<p>Elsewhere in cyberspace, I got into a discussion about whether there is ever any risk in reporting more than previous SAT I score if at least one of the scores is competitive for the applicant's desired college. The two positions that have come up in online debate are</p>
<p>1) it makes no difference how many SAT I scores you report, as long as one is a solid score for the college of your choice,</p>
<p>and</p>
<p>2) it's better not to report another SAT I score, lower than your best, because that might cause your best score to look worse.</p>
<p>What do you think? I suppose Ben Golub may have some thoughtful observations about how this issue works out in actual Caltech admission practice. And besides Caltech, what does any other college say about this? Is it best to report only one SAT I score? Is it a risk of being rejected for admission to have both an amply high score and also a lower score? </p>
<p>Thanks for all advice and links you can recommend about this issue. </p>
<p>Well, the entire score report is required so they see everything. Then the issue is whether the admissions officers actually look at the report from CB or just look at what's on the app.</p>
<p>Interesting question. The official policy is that the best score is the one that counts most. </p>
<p>The summary sheet in every folder does list all scores (which the CB reports). Usually I just glanced at all of them and took the best ones as authoritative. The only cases in which the lower scores mattered were the ones where there were many of them and the final one wasn't so great. So someone taking the Physics SAT II and getting 610,640,700 would probably be worse from my perspective than just getting a 700. The 700 might be a "real" 750 just having a bad day, whereas there is less doubt when you have a sequence like that.</p>
<p>But the more usual scenario would be SAT I Math: 730 then 790. Or SAT II Chem: 700, 760. In those cases we just took the highest one and didn't spend any time thinking about it.</p>
<p>So unless the scores betray a trend of very mediocre scores capped by a final mediocre score, then the lower ones really don't count for much at all.</p>
<p>Let me know if this is at all clear -- I'd be happy to answer questions if it isn't.</p>
<p>Where would you place the line between line between statistical uncertainty and actual improvement? Your example of 610, 640, 700 could be seen as random scores about an average in the mid-600s, but 600,700,800, for example, hardly seems consistent with the College Board's (iirc) +/- 30 point uncertainty. And, if you were to consider the 600,700,800 sequence a case of "actual improvement," would the final score of 800 be just as good as scoring 800 the first time? (Is the 600,700,800 example too rare to treat except on a case-by-case basis?)</p>
<p>I think Ben's reply is very clear, but I've been reading some rather thick comments about this issue on Brand X, so I may have to ask for follow-ups. Some people don't get things as fast as Caltech students. Just to be really clear, what if the trend is </p>
<p>middle-school age 800 </p>
<p>high-school age 710 </p>
<p>on the math section of the SAT I? (This is a made-up example; I'm mostly focusing on whether the "trend" evidence matters in such a case.)</p>
<p>In every case I have ever read about , students who took the SAT in middle school without exception scored hundreds of points higher in math by their senior year, simply because they contiinued to take more math classes in HS! My son is but one example- had to take SAT for CTY in 7th grade- scored 510 in math, senior year scored 790.</p>
<p>Hi, menloparkmom, I actually have heard of exceptions to the general trend you mention, which I agree is the majority case. A friend of mine who is a math team coach in another state told me that he usually hears about the SAT I scores of the MATHCOUNTS state winners in his state (as I do in my state) and that those scores are often 800 on the math section by age thirteen. Those kids go on to study post-calculus math while in high school, and in the cases he hears about they NEVER repeat their scores of 800. The only way is down from 800, and most kids taking advanced undergraduate math classes have lost at least a little bit of familiarity with SAT-level math by then. </p>
<p>My contention is that, knowing the college admission results (GREAT) of kids like that, there is no fear whatever in taking the SAT I incidentally twice or three times during high school age. Any low score(s) will simply be disregarded during consideration of the high score(s). Ben appears to back me up on this, and I haven't seen anyone with a contrary opinion cite any public statements from a college admission officer directly on that point.</p>
<p>" knowing the college admission results (GREAT) of kids like that, there is no fear whatever in taking the SAT I incidentally twice or three times during high school age."
Agreed, but the college admissions results for thse kids are not just a function of their SAT scores. Colleges scrutinize the transcripts, which can be as important, if not more importantin many cases than SAT scores, and any student who is successfully taking post calculus math is given credit for taking more rigorous courses than 95% of most college applicants, and will be evaluated at a different level.</p>
<p>^^ Exactly. The high school kids who are competitive at those most selective colleges really don't need to worry about their SAT I scores--they are competitively above the "don't even think about it" threshold, and are distinguished from other candidates for admission on other grounds.</p>
<p>I think an 800 in middle school followed by a 710 in high school would get a laugh or two in the admissions office and the kid would go in the 800 column (surely he gets points for doing better on the test in 8th grade than most people can ever, ever do). I mean, I doubt that it would ever dip that low on the second try, but even if it did it would not be a big deal.</p>
<p>So I am with tokenadult here. The only time when trend evidence hurts is when it confirms a worrisome hypotheis, i.e. low best scores that are backed up by a history of low attempts.</p>
<p>I can just picture someone on CC reading that last post and going "Caltech admissions has an 800 column?!?! Do people in it get an automatic boost?? <em>must get in it</em>"</p>
<p>"Those kids go on to study post-calculus math while in high school, and in the cases he hears about they NEVER repeat their scores of 800."</p>
<p>That's a little surprising, actually. I had a significantly bigger vocabulary in middle school than high school due to three years of spelling team, but my verbal score still went up 40 points from 8th-11th grade.</p>
<p>Still, I thought middle-school SAT results didn't even appear on the score report by high school. They didn't used to.</p>
<p>
[quote]
Still, I thought middle-school SAT results didn't even appear on the score report by high school. They didn't used to.
[/quote]
</p>
<p>The College Board routinely deletes from a student's record of scores all scores from eighth grade or below. I found out to my chagrin last year that my son's SAT II Math Level 2 score from eighth grade was poofed from his record for a time--nobody takes SAT II tests through the Talent Search programs--but I was eventually able to find a College Board supervisor with understanding of the issue and enough authority to restore that score to his permanent record of scores. Some parents ask for those scores to be preserved before they are removed from the students' records, but that takes affirmative action on the part of the parent. </p>
<p>But, yeah, the case you mention would be a possible case: kid gets 800 in middle school age, LOSES the score, and then gets <800 at high school age. (I'm referring to math; I agree that the most likely pattern on the verbal side is a score gain into high school.) It seems to me that the Caltech admission committee is smart enough not to obsess about such details.</p>
<p>I understand that you were referring to math. It is still surprising to me.</p>
<p>Frankly I can't really understand getting (significantly) less than 800 on math at the high school level if you'd done it in middle school either. Yeah, I took advanced math classes in high school, too, but it's not as if I forgot how to multiply... I've done 4 years of college and 2.5 years of grad school now and I still think I'd do fine if you sat me down and had me take an SAT math section right now.</p>
<p>With the old verbal section so much a function of straight-up vocabulary (less so now, but still the case to some extent), I would consider a pattern of decreasing verbal scores after having memorized long lists of words in middle school to be the trend, which was my point--but it didn't happen.</p>
<p>
[quote] tokenadult wrote:
SAT I scores of the MATHCOUNTS state winners in his state ... are often 800 on the math section by age thirteen. Those kids go on to study post-calculus math while in high school, and in the cases he hears about they NEVER repeat their scores of 800. The only way is down from 800, and most kids taking advanced undergraduate math classes have lost at least a little bit of familiarity with SAT-level math by then.
[/quote]
</p>
<p>tokenadult, what was the progression of their verbal scores during that time?
Were there any students with a statistically significant drop in the math scores (it takes a lot of wrong answers to go below 740)?</p>
<p>Any university will take it as obvious that having state or national level distinction in math competitions, 800 on SAT-II math, 5 on Calculus BC, and post-calculus math courses (which tend to come packaged with further high performance in physics or computer science), is far beyond whatever is measured by an 800 on the SAT-I math. Your friend's sample of gifted mathematicians just shows that for some applicants, their SAT math scores are rendered irrelevant, over and above the statistical insignificance of the minor drops in their SAT scores (800 to 760) that one would expect for such a group.</p>
<p>As the Caltech admissions reader indicated, selective universities can and do discern differences between different patterns of SAT's. Admissions officers are looking for signal within noisy data, which means that the best data (for the admissions officers, not the applicant) would be to take SAT several times and get the same scores. A single sitting is also good, because the scores are the scores by default; there won't be too much speculation about higher or lower possibilities unless the scores are seriously inconsistent with other information. Multiple scores all over the place can invite questions, whether or not they trigger a detailed investigation for any given application.</p>
<p>tokenadult, did your sample's verbal scores progress upward or downward? It's a simple empirical question whose answer you (and nobody else here) can easily reveal or determine.</p>
<p>Your apparent position on the multiple-SAT question in several threads here, has been that there is never a difference between (a) submitting multiple SAT scores and (b) submitting one score equal to the "superscore" of same. Ben Golub, in this thread, has informed you of one difference at Caltech. Admissions officers from Harvard, Princeton, and others have written about this issue and confirm that there are differences (or a definite potential for differences) in the handling of the two situations --- for example, all of them advise not to take too many SATs. People on CC who asked admissions officers at Harvard and elsewhere have posted information as well, and it indicates that "superscoring" is not the whole story. </p>
<p>In general, see above regarding signals and noise; any admissions officer (and I know some from Harvard, MIT and others) can tell you that this is the essence of their job, and wildly fluctuating SATs are an example of noise.</p>
<p>I am actually deeply baffled by this discussion. Admissions officers at most top places do not spend time reading SAT's like tea leaves to see what a certain fluctuation reveals. Certain extremes will set off alarms -- like 10 repeats or a 400 after an 800, but most of the time they are glanced at to make sure the person is basically in range for the school and we go to other things. Usually, for top academic achievers, there is such plentiful evidence that they are intellectually tops in math and science that SAT's can't do much to add or detract.</p>
<p>I understand that the intellectual debate about what a certain score reveals or doesn't reveal can be interesting, but in particular cases the point is very often wiped out by things we view as stronger signals. Just my two cents.</p>