*** official ap statistics 2013 thread ***

<p>Hey guys, when would you use a one sample z test or a one sample t test? What’s the difference? Also, when would you use a matched pairs test? Thanks.</p>

<p>@YellowDog99
You might want to not take the exam tomorrow if you don’t know the difference by now lol. Anyway, a z test is for proportions and a t test is for means. A matched pairs test is when you have two results from same people (e.g, someone hitting a ball lefty then righty) then comparing to see if there is difference.</p>

<p>Yes. Also, a Z test CAN be used for means too IF you know the population standard deviation (which is fairly rare)</p>

<p>Thanks. I know how to run all of the tests, I got a 4 on the mock exam. It’s just that I get confused as to when to use each one, lol.</p>

<p>@ ohhaiitslily, yeah that was my question. Our teacher taught us two types of one sample tests, and now I remember learning that we would never really use the z one.</p>

<p>Also, how are you guys memorizing all of the assumptions for each test? Like N1P1 is less than 10, n1(1-p1) less than 10, and so on, and all of the standard deviation and test statistic formulas?</p>

<p>Did any of your teachers make like a consolidated list of the inference procedures with their descriptions or something, i seem to get mixed up a lot lol</p>

<p>One sample Z test is to find the significance of a set of proportional values based on the expected proportion (P0). One sample z tests are used for “yes/no” data or that uses percentages. To find the P-value, you go to normalcdf on your TI-84 (2nd -> vars -> normcdf) and input normalcdf(min [usually is -1000000 if z-score is positive], z-score). If your p-value is greater than the alpha level (usually 0.05) then your data is significant.</p>

<p>One sample T is used to test the value of the population mean or matched pairs difference based on the expected values of u0. I assume that you can figure out how to find the p-value using tcdf</p>

<p>im super nervious, I feel like every time i do a problem i’m really confident going in, and then I forget to check conditions or something and loss points</p>

<p>Are we allowed to store information in our calculators? I know it was allowed for AP Chemistry I assume the same applies for statistics.</p>

<p>Haha um, no. A Z test is when you have perfect data, a T test is when you use sample data. For means values you can use Z or T, proportions must use Z. Don’t be so quick to tell other people they shouldn’t take the test.</p>

<p>btw because i saw somewhere here in here mentioned that a Z test is for proportions and T tests are for means… this is not true: What defines a Z test or T test for means is whether or not you know the population standard deviation (if you do know= Z, if you dont = T). Another thing to not get tripped up on (speaking from experience :P): frequently the AP will ask you by giving you the mean/std deviation of a random variable X and the same info for another variable Y and often ask for the mean and std deviation of the random variable X-Y or X+Y, if it does not specify that they are not independent, then you cannot calculate the std deviation</p>

<p>@yellowdog99: for your question about memorizing the assumptions for the inference tests and the std deviations of the different tests… the one that is hard to remember is the normality assumption; just think of trying to satisfy the normality approximation for the binomial distribution(btw those assumptions were (n1p1>10 not < i assume typo but just in case) . And about memorizing the different std deviations: i would recommend you memorize them to better understand the different tests, but if you really dont think you can memorize them all, on the formula sheet they give you at the AP exam the different std deviations for 1,2 sample T/Z test and the 1,2 Proportion Z test are provided</p>

<p>@DedicatedForLife: I was wondering that, too. Do you think the proctors will check our calculators?</p>

<p>We do not have to clear our calculators. At least that’s what my teacher said today.</p>

<p>lbouchard? more like l******ard
…I doubt proctors will check calculators but try to get it in your head just in case they do. Remember, stats is one of the lowest percentage requirements for a 5. Good luck everyone!</p>

<p>My AP Statistics class is right before the exam and my teacher will be checking all our calculators beforehand. It’s not too bad memorizing the conditions, but knowing which calculator function to use gets confusing sometimes; no not the formulas from the formula sheet. Quick note, don’t forget about your calculator batteries! If you haven’t change them in six months and use them constantly, then you should probably change them.</p>

<p>Wow, I make a joke trying to lighten the mood the day before the AP exam and I get lambasted. I know sarcasm doesn’t come across real well on the Internet and even worse on CC but Jesus.</p>

<p>In reference to the z vs t debate above: you only use a z test for a mean when you are given the population standard deviation, which is very rare, correct? And if that is correct, for all intents and purposes, getting a proportion means z and getting a mean means t, right? That’s basically how my teacher described, although people in here are making me a bit worried hahah.</p>

<p>Does anyone have a list of formulas not given on the reference sheet?</p>