Let be a functional that is continuous in at .
By the Glivenko–Cantelli theorem, almost surely, so almost surely, hence also in probability.
(b)
(i)
The mean functional is not continuous.
Consider and
Then has mean , has mean , and .
(ii)
The Cramér–von Mises functional is continuous.
For this, see that
(iii)
The quantile function is also continuous, but only if we assume that for all .
Note that
Let .
Then
Thus, we desire to show that
Alternatively, we could show that
This turns out to be easier.
First, we will show that .
Note that
Since, by assumption, (the strict inequality is important here),
we can make large enough to find .
The other inequality follows similarly.