More importantly, in one form or another they all point to the program’s structure as a research funder rather than an infrastructure builder as the root cause. I concur. The realities of life in the http://www.fastlane.nsf.gov are omnipresent, but the NSDL community has a strong shared vision of improving STEM education, and I firmly believe that the same people could have created a spectacular success.

The barrier, according to the reflections, has been a systemically caused inability to follow a unified strategy with clear goals and outcomes. Maybe that is true, but it could equally be a very different systemic problem. In the world of software and entrepreneurs, the mantra is: “fail quickly,” or “fail early and fail often.” This is not overly compatible with competitive grant-based funding. No one ever continues to receive grants for reporting that they dropped $150K in six months on something that turned out to be a very bad idea, yet that has to happen to uncover the one, simple, compelling value proposition that can turn a garage into a Google. Of course, you also have to recognize the good ideas and run with them as fast as you can.

Permit me, then, to pose a question. Can the NSDL, with the hand that it has been dealt and the culture that it has evolved, learn to “fail quickly” and if so, how?

Looking forward to responses.

Steve Mitchell
Science Library
University of California, Riverside

As a participant in NSDL for a couple years, I found this article interesting but not terribly coherent. The following quick comments, regarding NSDL and its future and challenges occurred to me as I read it:

NSDL embodies more potential now than it ever has had. It resides at what is the nexus of the President’s emphasis on education, science and social responsibility. NSF has a huge stake in improving science education at all levels and in improving public transparency and participation in setting science research priorities. NSDL is one of the few NSF efforts where this kind of thing can be encouraged and grown to an appropriate scale/scope – which is national.

Relatedly, a problem NSDL seemed to experience was that of participating individual institutions not really fully committing to a national level effort. This was something NSF was not effective at encouraging but could be effective at if it so chose. As a result we are left with digital education efforts where an institution preserves its “brand” at all costs but whose effort is often very redundant/diluted among those of hundreds of similar institutions and projects. These, in following, often lack the heft/resources to scale properly and to take advantage of the Web as a national and global sharing mechanism. Consolidation with larger efforts such as NSDL is politically hard at the home base. This is why NSF as a whole, to encourage stake holding in NSDL, would have to demonstrate serious long-term backing of NSDL, which it didn’t do…but could do.

I saw a quip quoted, in a comment on the article, to the effect that NSDL was like a start-up company without a CEO. And while there is some truth to this, it is also true that projects pioneering Internet services and attempting to use the Internet to its fullest potential in distributed education and research must develop new models of management that have novel mixes of centralization, de-centralization and fluidity. This is a trial and error process with which NSDL has bravely engaged and gained much wisdom, through concrete experience, about what works and what doesn’t. This knowledge is priceless and should contribute to a successful NSDL as a continuing effort – assuming it becomes adequately funded. Anything good and substantial takes time. It also takes proper funding and long-term commitment.

Really, NSDL must be viewed, in its current state, as a successful ongoing pilot given the great (perhaps overly ambitious) scope of what NSF wanted to achieve with it and given levels of funding that weren’t adequate to this scope. Note that, again given its ambitious goals, if the start up analogy was to be extended, NSDL was vastly under-capitalized, especially in its early years, as its course was being charted. Under Obama this situation could change quickly. Note that Ning, a just out of the blocks social networking tool with similarities to NSDL, has been fueled on over $100 million in startup money for the last two years alone. Compare this with NSDL’s $150 million, going to NSDL project’s and communities in almost every state, over eight years. With a proper budget to publicize and grow NSDL, its usage would be great. Again, reading the article, why doesn’t NSF fund ongoing science education? Why doesn’t it fund ongoing proper publicity for its projects? What are the hurdles? Can’t they be overcome — especially now in a political climate solidly favoring, in fact emphasizing, science and education innovation at every level?

A last comment on researcher/faculty searching that is increasingly Google reliant — since this is addressed in the article: I am a science librarian with two decades of experience and can simply say that, with some exception, faculty searching is often unschooled, even pathetic. Much is missed and this undoubtedly reduces the quality of and greatly increases the cost of doing science. This level of searching has been carried forward into the Internet and Google. In many ways Google has reduced the problem but in an increasing number of ways it is actually contributing to the problem. Researchers assume uses for it that are completely inappropriate. As a searcher who uses Google several times a day to answer science inquiries, I can say that it is great but only when seen as one tool among many others. I can also say that search success using Google, which attempts to be all things to all audiences/users, is getting noticeably worse. What I (and my patronage) need is increasingly diluted and drowned out by the simple presence of so much “muchness”. It is apparent to me that, if thoroughness in research is important, then those portals, like NSDL, which contain expert vetted content, may shortly offer the only alternative in effective Internet searching. If you haven’t discovered this for yourself, in your areas of interest, I wager that you will shortly. Again, NSDL needs a proper budget to be able to reach an appropriate threshold of national visibility and service. It needs the resources to educate and develop its potential user base.

As a last comment re: NSDL, it would be foolish for NSF to throw the baby out with the bathwater. The cost of doing this, of putting all our science education eggs in a basket like Google’s, whose approaches and algorithms may have reached their limits (note that to the degree that these algorithms may continue to scale probably implies that they will be trained/piggyback on the expertise inherent in the resource descriptions contained in expert created/described collections) could be immense. It is critical to have a multiplicity of approaches in finding tools and collections aiding science education and research and NSDL would be one of the most important of these.