4 replies to this topic

[Gretchen Peterson]

Here's a little analysis I did on the USDA farmers' market database in combination with CDC obesity data:
http://gislounge.com...rc3zU9s.twitter

[David Medeiros]

I have to admit I'm a little troubled by some of the aspects of this projects design. Primarily the use of county area as the normalization for farmers market rates. I would think both statistics (obesity and farmers markets) should be viewed as per population.

I'm also not sure about calling it a "study", that implies a systematic, scientific review of the topic. I would expect a study to be accompanied by more detailed information about the issues being examined and the methodology, with probably a handful of potential correlational relationships to examine and attempt to negate instead of just one.

I appreciate the promotional value of the work though, but I have to think that potential clients will have the same questions I did. I say this having spent quite a lot of time working on a similar correlational map project involving river obstruction sites and river species diversity in Oregon. Its still in progress, but my biggest obstacle is finding legitimate analytical questions to answer rather than simply overlay my various datasets.

[Gretchen Peterson]

The obesity data are already characterized as a rate per population. However, your point about normalizing the farmers' markets by population rather than by area is something I will explore this afternoon.

[dsl]

Ooops, should have read this first. I also thought maybe markets per capita might be a useful measure. Seems like it would be a pretty good indicator of healthy living. Neat idea.

Cheers,
David

[Gretchen Peterson]

I addressed some of the concerns over on the original article and also made the data available for anyone who wants to work with it. Thanks for your feedback.