| | | 
05-10-2011, 06:41 PM
|
#46 | | Senior Member
Join Date: May 2009 Location: South Florida
Posts: 1,412
|
margin of error = z*sqrt(S^2/n)
margin^2 = z^2*S^2/n
n = z^2*S^2/margin^2
n = 1.96^2 * 7200^2 / 200^2
n = 4978.7, rounds up to 4979 as the minimum sample size, so we take the next-highest answer choice, which is C.
| 
| Reply
| |
05-10-2011, 06:45 PM
|
#47 | | Junior Member
Join Date: Nov 2010
Posts: 77
|
Hey keasby where is the answer key to that test?
| |
| Reply
| |
05-10-2011, 06:45 PM
|
#48 | | Junior Member
Join Date: Dec 2010
Posts: 200
|
Here's a cram sheet: Statistics Study Sheet
Here are calculator functions: http://www.pagesf.com/gal/stats/prep...ator_sheet.pdf
Here are assumptions: http://apcentral.collegeboard.com/ap...tion_31840.pdf
And I also found this on the internet (I didn't come up with this, some teacher named Mr. Dooley did; I just copied and pasted it):
Here is a good way to remember the assumptions/conditions for inference for the slope of the true LSRL.
L – The mean response μy has a straight Line relationship with x. μy=α+βx The slope β and intercept α are unknown parameters.
I – Repeated responses y are Independent of each other.
N – residuals have approximately Normal distribution
E – residuals have Equal variance
| |
| Reply
| |
05-10-2011, 07:32 PM
|
#49 | | Junior Member
Join Date: Dec 2010
Posts: 116
|
Can someone explain numbers 3,4,17,27,38 on the 2007 Test? Thanks Powered by Google Docs | |
| Reply
| |
05-10-2011, 07:56 PM
|
#50 | | Junior Member
Join Date: Aug 2010
Posts: 299
|
#3. The 67th percentile means 67 percent of all values are below it, right? So essentially, we want to find the critical value that corresponds to a p-value of .67, which would be like invNorm(.67) on your calculator, which is .44, this means that for a value to be in the 67th percentile in a normal model, it must be .44*(standard deviation) greater than the mean. which is .3*.44=.132 greater than the mean.
#4 To start off, find the critical value (z*) that corresponds to a confidence level of .9, which is 1.644, so we can cross off B and D. C and E dont look like any test that I know of, so cross that out. Leaving only A. If you want to actually do it, the test is a 2 proportion z interval, so you would look on your formula sheet for the standard deviation that corresponds to that CI.
#7 This is a X2 GOF test I think. But dont worry about that part, you have the X2 value and the degrees of freedom (4-1 or 3), so plug those into your calculator and you get a p value of .045. The null hypothesis for X2 test would be that the observed results are the same as the predicted ones, because we have a p value thats less than our alpha level, we can reject the Ho. So B....but the answer key says A...what did I do wrong?
# 38 Common sense would lead you to D or E as the answer. Simply computing the probability for getting a 0 total gives E as the answer.
Last edited by AstroBlue; 05-10-2011 at 08:07 PM.
| |
| Reply
| |
05-10-2011, 08:06 PM
|
#51 | | Junior Member
Join Date: Apr 2011 Location: South Dakota --> Michigan '16
Posts: 91
|
3: invnorm(.67) will give you the z-score or the number of standard deviations above/below the mean depending on the sign.
(.4399)(.3)=.132
(C)
4: Interval = difference of proportions +- (critical value)(standard error/deviation)
Critical value = invNorm(.95)=1.645 eliminating (B) and (D).
Standard error/deviation follows the formula sheet
(A)
17: I just did a quick GOF test on my calculator and got a p-value of .045, so it's significant, which means A or B. Chi-square tests are always one-sided (greater than) so it is (A).
I'll post the others, if they haven't been posted already, after I finish the test.
| |
| Reply
| |
05-10-2011, 08:12 PM
|
#52 | | Junior Member
Join Date: Apr 2011 Location: South Dakota --> Michigan '16
Posts: 91
|
@Astroblue
On 17, the answers are referencing the statistic calculated not the p-value.
| |
| Reply
| |
05-10-2011, 08:14 PM
|
#53 | | Junior Member
Join Date: Aug 2010
Posts: 299
|
oh, okay, thanks for clearing that up  | |
| Reply
| |
05-10-2011, 08:26 PM
|
#54 | | Junior Member
Join Date: Apr 2011 Location: South Dakota --> Michigan '16
Posts: 91
|
Here's further explanation on 38:
For the first game you can get 0, 1, or 2. The same thing for the second. Adding all the possibilities gives 0 (0+0), 1 (0+1), 2 (1+1, 0+2), 3 (1+2), or 4 (2+2). This eliminates everything except for (D) and (E).
P(0) = (.3)(.3) = .09
P(1) = 2(.3)(.4) = .24
P(2) = (.4)(.4) + 2(.3)(.3) = .34
P(3) = 2(.4)(.3) = .24
P(4) = (.3)(.3) = .09
Therefore (E).
| |
| Reply
| |
05-10-2011, 08:30 PM
|
#55 | | Junior Member
Join Date: Apr 2011 Location: South Dakota --> Michigan '16
Posts: 91
|
27: This one's just confusing. I just did 1 - .056 and got .944. The widest confidence interval therefore that's narrower than 94.4% would be 93% or (B).
I don't know if this is the right way to go about this problem though.
| |
| Reply
| |
05-10-2011, 09:07 PM
|
#56 | | Junior Member
Join Date: Apr 2011
Posts: 40
|
Hey, can someone explain 6e of the 2010 stats test? (not form b)
Thanks!
| |
| Reply
| |
05-10-2011, 09:17 PM
|
#58 | | Junior Member
Join Date: Apr 2011
Posts: 40
|
Okay, thanks. I read the answer wrong and got confused ><
| |
| Reply
| |
05-10-2011, 09:55 PM
|
#59 | | Junior Member
Join Date: Apr 2010
Posts: 123
|
do they give us something besides the formula sheet? like a random number table, t distribution values, and standard normal probabilites?
| |
| Reply
| |
05-10-2011, 10:02 PM
|
#60 | | Junior Member
Join Date: Jul 2010
Posts: 92
|
They "just" give us the 7-page formula packet, and if there's a simulation-type problem I think they'll give us a list of random numbers to use if they don't want us to use randInt.
| |
| Reply
| All times are GMT -4. The time now is 07:40 AM. |
