god, I'm good on questioning.
This is a concept I still don't understand.
When we say degree of freedom is n-1, as we deal with a problem of a sample mean, such as mean=(x1+x2+x3........+xn)/n. Just because the constrain of x lies in a one demensional space.
But why do we have to use n, and then minor the 1.
I'm no statistical theory guru, but I know degrees of freedom is not an easy concept. In Mood, Graybill, and Boes, 3rd ed, p244, in a discussion of the chi square distribution; the sum of squared deviations from the mean for a normal random variable has df = n-1 because the sample mean = 0, which allows one to determine the nth deviation from the other n-1 so only the n-1 deviations are free to vary. I know, clear as mud.