Extension of Continuous Mapping Theorem for Bounded in Probability

288 Views Asked by At

The continuous mapping theorem states that if $X_{n} \to X$ in probability and $g$ is continuous almost surely then $g(X_{n}) \to g(X)$ in probability.

In little-O notation this is $$X_{n} - X = o_{P}(1) \Rightarrow g(X_{n}) - g(X) = o_{P}(1).$$

Are there any extensions for bounded in probability? More specifically,

  1. If $X_{n} - X = O_{P}(a_{n})$, what are conditions on $g$ which give $g(X_{n}) - g(X) = O_{P}(a_{n})$?

  2. If I have a sequence of functions $g_{n}$ and $X_{n} - X = O_{P}(a_{n})$, what are minimal conditions on $g_{n}$ to give $g_{n}(X_{n}) - g_{n}(X) = O_{P}(a_{n})$?

This is for self study so textbook references where I can read further are appreciated.