Let $X_1,X_2, \dots$ be iid random variables that take value 1 with probability 1/2 and value -1 with probability 1/2. Let $$S_n = X_1 + ... + X_n,$$ and let X be a Binomial(n, 1/2). Show that 2X − n has the same distribution as $S_n$ .

I am having a lot of trouble applying the central limit theorem to do this. The idea of $S_n$ being a distribution itself is confusing to me.

  • 1
    $\begingroup$ What makes you think the Central Limit Theorem is required? $\endgroup$ – mark999 Apr 26 '17 at 1:05
  • $\begingroup$ That is the chapter we are in right now in my statistics class $\endgroup$ – Ravmcgav Apr 26 '17 at 1:17

Given that $X_{i}=\pm 1$, with probabilities $P(X_{i}=-1)=P(X_{i}=1)=1/2$ and a random variable $S_{n}$ is defined by $S_{}=X_{1}+X_{2}+\cdots +X_{n}$. We are required to show that the random variables $2X-n$ and $S_{n}$ have the same distribution.

Define a random variable $Y_{i}=\dfrac{X_{i}+1}{2}$ which takes values either $0$ or $1$. Now, $$P(Y_{i}=0)=P\left( \dfrac{X_{i}+1}{2}=0\right)=P(X_{i}=-1)=1/2$$ $$P(Y_{i}=1)=P\left( \dfrac{X_{i}+1}{2}=1\right)=P(X_{i}=1)=1/2$$ As the random variables $Y_{i}, i=1,2,\cdots,n$ are iid $Bern(1/2)$, $\sum_{i=1}^{n}Y_{i}\sim Bin(n,1/2)$.

\begin{eqnarray*} \sum_{i=1}^{n}Y_{i}&=& \sum_{i=1}^{n}\left(\dfrac{X_{i}+1}{2}\right)\\ 2\sum_{i=1}^{n}Y_{i}&=& \sum_{i=1}^{n}\left(X_{i}+1\right)=\sum_{i=1}^{n}X_{n}+n\\ 2\sum_{i=1}^{n}Y_{i}&=& S_{n}+n\\ 2 X - n &=&S_{n} \end{eqnarray*}


Hint: $\frac{1}{2}(X_i + 1)$ is a $\operatorname{Bernoulli}(1/2)$ random variable.

Thus $\frac{1}{2}(X_1+1) + \cdots + \frac{1}{2}(X_n+1) = \frac{1}{2}(S_n+n)$ is a $\operatorname{Binomial}(n,1/2)$ random variable.

More detail:

  • Check that $\frac{1}{2}(X_i+1)$ is a $\text{Bernoulli}(1/2)$ random variable.
  • $\frac{1}{2}(X_1+1) + \cdots + \frac{1}{2}(X_n+1)$ is therefore a $\text{Binomial}(n,1/2)$ random variable (recall that a $\text{Binomial}(n,p)$ random variable is the sum of independent $\text{Bernoulli}(p)$ random variables)
  • Let $X$ denote the above sum; we showed $X \sim \text{Binomial}(n,1/2)$. Then using the definition of $S_n$, we can show $X = \frac{1}{2} (S_n + n)$.
  • So, $S_n = 2X - n$ where $X \sim \text{Binomial}(n,1/2)$..

The central limit theorem is not needed here.

  • $\begingroup$ Not sure if I totally follow. How does this imply that 2X - n has the same distribution as S_n? $\endgroup$ – Ravmcgav Apr 26 '17 at 1:28
  • $\begingroup$ @Ravmcgav See my edit $\endgroup$ – angryavian Apr 26 '17 at 2:07
  • $\begingroup$ I just still don't really see how the central limit theorem comes into play. Really sorry to keep this chain going I just want to make sure I understand! I see that $\frac{1}{2}(S_n + n)$ Could be manipulated into the expression I want if we set it equal to a new random variable, but I'm not sure how that would help us... $\endgroup$ – Ravmcgav Apr 26 '17 at 3:24

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.