Search This Blog

Friday, February 1, 2013

That's Some Bad Epidemiology

One doesn't look for cases of just about anything and then draw a circle around the area.

One defines what biological endpoint one is looking for, then determines how large an area to look in, breaks up a larger area into smaller areas in an objective fashion, and then makes comparisons.  One also has to use confidence intervals (ie, one small area may have a larger number of cases of a certain biological endpoint due purely to chance).

No comments:

Post a Comment