AppleJordan wrote:Okay, I know there are plenty of people here who have had statistics class, so I'm probably wrong.
But I thought that there can't be any points outside the "wiskers"! Isn't the graph used to show the distribution of a set of data over a number line?
That's what I thought, but upon further investigation, I found that a point can be placed outside of the whiskers if it is very far from the next term in the set. The reason given is that if the whisker was extending, it would give the false impression that there was a distribution over that range, instead of the single outlier.
For example, if you have an outlier of 100, and the next largest term as 20, to extend the whisker all the way to 100 would imply that that there is still an upper quartile range through that whole area, whereas if you cap the whisker at 20, with a point at 100, it becomes very obvious that you have one extremely different term, and the quartile of more similar data should be capped off at 20. Of course, if you had 2 or three 100's, perhaps the whisker should be extended out...
But apparently she is the only one
This is why I never took statistics.