1
$\begingroup$

A random sample of $n=484$ voters in a community produced $x=257$ voters in favor of candidate A. Compute the sample variance.

I have the solution but I don't follow some of the steps in it.

$$ \begin{align} s^2&=\frac{1}{483} \cdot \sum_{i=1}^{484}(X_i-\bar X)^2\tag1\\ &=\frac{1}{483} \left[ 257\left(1-\frac{257}{484}\right)^2 + (484-257)\left(0-\frac{257}{484}\right)^2\right]\tag2\\ &=\frac{1}{483}\cdot48\cdot\frac{257}{484}\cdot\left(1-\frac{257}{484}\right)\tag3 \end{align} $$

I have no idea how to get from equation 1 to equation 2.

Going from equation 1 I get...

$$ \begin{align} \frac{1}{483} \left(\sum_{i=1}^{484}X_i^2-484\left(\frac{257}{484}\right)^2\right)\tag4\\ \frac{1}{483} \left(\sum_{i=1}^{484}X_i^2-\frac{257^2}{484}\right)\tag5 \end{align} $$

I don't know what to do with $\sum_{i=1}^{484}X_i^2$

I had an idea that the problem should use the Binomial variance $np(1-p)$ but nothing in equation 2 looks like that to me so I'm just entirely lost.

Edit: I noticed that I erroneously excluded parenthesis around the factor in my eq 4 and 5 which I've since added in which, of course, made all the difference.

$\endgroup$
1
$\begingroup$

Here, you can view the samples as a string of ones and zeros (where one means they voted for candidate $A$), and the sample variance is the variance of that. Everything you have done seems to have been consistent with that. But now it should be clear that $X_i^2 = X_i,$ so you can interpret $$ \sum_{i=1}^{484} X_i^2 = 257.$$ You will see that the answer is somewhat reminiscent of the binomial variance, like you expected.

$\endgroup$
1
$\begingroup$

For your attempt remember $$(X_i - \bar{X})^2 \ne X_i^2 - \bar{X} ^2$$

We go from equation 1 to equation 2 by noting there are $257$ such $X_i$ that equal 1 and $484-257$ such $X_i$ that equal 0.

$\endgroup$
0

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.