This is where I have to admit to all of my ETS friends that I'm a statistician by profession. Given that, I tend to look at most of these kinds of studies with some level of doubt, and I ALWAYS wonder about the objectives/incentives of those who are doing the studies.

One thing to keep in mind is that this map is simply showing fives levels of mortality. From a county perspective, someone has to be on the high side and somebody has to be on the low side. Does that really mean anything? I haven't a clue.

The data were from a 34 year period. Would the results be the same for the previous 34 year period (not that the data are likely available)? Or from the next 34 year period? If they split the data into three 10 year chunks, would each of the time periods have the same result?

I think I'd be very interested in knowing the range of the mortality rates - from lowest to highest. If its not that wide a range, then this is all of little real significance - just random differences.

If they'd have measured some characteristic that had absolutely no impact from geographic location - maybe something like telephone defect rates - the map could very well have looked just like the one shown. Does it mean anything?? Not really.

In the end I suspect that the numbers of people who die from reasons other than "forces of nature" are so much higher that we're talking about buckets in an ocean here.

Ken

Ken