=STDEV(data) should be =SQRT(DEVSQ(data)/(COUNT(data)-1) which is supposed to
be less than DEVSQ(data) by a factor of 1/SQRT(COUNT(data)-1).
If you mis-stated the situation, and STDEV(data) is instead less than
SQRT(DEVSQ(data)/(COUNT(data)-1), then I will guess that you are using Excel
XP (2002) or older, and your data has a small CV (coefficient of variation,
also known as RSD -- relative standard deviation). Older versions of Excel
used a mathematically correct, but numerically unstable algorithm that could
give numerically incorrect results with a small CV or a very large n.
It is had to be more specific without more information about your particular
case.
Jerry
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.