Would it help to remove every outlier from the original data set and compute after that?
In reply to Re: A lesson in statistics by kyle in thread A lesson in statistics by 0xbeef