I have a population that I belive it will be represent a Gaussian (normal)
distribution if I am able to identify and remove the outliers properly. I now
understand I cannot use the Grubb's Test as outlined in a couple of websites
I found (see http://www.graphpad.com/articles/outlier.htm and
http://www.itl.nist.gov/div898/softw...r/grubtest.htm) because I have many outliers and the distribution is not normal.

Based on a discussion I found on this forums (subject: "how do I identify
statistical outliers with excel", dated: 1/7/2007), I decided to use the the
Interquartile Range method (IQR); but the following formula is giving me an
annoying #VALUE! error:

=if(or(D1 < A1 - 3*B1, D1 A2 + 3*B1), "outlier?", "")

I modified that formula so that D1 is the value I am testing, A1 is the
value of the 1st quartile, B1 is the value between the 3rd quartile minus the
1st quartile, and A2 is the value of the 3rd quartile. But I got the same
error message. Then I removed the spaces in the formula, but I still get the
same error. It is very frustating, and I don't know how to fix this.

Any suggestions?