<p>i used .79 not .77, you prob did also but just dnt remember</p>
<p>what were we supposed to do on 6c and d? that was the only part of the test that really ground my gears.</p>
<p>on part d i drew pictures.</p>
<p>6c involves using the data on the LSRLs for the t statistic, t = B/SE...slope over standard error (for example the magnet is .1811/.4583), and dof = 7 and 11 respectively, so you can find the p value for magnet and original, and reject/do not reject as needed.</p>
<p>for question 6 how did u guys interpert the slopes?</p>
<p>just to be sure..wut test did u guys use for 6a? for 6c u can actually just find p just by reading the regression analysis table</p>
<p>I said that for every 1 point a students pretest score rises, the posttest scopre increases by (slope) points, which is added onto the base score of (y intercept) that would be gotten if the pretest score was 0. Something like that.</p>
<p>6a: two sample t-test (independent) for the change in science test scores (the mean difference between pretest and post test for each block)</p>
<p>reject null hypothesis that the mean change in scores where the same for those that stayed in their original school was the same as the mean change in science scored for whose who change to the magnet school</p>
<p>how about for 6d?
r we not allowed to iscuss mc?</p>
<p>Crap! They did give you exact values of p, didnt they? Guh.</p>
<p>no idea :-S</p>
<p>i put that one group's scores (post with pre-test) could be predicted with a linear regression (the ones who stayed in their original school) but that those that moved to the magnet school could not (i got a large p-value for this group).</p>
<p>i know, i really just guessed.</p>
<p>yeah i think so, i remmeber that for the last questions i didn't do any computations. i just had to interpret the data they had already given me.</p>
<p>barron's was great help with this. my teacher never got into detail as to how to read a computer output.</p>
<p>shoot for 6a i used the wrong SD but i did everything else right, you guys think they will give me no credit for part A bc of that?</p>
<p>@shemarabbis</p>
<p>each part of a problem is independent from the last.</p>
<p>on some problems you can get part "a" COMPLETELY wrong, but if you follow up using your conclusions reasonably and do everything else right, you can still get a 3.</p>
<p>(not a 4.... for that, the solution needs to be perfect, i.e. with only minor mistakes)</p>
<p>can we talk about mc or are we not allowed to?</p>
<p>can somebody prove that last year's curve was 60-100 for a 5</p>
<p>just to clarify what i was talking about earlier with 4c...</p>
<p>pooling means (in the context of this problem) doing (42 + 21)/(50 + 30)...again, i'm assuming they wanted that for the problem & not just the straight average of the two..?</p>
<p>sorry about the big break b/w posts, i'm simultaneously studying for euro...:eek:</p>
<p>thats wat i did febreze, did anyone else do that?</p>
<p>Here is my solution to problem 6: (i felt confident about this problem, except for part D,). i should not be taken as the last authority.</p>
<p>a)two sample t-test (independent) for the change in science test scores (the mean difference between pretest and post test for each block)
reject null hypothesis that the mean change in scores where the same for those that stayed in their original school was the same as the mean change in science scored for whose who change to the magnet school</p>
<p>b)</p>
<p>y-hat=.1811x+73.27>>>>>>>>>magnet</p>
<p>every one point increase in score for the science pretest for the students who changed to magnet school corresponds to an increase of .1811 for the post-test score.</p>
<p>y-hat=.9204x+9.2.4>>>>>>>>>>>original school</p>
<p>every one point increase in score for the science pretest score for students who stayed in their original school corresponds in a .9204 point increase in post-test score.</p>
<p>c)</p>
<p>FOR MAGNET:</p>
<p>p-value= .706</p>
<p>conlcusion: at a statistical significance level of (alpha)=.05 we cannot reject the null hypothesis that there is no linear association between the mean the pretest score and the mean post-test score, for those that change to the magnet school</p>
<p>FOR ORIGINAL SCHOOL</p>
<p>p-value=0.000</p>
<p>conclusion: at a statistical significance level of (alpha)=.01, we reject the null hypothesis that there is no linear association between the mean pretest score and the mean post test score, for the students that stayed in their original school. </p>
<p>D)</p>
<p>NO IDEA :-S</p>
<p>my dad (whose work involves some statistics) said that he thought 6d had something to do with the fact that the observed differences were probably affected by the nonrandom assignment of students into groups (i.e. those attending original school had been rejected from magnet)</p>
<p>that was solely based on my garbled version of the problem though...</p>
<p>basically, i thought the question was just wayyy too broad.</p>