<p>It’s called the BoxPierce test. Why do you want to know?</p>
<p>Does anyone have the solutions to the 2010 Statistics FRQs?</p>
<p>bump 10char</p>
<p>Does anyone know if a scatterplot would have worked for #1 on the FRQs?</p>
<p>depends on the graders. residual plot would have definately been the best choice, but i could see a scatterplot getting full credit if you explained well why the scatterplot shows whether the linear model fits…dunno</p>
<p>I made a scatterplot and explained that it had a strong, negative, linear relationship, so the linear regression model was appropriate.</p>
<p>I think i made a scatterplot as well</p>
<p>A scatterplot really seemed to be the only graphical display I could come up with. I can see why some people are saying a residual plot would be better, but I couldn’t figure out how to make one with the data provided in the problem. I hope my approach with the scatterplot is an acceptable answer; it makes sense to me. If not, I hope it at least merits partial credit. I REALLY need a 4 on this exam…</p>
<p>Residual is best because it illustrates how well the linear transformation fits.</p>
<p>yea i think a lot of ppl didnt do residual because they didnt know how to make one. u just gotta do lin reg… go into l3 or somethin and go to list resid and then graph it against the x axis. from what ive seen tho almost everyone did a scatterplot</p>
<p>Yeah, I just couldn’t come up with the residual plot. It makes me feel better to hear that a bunch of other people used scatterplots.</p>
<p>How about some of the other FRQs? Does anyone have any answers they could share to give us an idea of how we did?</p>
<p>Sorry to keep bumping this thread up, but I’ve just been going through the FRQs a lot and it would be nice if someone could give some feedback and maybe confirm some of my answers. I have what I could recall from my answers posted here; I apologize as some are pretty vague, but hopefully they show I was on the right track. Here are some of the answers I got, from what I can recall:</p>
<p>1ai. The 5 concentrations of garlic.
ii. European starlings.
iii. Mean number of food granules consumed.</p>
<p>bi. I created a scatterplot.
ii. Linear regression model is appropriate because scatterplot shows a strong, negative, linear relationship.</p>
<p>2a. Approximately normal because of the Central Limit Theorem, (sample size is greater than 30).
b. I ran out of time on this one, but I basically attempted to set up some z-scores of some kind, but was unable to obtain an answer.</p>
<p>3a. In repeated samplings, 95% of the intervals would contain the true proportion of households in the county that own at least one dog.
b. No, there is not evidence because the confidence interval contains .39, so it is not an unusual value
c. 66 households. I showed the work on the exam, using the appropriate formula not given on the formula sheet, but which was in my review book and has just slipped my mind, haha</p>
<p>4a. I used a couple of formulas off of the formulas page, mu subscript x = np, and sigma subscript x = sqrt(np(1-p)) to get the answers.
b. I believe I used a z-score of some kind on this one.
c. Something along the lines of using a random number table for each of the models.</p>
<p>5a. I used a 2-Sample Difference of Means T-Test, stating the name of the test and providing the formula. I checked the conditions, stating that the problem mentioned that the samples were randomly selected, so SRS checked out. I checked for a normal distribution with two boxplots, one for each independent sample.</p>
<p>The parameter of interest was the mean length of fish.</p>
<p>Ho: mu1=mu2
Ha: mu1 not equal to mu2</p>
<p>Alpha was equal to .05.</p>
<p>I then carried out the test, and I believe I ended up failing to reject the null hypothesis, since the p-value was much higher than .05</p>
<p>6a. I used a triple bar graph with a key, one bar for each region and the bars were grouped into distances from the coast.
b. Further from the coast, less damage that is sustained. Then I made note of another couple of differences.
c.
10-20 miles: Average Rank:
Gulf Coast 1 1.8
Florida 2 1.4
North Atlantic 3 2.8
d. Q - value is calculated as follows:</p>
<p>Q = 5((1.8-2)^2+(1.4-2)^2+(2.8-2)^2) = 5.2</p>
<p>e. No clue. Just took a stab at it, but I don’t even remember. </p>
<p>Thank you for those of you who take the time to look at this. Please let me know if these are close to what other people got.</p>
<p>Here is the formula from a previous post by peter2328 which I also used for 3c but couldn’t recall:</p>
<p>"Here is 3 c:</p>
<p>use the margin of error formula given the 95% confidence interval, 0.417 +/- 0.119:
ME = c.v. * se(est)
ME = z<em>(95%) * sq.root (P’Q’/n)
0.119=1.96</em>sq.root( (0.417<em>(1-0.417))/(n))
0.119^2=1.96^2</em>(0.417*(1-0.417))/n
n= ((1.96^2)(0.417)(1-0.417)) / (0.119^2)
n= 65.95 ~66 households."</p>
<p>2b was zscore >.5 was about 28% chance</p>
<p>4b was z score shoulda been about 17% chance
4c was stratified random sample</p>
<p>5a i think alternate was U1<U2
6d i think the q value was 6.4
6e was reject the HO b/c if u used the table or the simulation, the chance of getting a q value of 6.4 or greater was .039, which is basically a low p value.</p>
<p>yea… went over these in class after the test i thought i was VERY prepared for the test, but missed a lot of pts on the fr 1-4, overall i think the test was harder than usual</p>
<p>Ok, I think I remember using stratified random as part of my answer for #2c. Oh yeah, now I remember I put that the alternate hypothesis was mu1 < mu2 on #5. Thanks. And Q-value of 6.4 for #6d? Could you please show me how you got that answer?</p>
<p>bump, anyone know how to get a Q-value of 6.4?</p>
<p>it depenends on how you ranked the 3 regions. i remember i got 5.2 to start, and then after double checking and switching the rankings it was 6.4. but… if u get a wrong q value and get the rest wrong, but correct assuming ur q value was correct, they arent going to penalize you twice</p>
<p>Oh, I think I see what I did now. Hopefully I can get some partial credit on that one.</p>
<p>This is the one exam that has me worried. Does anyone else want to look over post #395 and confirm some of my answers? It would be much appreciated.</p>