« Bayesian Inference | Main | The Space Shuttle Problem »
September 20, 2005
n vs. n-1
We tried to understand the mysteries of n vs. n-1.
We divide by n-1 when we calculate the sample variance because that produces an "unbiased" estimate of the population variance.
In calculating the sample variance, one "degree of freedom" is used up calculating the sample mean. This is "obvious" if you consider a sample of two numbers. One degree of freedom is the sample mean, and the other degree of freedom is how far each number is from the mean, which is precisely in the middle.
Posted by bparke at September 20, 2005 10:05 PM