Was looking at this link and wondering why for population covariance, the denominator is n, while for sample covariance, denominator is n-1. How does this 1/n(n-1) replace the 1/n^2 in the proof in this link? Covariance of two sample means

  • $\begingroup$ For other aspects of this question, please see the hits for variance divide n. $\endgroup$ – whuber Oct 7 at 15:47