To get the variance and covariance the following steps are taken:
In the step below to calculate E[Z^2] how do we approach this without a known pdf? For completeness would finding E[cos(theta)^2] be the same set of steps? But this is different because at least we know that theta has a uniform distribution?
Maybe what I'm looking for is a trick, theorem or a property? I'm not completely sure how to fill my misunderstanding in this single step. Otherwise everything else in the problem makes complete sense.