Dominic was saying that the issue is not sumproduct, but your expectation,
which seems to defy basic math.

Your sumproduct formula apears to be fine and calculates c1/n1 for region 1,
where c1 is the total of yes's and no's for that region and n1 is the number
of stores. Your divisional percentage is then
(c1+c2+c3)/(n1+n2+n3)
Why do you think that should be equal to (c1/n1+c2/n2+c3/n3)/3 unless
n1=n2=n3?

Also note that Excel (like all software that follows the IEEE double
precision standard) is not capable of 30 figure accuracy, since double
precision can only represent 15-16 digits reliably (hence Excel's documented
limit of displaying no more than 15 digits).

Jerry

" wrote:

That is my problem! I am NOT a sumproduct pro also... I placed out the
decimal point to 30 and it just does not come out correct. To my
knowledge there is not a mathematical issue here, so it must be in
Excel or most likely just in sumproduct. Does anyone else out there
have a clue as to why this happens???

Dominic thanks for your help though!


Hans