MS Excel 11.0 continues to have the same problem it has had forever. For
reasons I haven't been able to work out, it occasionally returnes negative
variance or some positive value that is completely wrong. Based on the
formula used to calculate variance as described in the documentation, a
negative value is impossible. What gives?

"N.C. Arens" <N.C. wrote...
MS Excel 11.0 continues to have the same problem it has had forever. For
reasons I haven't been able to work out, it occasionally returnes negative
variance or some positive value that is completely wrong. Based on the
formula used to calculate variance as described in the documentation, a
negative value is impossible. What gives?

By Excel 11 you mean Excel 2003?

VAR should have been fixed in Excel 2003. What do you get as the result of
DEVSQ applied to your data? If it's also negative, what's your data?