If x and y are two variables with independent frequency distributions with means μx and μy and variances \(\mathop \sigma \nolimits_x^2 \) and \(\mathop \sigma \nolimits_y^2 \), then the population variance of their difference \(\mathop \sigma \nolimits_{x - y}^2 \) is
1
\(\mathop \sigma \nolimits_{x }^2 +\mathop \sigma \nolimits_{y}^2\)
2
\(\mathop \sigma \nolimits_{x }^2 -\mathop \sigma \nolimits_{y}^2\)
3
\(\mu_x\mathop \sigma \nolimits_{x }^2 +\mu_y\ \mathop \sigma \nolimits_{y}^2\)
4
\(\mu_x\mathop \sigma \nolimits_{x }^2 -\mu_y\ \mathop \sigma \nolimits_{y}^2\)