FlyerTalk Forums - View Single Post - Distribution of Miles Flown across U.S. Population
Old Sep 3, 2015 | 9:36 am
  #8  
cubbie
10 Countries Visited20 Countries Visited30 Countries Visited20 Years on Site
 
Join Date: Feb 2004
Location: ORD
Programs: AA
Posts: 1,754
Pareto principle. Hunh. Makes it sound like an amazing discovery. 68, 95, and 99.7 percent of anything that is normally distributed lies within the ranges +/- 1, 2, or 3 standard deviations from the mean. As the second chart here shows, 80% lies within +/- 1.3 standard deviations:

https://en.m.wikipedia.org/wiki/68–95–99.7_rule

The first question, I think, should be is the thing of interest normally distributed? In this case, I don't think so. You can have done zero travel by air, but you can't have done less than zero. More likely it's a log-normal (zero-bounded, asymmetric, long-tailed) distribution. Not that the 80-20 approximation wouldn't still be in the right ballpark.
cubbie is offline