Quote:
Originally posted by JSDSKI
The numbers drop dramatically after the ban. I choose to interpret that to mean the ban had an effect on the numbers. You choose otherwise. To what else should we attribute the change?
|
Sigh. If you look at the numbers you'd see they aren't a nice smooth curve or line. 1996 had a spike in deaths - perhaps a school yard massacre or something that year? Because the variability in data, one has to look at trends, not at a small part of the data to the exclusion of the other data. This is the sort of thing one learns in statistics class so it seems obvious to me but I guess it isn't to everyone.
Here's an example of how this sort of thing works:
Over the course of a month in the fall, average temperatures will go down. Some days will be unseasonably warm and others will be more winter like. The
trend is for the temps to go down though. If a law was passed to restrict green house gas emissions on an unseasonably warm day, one cannot conclude that it was effective just because the days after that continued the cooling trend.
-Chris