<p>YES… the best news I’ve heard in a while.</p>
<p>I’m going to put all the inferences formulas/degrees of freedom/when situation to use them/conditions in my calc. </p>
<p>Scrolling through will save more time, than me trying to remember by memory…</p>
<p>@charrizard: oh sorry, i was overthinking it, although I have no idea why they would make two categorical differences - possibly to throw off people like me, although I would still get the right answer</p>
<p>^ Haha, no problem! They should just make all the df calculations the same… -.-</p>
<p>And @themeaningoflife, that’s going to take a long time to put into your calculator…</p>
<p>anyone understand 26, 27, 30. 38, 39 in the 2007 exam? if so, can you please explain? 
</p>
<p>sry if it’s a lot to ask!</p>
<p>Can we use the formula sheet on the multiple choice portion?</p>
<p>Did someone else find the 1997 exam relatively easy? I missed 5, but two still don’t seem right.</p>
<p>For 16, I don’t see how A could be true. Why would doctors be different than other patient. They’re humans too… Unless you assume the stereotype that they are generally healthier… which isn’t always the case.</p>
<p>Also, how do you do 20.</p>
<p>I tried multiplying the mean by the sample size (30)(.47) = 14.1
Used the t test and go p-value = 0.</p>
<p>EDIT: Will 30/35 get me a 4 assuming I do decent on the FRQ and quite possibly miss all of 6. lolz</p>
<p>but for 19, </p>
<p>it’s asking for the weights of the two containers on the same flight so it’s asking for a scatter plot of two points.</p>
<p>The two containers can both be the same weight. If X is the limit then both containers can each weigh X/2. </p>
<p>I know I’m wrong but now I’ll be refuted and understand better by your counterarguments.</p>
<p>There’s a limit to the total weight. Assuming the two containers collectively weigh the maximum capacity allowed, as the weight of container X increases, the weight of container Y must decrease. Therefore, the correlation of the graph must be below zero.</p>
<p>I think that’s right.</p>
<p>My question: we can use the formula sheet on both parts of the exam?</p>
<p>
</p>
<p>The thing is, it doesn’t matter whether or not you know that doctors are generally healtheir. If your sample was constructed so that only a certain subset of the population could have been picked (in this case, only doctors could be picked), your results are valid only for that subset. That doesn’t depend on suspecting some stereotype one way or the other.</p>
<p>
</p>
<p>How did you go about using a t-test without null and alternative hypotheses?</p>
<p>Can you explain it to me?</p>
<p>I’m off to take my test, good luck to you all! (:</p>
<p>Can one of you guys explain the definition of a confidence interval to me?
As in not what a CI is/how to make one, but just the definition.</p>
<p>i.e. if you find out that the 95% CI for something is {5.2,8.1}, and you’re told to ‘interpret’ it, what is a nice definition-esque answer?</p>
<p>Thanks :)</p>
<p>If you’re told to interpret it, you can usually just say “We can be 95% confident that the true mean is between blah and blah”. I’m not quite sure what else you want; you want a definition, but not what it is?</p>
<p>95% of all intervals created using said procedure would contain the true mean/proportion</p>
<p>So it is not that we’re 95% confident the true mean is in this interval it is:</p>
<p>We are confident that in 95% of these intervals the true mean will be contained.</p>
<p>How do we say that elegantly and consistently?</p>
<p>Also, if power is 1-B (type 2 error), then the most powerful tests would be one where alpha is larger? So when alpha and n are larger it is more powerful, albeit more likely to make a type 1 error.</p>
<p>Nope, it’s a combination of the two. Both what amarkov and I said were essentially correct.</p>
<p>So what exactly is the optimum way to conclude a confidence interval of {5.2,8.1} with 95% confidence?</p>
<p>Thanks.</p>
<p>Hm, kay.
That’s what I was confused about. I’ll just use something along those lines then.</p>
<p>“We are 95% confident that the true mean of the population will be between 5.2 and 8.1.” will suffice.</p>
<p>Oh, I get it now. Because in 95% of the intervals the true mean is between what you find, you can say that in this particular interval the probability that it is within there is 95%.</p>
<p>It helped looking at that graph with the different intervals that crossed over the true mean while 5% didn’t. (On wikipedia)</p>