19.01.2015 Views

Statistics for Decision- Making in Business - Maricopa Community ...

Statistics for Decision- Making in Business - Maricopa Community ...

Statistics for Decision- Making in Business - Maricopa Community ...

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

follow<strong>in</strong>g data was collected on the revenue generated by randomly sampled red-lights<br />

across the valley. The goal is to have, on average, each camera generate $750 and no less<br />

than $640 per day.<br />

883 522 590 779 887 615 690 771 843 509<br />

872 840 536 892 880 588 547 770 687 842<br />

832 840 676 555 884 617 517 586 505 552<br />

a. Can the state be 95% confident that the desired average is possible<br />

b. Generate a 99% confidence <strong>in</strong>terval <strong>for</strong> , the population average daily revenue<br />

per camera. Expla<strong>in</strong> <strong>in</strong> a complete sentence what this means.<br />

c. Is the CLT valid <strong>in</strong> this problem Expla<strong>in</strong>.<br />

d. Us<strong>in</strong>g the assumption that the distribution of ̅ is normally distributed, calculate a<br />

theoretical 95% confidence <strong>in</strong>terval <strong>for</strong> (you will need to estimate the<br />

√<br />

standard deviation of ̅‟s and ̅ to estimate .<br />

e. In reality, anytime we estimate parameters, like you did above <strong>in</strong> part d), we<br />

actually shouldn‟t assume a normal distribution. Instead, we should assume what<br />

is known as a -distribution, which is symmetrical, though has more variability to<br />

account <strong>for</strong> the uncerta<strong>in</strong>ty <strong>in</strong> our estimates.<br />

Watch this brief <strong>in</strong><strong>for</strong>mative video:<br />

http://www.youtube.com/watchv=yV-0ReCXW64<br />

Pull up the follow<strong>in</strong>g applet: http://www.stat.tamu.edu/~west/applets/tdemo.html.<br />

You can type <strong>in</strong> the percentile correspond<strong>in</strong>g to means you want to consider.<br />

stands <strong>for</strong> “degrees of freedom” and can be calculated by tak<strong>in</strong>g the sample size<br />

m<strong>in</strong>us 1 ( ). (From the video, we know that, if the sample size is really, really<br />

big, then the difference between the normal distribution and t-distribution<br />

becomes <strong>in</strong>dist<strong>in</strong>guishable.) The output of this applet will give you the number of<br />

standard deviations your endpo<strong>in</strong>ts will be on either side of the mean.<br />

For example, you will f<strong>in</strong>d that a 99% confidence <strong>in</strong>terval <strong>for</strong> a sample of size 100<br />

has endpo<strong>in</strong>ts that are 2.626 standard deviation from the mean (left and right).<br />

Let‟s say your sample mean is ̅ and standard deviation . Then, the<br />

confidence <strong>in</strong>terval will be an <strong>in</strong>terval around the sample mean. That is, one<br />

standard deviation is √ √<br />

(remember, the standard deviation of means<br />

requires that we divide the standard deviation among <strong>in</strong>dividual ‟s and divide by<br />

the square root of the sample size). So, 2.626 standard deviations would be<br />

2.626(0.5) = 1.313 units away from the mean. The endpo<strong>in</strong>ts would be 40 – 1.313<br />

and 40 + 1.313, or 38.687 to 41.313.<br />

Formulaically, we found:<br />

<strong>Statistics</strong> <strong>for</strong> <strong>Decision</strong>-<strong>Mak<strong>in</strong>g</strong> <strong>in</strong> Bus<strong>in</strong>ess © Milos Podmanik Page 201

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!