> I disagree with Chris, although his answer is one that many statisticians > would give.
> So, I would summarize a symmetric distibution with 2 outliers as symmetric > with 2 outliers -- even if I didn't know a reason for the outliers' > deviation from the rest of the data. > > For teaching, I think students should know that not all statisticians and > data analysts would agree on this. >
And, teachers should know that some statisticians and data analysis speak with authority and others (moo-wah comes immediately to mind) speak with implied question marks at the end of every sentence! :) :)