AP stats discussion

Test Preparation AP Test Preparation
481

<p>2b) Using bionompdf/the formula gave the answer right? Darn it, realized I read the question slightly wrong… Still partial I guess.
2c) At first I got like 22% but then I took it to the fifth power, which I think is wrong.</p>

<p>6 last part) I just used the formula for calculating outliers and said that if there are any outliers, (they number is greater than the max or less than the min.) then the mean will be diff. from the median suggesting a somewhat skewed distribution. Is that good enough??</p>

<p>I think I did worse than I thought I did…</p>

482

<p>Shizzle, it said “define a formula for a test statistic.” A formula.</p>

<p>I made an even more foolish mistake though.</p>

<p>Y = (Q1 - min) - (median-Q1) + (Q3 - median) - (max - q3)</p>

<p>My justification in part d was “if y = 0, then the distribution is normal.”</p>

<p>I only had 7 minutes to do the problem though; still, I sucked so much on 6.</p>

483

<p>Anyone else put (max-median) / (median-min)?</p>

484

<p>laplast, that’s just what I did! It was the only thing I could think of, and makes sense to me.</p>

485

<p>haha same thing here:
(max-med)/(med-min)</p>

<p>its perfect ;)</p>

486

<p>Innocuity, that’s what I did also. What’s wrong with that?</p>

487

<p>You can’t use outliers, because there might be one point that was WAY unrepresentative of the rest of the data… like the data set (1, 1, 2, 3, 3, 4, 4, 4, 5, 5, 1000). Using max or min could way skew your formula… You have to use, like Q3 and Q1… I think. I’m 95% confident about that. (lol)</p>

<p>That graph on the first one KILLED me. I was like “What. The. Heck.” when I saw it… eheh, I felt like such an idiot…</p>

488

<p>Yeah the first thing I wrote down was (Q3-med)/(med - Q1). But then I replaced the Q1 and Q3 with min and max. I mean, I’m sure I’ll get a partial for my answer.</p>

489

<p>“You can’t use outliers, because there might be one point that was WAY unrepresentative of the rest of the data… like the data set (1, 1, 2, 3, 3, 4, 4, 4, 5, 5, 1000). Using max or min could way skew your formula… You have to use, like Q3 and Q1… I think. I’m 95% confident about that. (lol)”</p>

<p>Ehm, that data set is pretty much the definition of a skewed distribution. I don’t see your point? Q3 and Q1 doesn’t work as often as max and min in predicting skewness, for ex.</p>

<p>(2, 3, 5, 5, 6, 10), (14, 15, 18, 19, 20, 21, 22, 22, 22, 36)</p>

490

<p>was it ok for the statistic if i did:</p>

<p>(max-median)-(median-min)</p>

<p>and if that is > 0, it is skewed right?</p>

491

<p>that sounds right to me</p>

492

<p>i agree with hipster…max and min are more appropriate than Q1 and Q3
outliers are one indication of skew…</p>

<p>but honestly…i don’t htink they’ll be picky and count off…(i could be wrong) but that question seems to lend itself to many many possible responses</p>

493

<p>Although the answer key is usually strict about answers, I feel like if they asked us to create our own formulas, they knew there were multiple right answers.</p>

494

<p>I did put a formula, it was just the outliers formula. It’s still a formula… but I’m pretty sure it doesn’t qualify as a statistic. </p>

<p>Oh well, I heard of people entirely skipping question 6 and getting a 5.</p>

495

<p>I just did a bar graph for the first one… isn’t that ok? I made a key and shaded in the boys one and left the girls one blank.</p>

496

<p>for 3b, I put it counfounding variable and not lurking variable? Which one is it?</p>

<p>for 6a, it asked for the parameter in question and the null hypothesis and alternative hypothesis</p>

<p>What is the parameter in question? I didnt understand that…</p>

497

<p>Hey guys will I get a partially correct if I put in Chi square test for homogeneity instead of independence?</p>

<p>And will i still get some credit if I used binompdf for question number 2?</p>

498

<p>ya u’ll still get credit for both…im guessin 3/4</p>

499

<p>parameter in question i believe was the population mean of fuel efficieny for all cars of that model</p>

500

<p>^crap i think i missed that…i mentioned something with population mean but not in that way…i mentioned like some mean test or something</p>

<p>If you just missed 1 condition on a problem, is that an automatic 3?? and counfindg = lurking? for that 3b, i put some people might be intimediated to do the disection or something to that extent…</p>