How do you compare SAT subject test scores?

Colleges and Universities A-Z Massachusetts Institute of Technology
21

<p>^ That <em>may</em> have applied to SAT IIs, but as we all likely know, SAT IIs have larger curves than the SAT I. A 700 and an 800 are distinguished mainly by sloppy errors or tricks or a bad test day on the SAT I, but there’s a HUGE gap between the performance levels on, say the SAT II Math. If you study, you can get an 800 or close on that - the percentiles likely say as much.</p>

<p>Whether or not MIT cares is one thing, but I think it unwise not to try very hard for 750+ on the math and science SAT IIs if serious about attending a good school and studying a quantitative and/or scientific field.</p>

22

<p>

</p>

<p>Admissions officers should not assume that this is the case, though. </p>

<p>

</p>

<p>I completely agree; I was merely attempting to convey what I believe Chris has said, even though I firmly disagree with what he has indicated as being MIT’s admissions policies with respect to consideration of test scores.</p>

23

<p>So you’re saying that MITChris is lying to us?</p>

24

<p>^ No, I am saying that I disagree with what Chris has described, not that it occurs.</p>

25

<p>^ Yeah honestly, I would say below 750 is a huge red flag in SAT II Math, and below a 5 on the AP is too…i.e. having a bad day is OK if and only if you have an explanation and sufficient redeeming qualities. Yes, I know the AP is not considered, but I always felt there should be non-M/C standardized testing considered for any school.</p>

<p>Honestly, I myself have always been firmly on the sloppy side, am not a great test-taker and an 800 on SAT II Math was no problem for me back in the day.</p>

26

<p>I’ll have to agree with mathboy on the SAT I front here - I do believe that anyone with, say, a 750 in math (who didn’t put a huge effort) would be able to get that 800 on a better day. It’s not exactly a matter of skill, just a matter of a few sloppy errors - though I guess you could count avoiding those errors as skills in a way.
I also agree about the SAT II’s. The curves are much more lenient (-10 = 800 in physics) so even the difference between a 790 and an 800 is great, let alone a 700 and 800. And I do believe the bar for SAT II’s should be higher than a 700 - it’s generally a lot easier to get it on the SAT II than on the SAT I.</p>

27

<p>

</p>

<p>Yea sorry about that, I guess I misunderstood what you were saying.</p>

<p>The thing with SAT II’s is that they depend a lot on the quality of teaching and the material covered. My high school’s only physics class took a semester to cover Newton’s three laws; I don’t think anyone who took that could even think about scoring above a 700 on the subject without serious outside preparation. Also, SAT II’s are pretty obscure.</p>

28

<p>

</p>

<p>What do you mean?</p>

29

<p>I’m curious as to what your guys’ qualifications for judging proper SAT cutoffs are.</p>

30

<p>

</p>

<p>I know that I am not answering your question, but I am opposed to cutoffs in general: the difference threshold facilitated by the metric should dictate, in my opinion, the difference threshold for discrimination among scores at the admissions end.</p>

31

<p>Thanks Chris, for expanding on what had already been stated before. This does give me a clearer idea.</p>

<p>Yep, anything below 800 is sloppy in Math level II, but there’s only a marginal difference between say, (-6 ) raw score 800 and a (-7) raw score 790, so cutoffs do make sense, when they are adjusted according to the respective curves ( I hear German has a nasty curve).</p>

32

<p>@shravas</p>

<p>I think with the transparency of what’s covered on the SATs plus the availability of preparation tools, it’s no longer the classes which officially prep you for them. I agree the SAT IIs are strange, and to give an idea, I never took the physics one, and never looked back. But for instance, many things on the SAT II math weren’t officially covered in my curriculum; I never had precalculus. I think it’s no excuse not to study hard for the tests and show you can do basic things.</p>

<p>@Piper </p>

<p>At least from my end, I of course don’t claim to know what schools actually do. How I judge tests is from seeing what they cover, distribution-wise, hopefully having taken them, and noticing what skills one should have for the future to do math and science work…or what skills should be easily demonstrable by someone who shows promise, even if themselves not so useful.</p>

<p>People can get somewhat lesser results on tests for a number of reasons. I guess I meant “red flag” as “This is bad, unless other evidence counteracts it”. </p>

<p>I guess how do we judge anything at all? We’re students. We spend a lot of time thinking about being students, and certainly should have some intelligent things to say about the things we had to do in the past, as they correlate to the future. We’re actually [many at least] doing some math/science. We’re the professionals of tomorrow in these fields, and have had recent experience making the transition [or are about to make it].</p>

33

<p>the difference threshold facilitated by the metric should dictate, in my opinion, the difference threshold for discrimination among scores at the admissions end.</p>

<p>…or what skills should be easily demonstrable by someone who shows promise, even if themselves not so useful.</p>

<p>And it appears that Admissions has not found the upper differences to be important in the performance of a student at MIT. I promise you, they didn’t arrive at their magic number because it sounded like a fun thing to do :)</p>

<p>(I also want to point out that it’s not a strict cutoff.)</p>

34

<p>

</p>

<p>I wonder how valid their statistical analysis was, though: if those with higher scores had been given an admissions advantage, those who got in with lower scores must have had other qualifications that correlated with academic success.</p>

35

<p>I’m not in favor of strict cutoffs either, but I am in favor of noticing that below certain scores <em>requires explanation or other promising evidence</em>. </p>

<p>I’m not analyzing things based on statistical games indicating success or failure at MIT. For all I know, a 700 and an 800 on the SAT II Math may be of negligible difference in terms of performance at MIT.</p>

<p>However, these are loaded terms. What does “performance at MIT” mean? Does that mean the rate of passing all your classes? Does that mean rate of acing all your classes? Does that mean rate of getting into elite grad schools / elite careers?</p>

<p>I should say that I post where I do because I’m a math/science enthusiast [not out of deep knowledge of a school, aside from knowing it’s serious in these areas], and can certainly say to a reasonable degree when there’s a funny assumption about related basic educational principles. Frankly, someone who can <em>thrive</em> at a science or math discipline in any serious school has to have so many skills and so much preparation and ability that the skills and level of test-taking required on the SAT Math II in order to ace it should be easily demonstrated…no statistics are necessary to do this, only analysis of the test, what it entails, and what skills being a solid thinker entail. However, factors such as motivation [giving a damn about the test], having good health on the day of the test, etc, figure in. * If other positive evidence invalidates the concern caused by a lower score, * then fine, who really cares about the score. Like I said, scores are markers of basic competence, the absence of which raises some worry, which can then potentially be set to rest.</p>

36

<p>Frankly, someone who can <em>thrive</em> at a science or math discipline in any serious school has to have so many skills and so much preparation and ability that the skills and level of test-taking required on the SAT Math II in order to ace it should be easily demonstrated…</p>

<p>I disagree with the preparation aspect. While preparation certainly helps, people do come in without adequate preparation and learn to thrive anyway. IME, this usually just makes the difference between what year you find hard. People with no preparation will find freshman year hard, but learn enough to have an easier time through the rest. Other people will feel the pressure ramp up during junior year or so.</p>

<p>As for the validity of their analysis/loaded terms - this basically amounts to, “This job is hard”. I don’t think anyone in Admissions would disagree?</p>

37

<p>Really, I don’t see how what you’re saying actually disagrees with me. It’s either juggling terminology or expanding on how what I said may be viewed. Of course, different people are familiar to different extents with what is taught in the general requirements in frosh year, and that affects when they feel their first struggles. If someone manages to thrive later on, and can pick up on the basics well during frosh year, then they’re well prepared to thrive. </p>

<p>But there have to be strong indicators that this will happen, I’m sure anyone will agree - if not, then you might as well roll dice instead of evaluate candidates.</p>

<p>The point of my comment that “performance at strong school X” is a loaded term of course entails that the job is hard. I know nobody would disagree. </p>

<p>This is why I keep repeating - I’m not an adcom. But I am sincerely interested in math/science education, and think about which skills correlate to what, what should be a very feasible task for someone with a future in serious math/science thinking, based on my knowledge of the skills required … and then, of course there are cases where the tasks are not completed not because the level of preparation for future success never was there, but because people either had exceptional circumstances or just didn’t care [which is fine, and doesn’t disqualify them as future academic gems, but raises the concern as to what does qualify them - and that’s where the adcoms’ magic needs to come in].</p>

<p>I just think it’s completely unfounded not to wonder why someone got, say a 700 on the SAT Math II as opposed to closer to an 800, because assuming the person studied, which they hopefully did, there’s a HUGE gap between the two scores, and that test honestly requires very little sophistication. I never had the explicit preparation for a lot of what’s covered there, never had a precalculus course, am not a great test-taker, and it was still no problem for me. It’s fine to have a bad day if you make it up somehow, but it’s a serious claim to me to say that the difference in those scores shouldn’t cause some concern, when I think most people who have done basic math and science would probably disagree.</p>

<p>Basically what I’m saying is, if you can say you’ll sit down and thrive at really hard math and science some day, you <em>very likely</em> can sit down, read some basic prep material for such a test, and ace it, since they let you make tons of mistakes and the curve is insane.</p>

38

<p>It’s a 60 minute test with 50 (I believe -?) questions. I think they adjust for slow test-takers or compensate for careless errors using the curve. Even the SAT I is 54 (simpler by far) questions - with 70 minutes. The curve, in my opinion, wouldn’t be fair otherwise. That said I don’t know the actual difficulty/content of the test as I’ve yet to take it, so my opinion might change when I try one out. </p>

<p>At any rate, this argument is sort of moot. It always comes down to the same conclusion - just do your best on the tests and apply. And take the interview :)</p>

39

<p>Well, I don’t even call this an argument, it’s unclear if people are even talking about the same things and really disagreeing, and if so, what the basis is. I don’t think there really is much more to say than what Jimmy says - do the best you can [as in, really try, and don’t BS yourself about which score ranges are the same], but always prioritize what you think is useful for your future. If you find yourself for some reason really disadvantaged by a standardized test, obviously don’t get fixated on it and spend your life on it…focus on what you think will make you succeed in your long term goals. And keep in mind, odds of getting into any great school, including MIT, are against you, so keep broader goals in mind.</p>

40

<p>I’m not sure if this is a reply directly to anyone, or just to the thread as a whole, but it’s interesting to note that the graduate school version of the SAT, the GRE, is widely disdained as a predictor of graduate school success, at least in the fields I know well. In some places, like MIT’s incredibly competitive EECS PhD program, GRE scores are neither required nor considered.</p>

<p>At most programs, applicants must submit GRE scores, but they are considered in a fairly superficial way, consistent with the the way MIT undergraduate admissions is looking at them – something above a 700 in the quantitative section is fine for any top PhD program in the sciences, even though the GRE is considerably easier than the SAT, and fully ten percent of test-takers score a perfect 800, and the professors who make up the admissions committee are explicitly looking for brilliant future scholars rather than caring about anything other than brilliance.</p>

<p>

This is something the admissions office, of course, analyzes on a continuous basis, with a huge dataset going back many years.</p>