Show Var(x) = Cov(x,x) for continuous random variable x, using the integral definition of covariance

47 Views Asked by At

We start from the definition of covariance:

$\text{Cov}(X, Y) = \int_{-\infty}^{\infty} \int_{-\infty}^{\infty} (x - \mu_X)(y - \mu_X) f(x,y) \, dx \, dy $

Plugging in Y = X (abuse of notation):

$\text{Cov}(X, X) = \int_{-\infty}^{\infty} \int_{-\infty}^{\infty} (x - \mu_X)(x - \mu_X) f(x,x) \, dx \, dx $

Now, I'm not able to intuitively understand how f(x, x) turns into f(x). Can someone shed some light? Basically, I'm confused as to how the definition of covariance for two different variables generalizes when we're dealing with the same random variable.