A researcher wants to estimate the proportion of students enrolled at a university who eat fast food more than three times in a typical week. Would the standard error of the sample proportion $\hat{p}$ be smaller for random samples of size $n=50$ or random samples of size $n=200 ?$