1
$\begingroup$

Let $(X_n)$ be a sequence of random variables and $(a_n),(b_n)$ be two sequences of non-negative real numbers such that $a_n\downarrow 0$ and $b_n\downarrow 0$ when $n\to\infty$.

If for any $t>0$, $$ P(|X_n|\geq a_n+t)\leq b_n, $$ can we conclude that $X_n\overset{P}{\to}0$ as $n\to\infty$? From the hypothesis I know that for any $t>0$ $$ \lim_{n\to\infty}P(|X_n|\geq a_n+t)=0. $$ But I do not see why the fact that $a_n\downarrow 0$ implies that $$ \lim_{n\to\infty}P(|X_n|\geq t)=0. $$ I was trying to use Slutsky's theorem, but I don't know anything of the convergence of $|X_n|-a_n$. Other thing is that for $n$ sufficiently large $a_n\leq \epsilon$ for any $\epsilon>0$, so $$ P(|X_n|\geq a_n+t)\leq P(|X_n|\geq \epsilon+t), $$ but that doesn't help either.

Any suggestions?

$\endgroup$
1
$\begingroup$

Fix $s > 0$. We want to show that $\mathbb{P}(|X_n| \geq s) \to 0$ so fix $\varepsilon > 0$ and we aim to show that for large enough $n$, $\mathbb{P}(|X_n| \geq s) < \varepsilon$.

By the assumption applied with $t = \frac{s}{2}$, $$\mathbb{P}(|X_n| \geq a_n + \frac{s}{2}) \to 0$$ Combining this fact with the assumption that $a_n \to 0$, there is an $N$ such that $n \geq N$ implies that $a_n < \frac{s}{2}$ and $\mathbb{P}(|X_n| \geq a_n + \frac{s}{2}) < \varepsilon$. This means that if $n \geq N$, $|X_n| \geq s$ implies that $|X_n| \geq a_n + \frac{s}{2}$. Therefore, for $n \geq N$, $$\mathbb{P}(|X_n| \geq s) \leq \mathbb{P}(|X_n| \geq a_n + \frac{s}{2}) < \varepsilon$$ which gives the desired convergence.

$\endgroup$
1
$\begingroup$

Let $\epsilon>0$ be fixed and let $t=\frac{1}{2}\epsilon$.

Then for $n$ large enough we have $a_{n}+t<\epsilon$ so that $P\left(\left|X_{n}\right|\geq\epsilon\right)\leq\left(\left|X_{n}\right|\geq a_{n}+t\right)\leq b_{n}$.

This leads to $\lim_{n\to\infty}P\left(\left|X_{n}\right|\geq\epsilon\right)=0$.

Then the fact that this works for every $\epsilon>0$ allows the conclusion that $X_{n}\stackrel{P}{\to}0$.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.