*** official ap statistics 2013 thread ***

<p>ok there was def no vegetable question on my test… was this MC or FRQ? I’m pretty sure we had diff questions now. Also, question 6 was the headache question right? Question 5 was the surgery one… </p>

<p>question 6 c asked about power and p values and gave you a histogram… how did you solve that one</p>

<p>and did you ppl have a lot of Ds on the MCs… i know i did</p>

<p>I think you got the form B or whatever…
My question 6 had to do with typhoons.</p>

<p>im same as ibworld student lol and the typhoon one was mindblowingly easy compared to past #6’s</p>

<p>Sound like we’ve had different forms:/ Our #6 was so cruel and literally no one figured it out.</p>

<p>Okay I’m sorry I was wrong. It was a 99 percent CI, which is a CV of 2.626. I still stand behind 413. How do you get 350 with a definite CV of 2.626 times (36/root n)???</p>

<p>It literally was interpreting a bunch of line graphs and plotting a point. No inference statistics or probability at all.</p>

<p>@TheBombingRange idk what CV = 2.626 means
m = t (s/ root n)… rearrange the equation you get n = (t x s / m)^2,
plug in the values and ta da!</p>

<p>@magmar95 yes lots of Ds! I checked and rechecked and finally managed to change one of them into a C ._. Your FRQ were the same as mine. No one on my class knows how to do #6, so if you didn’t get it either, just relax.</p>

<p>Such an easy exam–especially the free response. What a nice surprise.</p>

<p>Wow I just realized that when you standardize the values of a linear regression line, the mean is reduced to zero and standard deviation would be 1, therefore the line starts at zero and we have no way of knowing if it still passes through the original mean values. I’m pretty sure it wasn’t d or e because those were the mean values but just interchanged. c was that the slope was 1, but the slope would be equal to the correlation which can be any value since b=(correlation)(Standard deviation of y/ Standard deviation of x). I didn’t remember a, but it seemed unlikely so I crossed it off. B seemed the most viable option which was that the correlation didn’t change. This is actually pretty strange since the correlation is already calculated using standardized values haha.</p>

<p>This test was much easier than expected. Hoping to get a 5! :)</p>

<p>I’m almost positive I got #6 right… But you guys are right no one in my class knew either</p>

<p>@Arion. Choice A was that the r coefficient is the same as the original .</p>

<p>Hi guys, so I’m confused here: for FRQ #6, people have been saying that there was some sort of histogram or some sort of Point that you needed to plot, and I never read that direction!</p>

<p>My question #6 was the typhoon question, did I have to plot a point on one of those graphs? Like part (d) or something?</p>

<p>I filled in the square on part (c) with the number I calculated, did I have to plot that somewhere?</p>

<p>Also, was part (d) even a question, or just a graph?</p>

<p>The question in d) was to complete the rest of the graph. Which was to just plot and connect the 2010 4yr average and you get credit. Dead easy. Seemed like people skipped it though.</p>

<p>Ah, I missed it. :frowning: And I tried to figure out why part d had no apparent question… I guess the question completely passed my vision.</p>

<p>One other thing, one of the free response Q’s had something about normal approximation and how to prove that we couldn’t use it. </p>

<p>I said np>10 nq>10 was not satisfied (b/c one of the 4 values was = 0). I’m not sure if this is correct though.</p>

<p>@Magmar95, you and I got Form I, whatever that is. Q 6 was a headache question, the power of the test, according to one of my classmates, β = .55 because 55/100 simulated p-values were below .10 (alpha). that means that there is a 1-β = .55 chance of getting a type-II error</p>

<p>Omgg I can’t believe that was the AP test. Felt REALLY doable. Easier than 2007 MC as well… And the fact that I was feeling pretty good going in def helped. Oh man for the FRQ’s I wrote PARAGRAPHS for each question to make sure I wrote every single bit of small nit-picky flea detail down. AP graders please be lenient with me … !!!</p>

<p>Kimaroo I put because n<30 failed.</p>

<p>The central limit theorem (n<30) only works for sample means. For proportions you’re supposed to use np. So you couldn’t use a normal approximation because np was 0, or less than 5.</p>