Much like with voting maps, I bet if the study was done on a per county basis we’d see a similar pattern with a wide range in each state.
Bingo; all of this type geographic statistic should be on a county basis, as the dense urban populations can always prejudice the results across a state. And counties are generally large enough, but isolated enough, to substantiate a local trend. For instance in Nebraska because of the overall low population density, most of the counties would show a similar lifestyle, except like a poster above says, Lincoln and Omaha.
At The Suntrade Institute we generally use counties, the compiled statistics are generally available, and far more telling.
I would indeed like to see a map of the happiest counties in the USA.