<p>What did you say for 2c?</p>
<p>I wonder how many points if #6 likely to be out of?</p>
<p>If you need any more input, I used matched pairs for the beef, too. For 1a, I put "average distance from the mean" (which might not be right) and "measure of spread" then described what would happen if it was larger or smaller.</p>
<p>the meat question was definitly matched pairs. #6 will be out of 4 points, just like the rest of the FRQ. however, it is weighted more than the rest of the frq. it counts for 25% of the free response score. what was the question for 2c?</p>
<p>there is some discussion among statistics teachers on 3b.
Is the standard deviation s/sqrt(n) or just the given s?
people are suspicious because it was too easy to be just the given standard deviation. Any other ideas?</p>
<p>hey if a person</p>
<p>omitted 6 c and 6 e (but got the other parts of 6 right)</p>
<p>omitted 3c (but got the other parts right)</p>
<p>and correctly did a 2 sample t test on 4</p>
<p>is he still in running for a 5</p>
<ol>
<li>was not a 2-sample-t-test.</li>
</ol>
<p>I wasn't sure about 3b either.
I just used the given s because it said that it was for the sampling distribution for n=50, and we were testing a sample of 50.</p>
<p>adroittactician, there is confusion regarding #4, but personally i can't see how it isn't a matched pairs. If they're testing methods of measurement, it makes no sense to test different samples, seeing as the methods don't alter the physical properties of the samples</p>
<p>How can #6 be out of 4 points? </p>
<p>Is it one point for a) and b). One point for c) and one point for d) and one for e)?</p>
<p>So what get a) right and b) wrong.</p>
<p>I don't know if this was asked before, but what were the conditions needed to run a Z test? and were they met...</p>
<p>I put Random( yes), n>30 (yes) and that's all i put, there are a few more conditions i think</p>
<p>I understand about #4...however, will that person still have a chance to get a 5 if he committed all those errors?</p>
<p>usually they say that if you get all parts essentially correct, you'll get 4/4, if you get 4 parts essentially correct or 3 parts essentially correct with the other parts partially correct, you'll get 3/4, if you get 3 parts essentially correct with the others not partially correct you'll get 2/4, otherwise you'll get 1/4. it's different for every question, though. that's just a guideline</p>
<p>what about 4 parts essentially correct and 1 part partially correct? Is that a 4? I would think so depending on communication of the response.</p>
<p>probably depends on how thorough you were... i dont really know</p>
<p>The assumptions for a two-sample proportion z-test is SRS, whcih it was, and np > 5, and n(1-p) > 5 for both samples. The experiment fails on of the np assumptions, so it cannot be used.</p>
<p>For 3b, I actually had 50*rt(3) since I knew Stdev(x-bar) = Stdev/rt(n) and we were given that Stdev(x-bar) = .3 and n = 50.</p>
<p>It seemed perfect while I was taking the test, but no one seems to have done this, which leads me to think that it's probably wrong.</p>
<p>I think you jsut use .3. I had a z-score of -1.67.</p>
<p>did 3b have anything to do with z-score?</p>
<p>I just did ncdf(-E99, 7.5, 8, 0.3)... what were you supposed to do?</p>
<p>as far as i remember, that is all you have to do. you dont do 0.3 / sqrt(50) because it said in the question that the standard deviation for 50 people was .3</p>