I agree with this too. Its easier to map a food desert than disease rates, and its easier to involve a community in the food source mapping than in the diabetes data. .
If your data don't include 'time in residence' then its useless for this type of analysis for policy action. Further, if it does, there's only some small guarantee (from cancer cluster work) that you can make some factor or risk from such data. That is: how do you make a metric for monitoring?
Trying to find it again. Haven't succeeded. I was pretty sure it was a link to something by Seth Godin whose blog is ostensibly about marketing, but bleeds over into many other areas. He's a bestseller with books like "Tribes" and "The Dip" which are about more than marketing - building social movements, the changing face of the market, etc. He also talks a lot about non-profit management and marketing. his posts tend to be quite brief, but they usually link you to more in-depth and often very useful info.
I'll keep poking around and see if I can locate it. The article was brief, but its got me thinking. His main point was, in this age of overwhelming information, we should not be shy about moving forward with what we know. Its easy to get paralyzed feeling that you need more and more data, but as far as non-profits, his point was that we usually have more than enough info to work with and push the inquiry in the right direction.
Thanks for the heads-up to Seth Godin's blog - regardless of whether I can locate the source article, the idea (proceed with what you got!) is a nice reminder - appreciate the info!