<p>^ df = 3 </p>
<p>10char</p>
<p>How'd you find that link?</p>
<p>for 1c can u write that the mean of 3/4 cup is greater than the 1-cup?</p>
<p>shema, I also said they preferred both habitats 2 and 3. It's a perfectly fine answer as long as you can show that your evidence supports it. Plus, they asked for habitat typeS.</p>
<p>i found it through college board, they released the test, but do u know how to do the problem? (number 4 part c</p>
<p>conceratejungle how did u support your answer?</p>
<p>^ compared expected moose to observed moose. I think habitats 2 & 3 (the ones near the edge) is more correct than just habitat 3, since the question was worded plurally.</p>
<p>did u knw how to do question 4 part c?</p>
<p>@Delusioned...That's what I wrote.</p>
<p>Oh darn it..I did Linear Regression too. I thought "consistent" meant goodness of fit? No?</p>
<p>Did you guys to Chi-Sq Test for Independence? Some how I don't think that's right?</p>
<p>Chi square for....sameness. I dont know what its called, but its </p>
<p>Ho: Blah and blah are the same
Ha: Blah and blah are different</p>
<p>In this case, blah were the two proportions</p>
<p>hey any of you guys know how to do question 4 part c, please its really bohering me plz help!! here is a copy of the exam <a href="http://apcentral.collegeboard.com/apc/public/repository/ap08_stat_frq.pdf%5B/url%5D">http://apcentral.collegeboard.com/apc/public/repository/ap08_stat_frq.pdf</a></p>
<p>for 4 c wasn't it square root of p(1-p) / n</p>
<p>that's the standard error for a proportion</p>
<p>On that question 5, I also did the chi-square test for goodness of fit. And isn't the standard error of a sample proportion sqrt[(p^hat(1-p^hat))/n]?</p>
<p>That's what I did at least.</p>
<p>yeh that wat i did for 4 c but wat numbers did u plug in stephen</p>
<p>im not gonna do the math and i forget the exact proportion but i remember i took the average of the two proportions, then i used that as the proportion and then put it over the sum of the two populations. </p>
<p>i could be completely wrong.</p>
<p>yea i might be wrong with this but i pooled the proportions for 30 degrees & 50 degrees and used that value (p^) to calculate SE..</p>
<p>sqroot [ (p^q^/n1) + (p^q^/n2) ]</p>
<p>where n1 = n for 30 deg & n2 = n for 50 deg</p>
<p>that's how you do SE for 2 prop z tests but now that i think about it i'm not positive that's right for this problem....</p>
<p>wasnt the estimate for 40 degrees just the number of devices working for 30 degrees+the number of decives working for 50 degrees divided by the number of devices tested for 30 degrees+the number of devices tested for 50 degrees?</p>
<p>hmm i dont think thats it but then again, im no statistician and am not even completely sure on what i did. </p>
<p>^--- this comment was to febreeze not shemarabbis.</p>
<p>i did the same thing you did shem.</p>
<p>edit...</p>
<p>no i didn't do the same thing. i took 42/ 50 + 21/30 and got some number then i divided that number by 2. if thats what you said im sorry.</p>
<p>nice stephenn but for standard error i did the SD for a ONE prop z test not a Two prob like febreeze and the p i used was the proportion for 40 degrees</p>
<p>my SE was sqrt(.77(.23) / 80)</p>