A direct calculation of moments of the sample variance (Q419424): Difference between revisions

Let \(V\) be the sample variance of an i.i.d. sample \(X_1,\dots,X_N\). The authors propose a new method of representation of \(\mathbf{E} V^j\) as a polynomial from the moments \(\mathbf{E}(X_1)^i\). It is based on the representation of \(\mathbf{E} V^j\) as a weighted sum of \(\mathbf{E} \left(\sum_{j=1}^N X_j^2\right)^m\left(\sum_{j=1}^N X_j\right)^n \) and calculation of this expectation directly, using some new combinatorics results. This method can be considered as an alternative to the Polykays rules. The derivation of the Gauss formula for \(\text{Var} V\) is presented as an example.