<p>Hi, everyone, </p>
<p>I have another CC participant to thank for pointing out a College Board chart a few years ago showing how many students got each level of composite scores on the old SAT I. Now I see that there is a similar chart of new SAT I scores </p>
<p>demonstrating that it is even harder to ace the new three-section format than the old two-section format.</p>
<p>Thanks for posting, Tokenadult.<br>
2280 is the cutoff for 99+%; 2200 is the cutoff for 99%. Quite a gap from 2400.</p>
<p>I know one person ever who has gotten a 2400. I know 3 or more who have 2390s. The thing is you can get 2400, but most dont because they dont have the stamina to take a 5 hour test and be flawless. On my 3 SATIIS combined I scored 250 points more than my 3 sections combined on my SAT. I think college board should change SAT from one test to SATII format where you take 3 one hour tests and you can either do all in one sitting or one each month. I think 2400s will go up then.</p>
<p>Interesting, and a good bit of marketing by the College Board to get everyone to think in terms of all three tests when colleges have been uncertain how much weight, if any, to give to the writing section. Perhaps more meaningful to speak of "1600 with an 800 in writing"</p>
<p>You don't have to be flawless to get a 2400. On my test date, missing 1 CR, 1 Writing Mult Choice, getting an 11 on the essay, and missing 0 math = 2400.</p>
<p>Yes, since recentering, perfection in the item content is not necessary to attain the highest standard score on the SAT I. But it's still quite impressive to ace all three sections.</p>
<p>My friend's niece got 2400. I was impressed before--now I'm REALLY impressed! Only 107 girls in the country!</p>
<p>Also, this chart points out how crazy the scores on CC are. Almost every kid who posts on the Chances? board has scores over 2100--any lower, and all the other kids say the OP has no chance anywhere. Well, 2100 is 97th percentile. That sounds pretty good to me.</p>
<p>I know of one or two 2400 scorers who are NOT on that list--because they are in high school class of 2007, not 2006. In other words, the number of young people who have scored that high is greater than what the chart shows, because the chart reports numbers of students in a single graduating class. </p>
<p>Being "only" in the 97th percentile is NOT particularly competitive for students desiring the four or five most competitive colleges, unfortunately. That skews what is said on CC about applying to other colleges, which certainly do admit students with 97th percentile scores.</p>
<p>A friend of my sons got 2400 if (like some colleges) you take the best scores from each time. I know this isn't the same since his best individual score was 2360. The interesting thing was his GPA was about a 2.8 (a genius who refused to work in high school).</p>
<p>There is something that confuses me about this distribution. Since the test isn't designed to test the upper reaches of a student's knowledge, and since perfection isn't required for a top score, I would have expected that the number of students scoring 2400 would exceed by some meaningful margin the number scoring 2390 or 2380. The 2400 score would essentially aggregate a much longer "tail". I thought it worked that way when 1600 was the top score.</p>
<p>But that seems not to be the case. The number of students scoring 2400, 2390, and 2370 is basically the same (and only 2380 noticeably less).</p>
<p>Also, let's appreciate how much of a difference the writing test makes. For the last few years, the number of kids with 1600s hovered around 1,000, and maybe 2,000 in the 1580-1600 range, and 5,000 in the 1540-1600 range. With the writing test added, the top 1,000 goes down to 2360, the top 2,000 is basically 2340, and the top 5,000 goes down below 2300. In other words, 2360 on the new test is basically equivalent to 1600 on the old. </p>
<p>However, interestingly, 2340=1580 and 2300 = 1540. The top 1,000 scores are distributed over five different data points (vs. one on the old test), but the next 4,000 scores are distributed over only 7-8 data points vs. 6 on the old test -- really not that much more differentiation.</p>
<p>Your observation actually plays out when you look at each section individually -- <a href="http://www.collegeboard.com/prof/counselors/tests/sat/scores/data_tables.html%5B/url%5D">http://www.collegeboard.com/prof/counselors/tests/sat/scores/data_tables.html</a></p>
<p>For some reason the effect disapears when you combine the sections.</p>
<p>There's another factor to consider when looking at scores between 750 and 800. On some administrations of the test, it is impossible to earn some of the specific scores in this range.</p>
<p>For example, my daughter got a 760 on the Math section on a particular administration of the test where 760 represented two wrong answers. On this particular SAT math section, a paper with one wrong must have earned either a 770, 780, or 790. I don't know which. I do know, though, that on this particular test, two of those three scores were never given out. How could they be?</p>
<p>When the college board releases these numbers, they mean for one sitting, correct? So there may be a lot more students who can claim 2400 if they mix and match their scores. Do we know how many students got a 12 on the essay? I know they said, less than one percent, but do they list the actual numbers/</p>
<p>Marian: If you look at the individual tests, there were some of each score given out, but the distribution is very uneven. For some reason, it was very hard to score 770 on the math test, and extremely hard to score 780 on the CR test (1,600 790s, 4,000 770s, and 500 780s). </p>
<p>I am sure this relates to how unanswered questions are scored.</p>
<p>"For example, my daughter got a 760 on the Math section on a particular administration of the test where 760 represented two wrong answers. On this particular SAT math section, a paper with one wrong must have earned either a 770, 780, or 790. I don't know which. I do know, though, that on this particular test, two of those three scores were never given out. How could they be?"</p>
<p>My son took that Math section. :) Presumably he would have scored a 770, 780 or 790 if he had omitted both questions or had one wrong answer and one omit instead of two wrongs. Admittedly that's two possiblities for three possible scores.</p>
<p>mathmom, that's what I thought, but my daughter says no. She says that because there is a rounding step in the scoring process, the so-called guessing penalty really doesn't apply for the first two wrong answers.</p>
<p>Here's a quote from the College Board's book The Official SAT Study Guide:</p>
<p>
[quote]
You receive one point for each question answered correctly. For each question that you attempt but answer incorrectly, 1/4 point is subtracted from the total number of correct answers. No points are added or subtracted for unanswered questions. If the final score includes a fraction, the score is rounded to the nearest whole number. [Emphasis added]
[/quote]
</p>
<p>My daughter says that this means that because of the rounding, 2 errors = 1 omission + 1 error = 2 omissions. Also, 1 error = 1 omission. The penalty for wrong answers doesn't become meaningful until you have three wrong answers. She also says that this is actually an important strategy point for kids who are working in the extreme upper score range (above 750). In this range, it makes sense to guess.</p>
<p>Of course, my daughter could be wrong. After all, she only got a 760 in math. ;)</p>
<p>See </p>
<p>for comparable figures from an earlier year, showing distribution of scores on the old, two-section SAT I. </p>
<p>Yes, these are single-sitting scores. The students who take the test more than once gets only their highest single-sitting score reported on the chart. The chart does not show the distribution of scores for class of 2007 students who have already taken the new SAT I--there are known instances of 2400s among those students.</p>
<p>Our son had 2400 in April, with one wrong answer on reading part (which he noticed immediately after he walked out of the test room). He got 12 on his essay too. He is in the class of 2007, by the way.</p>
<p>Son got a 2400 in Dec 05. Plus/minus mid 700's that spring when he got a 35 on ACT -no studying, of course. I wonder how many of the 238 students with perfect scores had to study or did the score reflect natural ability? Son was supposed to be retaking the math SAT II for which he hadn't studied ("we're reviewing precalc math in calculus"...), but the test wasn't offered so he took the whole thing over. Don't ask about his logic- he turned 16 last fall. Also, don't ask about his senior grades- doing and turning in homework, not just acing tests counts in the grades here. I should start a thread about gifted and talented parental trials and tribulations...</p>