Given a dataset $X$ with mean $\mu_X$ and variance $\sigma_X^2$, what is the expected value of the sample variance $S^2 = \frac{1}{n-1}\sum (x_i - \bar{x})^2$ in terms of the population variance $\sigma^2$?