>I need a short and sweet lesson here. I would like to do some very basic statistics on data. Say the values are: 5,21,26,17,28,31,36,41,29,90. What is an acceptable way to reject highs and lows mathmatically. The extremes throw off an average. What is the difference between mean and average? Any formulas or help would be greatly appreciated.
>
Hmmm...well, the big question statistically is usually whether you can assume data is approx normal and use that (actually the 't' distribution which is similar to normal). Otherwise probably use non-parametric methods, which I won't go into. If you can use normal approximation, see my reply to Arnon, a 95% interval is often good in real world situations...(mean +/ 2*std dev). If you have SAS or something similar, you can get nice listings of statistics, outliers, etc.
But careful rejecting measurements that appear to be outliers unless there's a good reason.
The Anonymous Bureaucrat,
and frankly, quite content not to be
a member of either major US political party.