Years ago proposals to the National Science Foundation were simply to conduct scientific research. But, as has often been the case when science is funded from the public, there were complaints about the relevance of science, etc. And so NSF rolled out a requirement that proposals also include a section addressing the broader impacts of research grants, probably to make science more criticism-proof. At first, this seemed mainly an information request–what other good might incidentally arise from doing this science. But as time has gone on, this has become grounds for rejecting otherwise acceptable scientific proposals. And GG will argue that this is mistaken (though he has suggested some more realistic claims that could be listed as broader impacts).
Now some will defend the broader impacts requirements and argue that they are in fact too weak (Small Pond Science, for instance, delves into broader impacts quite a bit). In a way, we are agreeing on impact while disagreeing on a cure. Certainly there are lots of reasons to encourage the dissemination of science and to encourage scientists to interact with communities. But the reality is that an awful lot of broader impact work is, well, awful. For instance, many scientists decided that a good outlet would be to train teachers–but not knowing how to train teachers, not having run workshops or prepared educational materials, having no concept of how to measure success, their efforts fell flat on their face. Other efforts had outreach in the forms of unread blogs or poorly attended lectures (full disclosure: this blog has never been claimed as a part of broader impacts in any proposal).
So what is the result? Scientists aren’t stupid and aren’t interested in wasting their time: trying outreach and flailing, they look for help to do it better. We now have PIs partnering with professional outreach staff on proposals to more successfully do some outreach. Proposal budgets, of course, have to go up while the total NSF research budget has been slowly declining since 2009. While this is a boon for outreach staff, is this really the intent of funding the National Science Foundation?
And just how effective has been the outreach that NSF has piggybacked onto grants? GG is sure there are some big successes, but is willing to wager that most efforts produce nearly nothing. (Realistically, the biggest summed impact is training of a new generation of STEM scientists). Without some measure of success, we don’t know if money is well spent doing this or not–and further diluting the research budget by demanding such accounting is not a great solution. Furthermore, this is increasingly encouraging scientists to exaggerate findings to get more press (which is of course a form of outreach), a form of activity that is driving more and more scientific misbehavior.
Arguably one appearance by Neil DeGrasse Tyson or Bill Nye on a late night talk show is a more effective form of outreach than hundreds of NSF-supported broader impact activities. And, probably, a lot cheaper.
Here’s an idea: maybe we should simply directly fund people whose jobs are to disseminate science and omit such funding from regular NSF grants. This does a few things:
- Removes research scientists from developing outreach projects that are not their specialty.
- Keeps research scientists from trying to evaluate outreach components of proposals in mail and panel reviews that they really can’t evaluate.
- Allows for substantive outreach efforts to be compared with one another and (ideally) be evaluated by people with expertise in these efforts.
- Permits prioritizing some forms of outreach over others (e.g., outreach to underrepresented communities).
- Allows for adding in an evaluation component to outreach efforts that would be awkward or unworkable in the more numerous small outreach efforts currently underway.
- Permits the public and Congress to see more directly where money is being spent.
Perhaps you have this program stay within NSF; perhaps you put it in the Department of Education. Maybe you still require grant recipients to list whatever outreach they happened to do in their reports (e.g., student development, incorporation of materials in classes or textbooks). Maybe you continue to demand actual outreach sections for larger proposals (say with total budgets over $1M); either they partner with an outreach proposal or have their own substantive outreach component as at present.
But lets get real. When you decide to build a road at public expense, there is not a requirement in there that the road builders do public outreach; you are hiring them to build a road. Sure, some might decide to encourage a school group to come through, or they might host a group of environmental planners, or they might sponsor the Little League team, but you know, you want the best road in the end, anything more is gravy. Let’s have the National Science Foundation seek the best science. And then maybe help Ken Burns make a documentary about it, or pay for Bill Nye to tour the late night shows, or pay for new displays at museums incorporating the new science. In the end, we’d get better science and better outreach.