Mapping with Google Insights
David McRaney recently used Google Insights to discover and map the spatial distribution of Google searches involving the search term “sweet tea”. I love the idea, but I wish that Google allowed us to normalize the data against something like the number of total searches made per state or the populations for each. Or does it, in the former case, already normalize the data against the total number of searches for that state? I think it’s just against the aggregate for total searches performed in that country or even for the world as a whole.
My sense is that normalizing the data would lead to a map very similar to the one he produced, but what if it didn’t?What if the normalized map revealed that people in the southeastern United States actually conduct less searches of “sweet tea” per capita? I mean, southerners are more likely to already know how to make it, right? I learned how to make sweet tea from my mother, born in Atlanta, who learned how to make it from her mother, also born in Atlanta, who learned how to make it from the aunt who raised her, also born in Atlanta, and on and on like that for generations. Somebody in, say, Boise, might not have a family member to teach them the wonders of sweet tea. But they have Google: in the first ten search results for “sweet tea”, four are recipes.
A very simple yet effective Flash map of sweet tea locations in Virginia inspired McRaney’s project.
photo credit: ratterrell
1 Comment