What is an outlier?

In article <01I30T325HGO8WXHT8@HUSKY1.STMARYS.CA>,
RDAWSON@HUSKY1.STMARYS.CA wrote:

..
>
>What is an outlier?
> I don't know; let's catch a few and see what they are in this case.
>
>Why should it be identified...
> So that we can find out what it is.
>
>..and treated any differently from any other observation?
> Maybe it shouldn't, but let's find out what it is before we decide.
>
>What is the effect on the analysis...?
> That will depend on what the nature of these particular outliers is,
> which we won't know till we've caught them.

-- sounds like a snipe hunt to me...

The problem is that you can't catch an outlier without a model (at least a mild one) for your data. Else how would you know that a point violated that model? In fact, the process of growing understanding and finding and examining outliers must be iterative. This isn't a new thought. Bacon, writing in Novum Organum about 400 years ago said:

"errors of Nature, sports and monsters correct the understanding in regard to ordinary things, and reveal general forms. For whoever knows the ways of Nature will more easily notice her deviations; and, on the other hand, whoever knows her deviations will more accurately describe her ways." [ II 29]
This is my favorite quotation from Novum Organum. Anyone who would analyze data should post it where they can read it frequently.

-- Paul Velleman


Eli Brettler
Last modified: Thu Sep 5 07:10:40 EDT