4
$\begingroup$

This is from an assignment.

The problem:

Given $n=15$ and assuming $Z_1, ..., Z_n$ are iid $N(\mu, \sigma^2)$, let $\bar{Z}$ be the sample mean and $S^2 = \frac{1}{n-1} \sum^n_{i=1} (Z_i - \bar{Z})^2$, define the following random variable $$ V = \frac{(Z_1 - \mu)^2}{\sigma^2} $$ and calculate $$ P(V > 1.801) $$

My solution:

The second part is not an issue, I am a little stuck on what distribution $V$ follows. I know that $(Z - \mu)$ subtracts the mean from the normal distribution and gives $\mu = 0$, and I know that $Z^2$ gives a $\chi^2$ distribution. I think that dividing by $\sigma^2$ probably standardizes the variance to produce a standard normal $\chi^2$ if that exists. To my mind though, that would look like $$ \left( \frac{Z_1 - \mu}{\sigma^2} \right)^2 $$ to first create the standard normal then square to produce a $\chi^2$ distribution.

Any prods in the right direction would be appreciated.

Edit: Some of the above information may be superfluous and for different parts of the question which I have not asked about.

|cite|improve this question
$\endgroup$
  • 1
    $\begingroup$ $X = (Z-\mu)/\sigma$ has a standard normal distribution and $V = X^2$. What is the distribution of the square of a standard normal? $\endgroup$ – linksys Feb 25 '16 at 17:37
  • $\begingroup$ Dividing by the standard deviation causes the random variable to have unit variance. This is because constants are squared when they factor out of the variance operator. $V$ is just a standard normal squared. $\endgroup$ – dsaxton Feb 25 '16 at 17:38
  • $\begingroup$ @linksys Is it a $\chi^2$ with 1 df? $\endgroup$ – Chris C Feb 25 '16 at 17:41
  • $\begingroup$ @dsaxton Thanks! I didn't know that, and it makes sense as to why $\frac{\bar{Z} - \mu}{ \sigma / \sqrt{n}} \sim N(0, 1)$. A standard normal squared would be $\chi^2$ with 1 df? $\endgroup$ – Chris C Feb 25 '16 at 17:42
  • 1
    $\begingroup$ Either there are a very large number of typos in this question or else most of it consists of intentional distractors. Since the definition of $V$ involves only $Z_1$, $\mu$, and $\sigma$, then--once you have double-checked that you copied it correctly!--you ought to completely ignore the information about $n$, $Z_2, \ldots, Z_n$, $S^2$, and $\bar Z$. $\endgroup$ – whuber♦ Feb 25 '16 at 17:43
3
$\begingroup$

Dividing by the standard deviation causes a random variable to have unit variance. This is because constants are squared when they factor out of the variance operator. $V$ is just a standard normal squared which follows a $\chi^2_1$ distribution.

|cite|improve this answer
$\endgroup$

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.