centralized operation, Core Integration, that can receive these individual innovations and commit to reliaby supporting them for a reasonble period of time. Without this type of support committment by CI, there will never be large-scale reuse of services in NSDL.

Finally, concerning your broader comment about the speed of the maturation of YouTube with respect to educational technologies, it all comes down to flexibility and money. YouTube is an interesting example. They received $11.5 million in venture capital in 13 months from Sequoia Capital. This is equivalent to the entire annual budget of the NSDL program. They also have a streamlined management structure and are not accountable to community processes and community governance, which are expensive and time-consuming. In short, they can do things at fundamentally different scales and speed; they have a much more nimble decision-making structure, and two orders of magnitude more money to innovate and market with. I have been thinking of late about the tendency of NSF to adopt community-based governance processes for large projects. The benefits of this approach are buy-in from NSF’s bread-and-butter consituencies of university-based researchers. The downsides are that for many educational innovations, this constituency is not the same as the intended user audience. Thus, projects are having to spend a lot of cycles and money to engage the research community, instead of focusing on satisfying their customer needs. As you say in your post, I am not sure how to resolve this dilemma.