*** official ap statistics 2013 thread ***

<p>Independence/Association. Although the end-value would have no difference, there was a difference in the procedure. They randomly selected individuals to make up a heterogeneous group first, then separated them categorically. In a test of homogeneity, they would randomly select from several homogenous groups, and compare these homogenous groups.</p>

<p>um test of independence is 2 variables… test of homogeneity is a subset of goodness of fit which is one variable lol… and they kinda asked you in the question is there an association between the variables (plural)</p>

<p>Are we allowed to talk about the FRQs until they come out on the college board website? we don’t want our scores cancelled guys?</p>

<p>can someone give the conditions that must be checked for a chi-squared test for independence? Thanks</p>

<p>It must be a random sample and at least 80% of the expected values must be greater than or equal to 5.</p>

<p>Ah, I think I did the checks for independence, but then said we could carry out a test of homogeneity. But in my conclusion, I said it proved an association at the .05 level, but not the .01 level. Ah, I’ll probably get a P on that part. Oh well.</p>

<p>Okay I think I found the problem: was it a proportion or a means question? Proportions use the infinity value of 2.576, means would use 2.626 and a df of 100. I think I might have found the error…</p>

<p>Even though it’s good to err on the side of caution, I highly doubt that College Board will bother showing up at our houses to accuse us of cheating.</p>

<p>MC was somewhat difficult, FRQ was a joke.</p>

<p>But that’s only because my teacher couldn’t explain the difference between a z-test and a t-test. Eventually, I had to seek help elsewhere–a teacher at a neighboring school, who was an AP, grader a few years ago, gave me a list of mnemonics to remember. Boy, did that save me on some of those objective problems.</p>

<p>@APNeuroSurgery: conditions for chi squared tests are SRS of your thingy and none of expected values are less than 1 and no more than 20% are less than 5
@TheBombingRange: Regardless of proportions or means you will use z for calculating the sample size questions (again you can’t even calculate your df without knowing the sample size in the first place)</p>

<p>But all of the possible sample sizes for that question had a df of 100. And you do not use df for proportions, only means. It was a 99% confidence interval and it was the sample Standard Deviation, not population. Why do you think it was a Z?</p>

<p>I can honestly say that I know hate typhoons</p>

<p>I wasn’t sure how to describe the distribution of the time plot on question 6. I basically talked about how the the pacific one fluctuated more and had a bigger decrease between 1997-2010. Does that sound anything like what I was supposed to say?</p>

<p>One kid in my class plugged in all of the values and described both distributions by center, shape, and spread. I didn’t think that’s what they wanted us to do, it took him 20 minutes just for part A.</p>

<p>for part a i found the min, q1 med, q3 and max, for each, and for part b i said where it decreased and such</p>

<p>for a) i think you did have to do center shape and spread discussion
for b) i think you had to talk about parts in the graphs where trends started to occur</p>

<p>I thought that you were supposed to describe the strength, form, direction. And it’s impossible to talk about shape on a line graph because the x variable is qualitative.</p>

<p>They just released the free response. <a href=“http://media.collegeboard.com/digitalServices/pdf/ap/apcentral/ap13_frq_statistics.pdf[/url]”>http://media.collegeboard.com/digitalServices/pdf/ap/apcentral/ap13_frq_statistics.pdf&lt;/a&gt;
Does anyone want to try to go through each question?</p>

<p>^lets go for it</p>

<p>did we have to show our conditions for 1b? i forgot to</p>