<p>I know there's probably a lot of these threads floating around the forums, but I just thought it'd be nice to have an official thread with people's experiences with the SATs in different months.</p>
<p>Which months appear to be the best "curve" months for the SAT?</p>
<p>May, June, October. This is when the typical crowd takes the SAT. A month like January only attracts the hardcore SAT nerds.</p>
<p>I would say possibly November because most people who have gotten like 1500s have already gotten them in May, June and October. November is more of a last ditch effort for people who are not doing that well to do better.</p>
<p>I thought January was best, because that's when Juniors finally take it for the first time?</p>
<p>I knew some seniors who took it in Jan because if you do rush reporting, it still makes it to colleges on time.</p>
<p>November kind of makes sense? </p>
<p>But I did get a bit higher on my OCtober SAT score than my practice tests. And when I took it in January, it was lower than my PSAT score. Does that say anything?</p>
<p>is there any "best curve" month at all??</p>
<p>I actually thought that it'd be January because rationally, that's when all the seniors stop taking it......, so I was planning on taking it January.</p>
<p>I guess I'm wrong, I don't know.</p>
<p>There isn't. This is all a myth, a horrible, widely-spread myth.</p>
<p>Is it really just a myth? Why is that?
It seems like there would be one..since the number of people taking it varies throughout the year.</p>
<p>Well I took it last january as a junior and the curve for CR and writing was reaaally good...like -3 =800 for CR and writing -3 was still like 780...dont remember about math..ill look it up when I have time.</p>
<p>It's more or less true that no date is better than any other, because the test is not curved relative to the group of students who are taking it on the same day, it's normed relative to the entire set of test-takers over multiple years. However, because of the way the equating sections work, there are some subtleties in this. Based on what CB says about their method of converting raw to scaled scores, I suggested that there might actually be better dates in an earlier thread:
<a href="http://talk.collegeconfidential.com/sat-act-tests-test-preparation/518574-curve-sat-test-date-better.html%5B/url%5D">http://talk.collegeconfidential.com/sat-act-tests-test-preparation/518574-curve-sat-test-date-better.html</a>
My suggestion is just based on a hypothesis, but it should be empirically testable by the CB statisticians, in principle.</p>
<p>I think your hypothesis is more believably about whether test-prepped people can improve their scores than about which tests dates are the "best".</p>
<p>But we don't need CB statisticians. Looking at all the QAS curves so far since the SAT changed in March 2005, I just don't see a month that has a consistently generous or consistently harsh curve. For months other than Jan, May, or Oct, I imagine there are plenty of people here on CC with curve data to fill in the gaps.</p>
<p>fignewton, take a look at how CB converts the raw scores to scaled scores. Now, suppose that you could take the SAT along with a large set of co-testers who do better on the equating section than they do in general. That would definitely tend to boost your score, assuming that you are good compared to your co-testers--there's nothing hypothetical about that. </p>
<p>My hypotheses are that extensive test prep tends to raise students' performances on the equating section more than it raises them on the general test, and that heavily prepped students are more prevalent among the testers in some months than others. If these were true, it would actually make some months better than others. If you can get the recent curve data for all of the months, we could see if there is any pattern.</p>
<p>I understand your hypothesis, I just don't find it believable :)</p>
<p>I think it is hard to make the case that test prep companies hire people specifically to memorize the equating section questions (there is no other way legally to see those questions). Any effect on the equating section score distribution would require large numbers of these "equating-prepped" people to take the test in a particular month. Your hypothesis must also explain why there is no consistently generous curve for the months of Jan, May, and Oct.</p>
<p>But again, the people on CC probably have the curve info, why not start a thread to see if your hypothesis can be supported by data?</p>
<p>Well, fignewton, you might be right. :) The effect would probably be relatively small (plus or minus 20 points near the high end) in any event, and it could be swamped by other test features. But I agree that it would be interesting to use CC to test for any pattern in the curves for various months. It would also be interesting to know whether there is any observable difference between the months when the question and answer service is available (months viewed collectively), and those when it's not.</p>