A New View of Statistics

© 1997 Will G Hopkins

Go to: Next · Previous · Contents · Search · Home

Summarizing Data:
Simple Statistics: THE SPREAD continued

 Percentile Ranges

The most common of these is the interquartile range, although even this is a seldom-visited feature on the statistical map. It is used with the median to give an idea of centrality and spread of skewed or otherwise grossly non-normally-distributed variables. Measures of training are often skewed enough to merit use of percentiles instead of the mean and standard deviation. For example, weekly training in a group of novice athletes might have a median of 5 and an interquartile range of 3-12 hours/week.

Here's something challenging for the real lovers of numbers. The mean ± SD encloses 68% of the data on average for a normally distributed variable. So if you want to use a percentile range that corresponds to the mean ± SD, what should it be? Answer: 16th-84th. If I had my way, this measure would replace the interquartile range. We could call it the standard percentile range…

We're right out on the horizon now. Let's get back to familiar territory.

Go to: Next · Previous · Contents · Search · Home
resources=AT=sportsci.org · webmaster=AT=sportsci.org · Sportsci Homepage · Copyright ©1997
Last updated 29 April 97