<p>I've heard of a few cases of higher SAT I results; however, the vast majority of the posts on this topic that I came across on CC were about better scores on the ACT. Is it true that most (or most CC kids) score higher on the ACT? If so, why and how will this trend shape the standardize test landscape?</p>
<p>In my house, I have one D who prefers & tests better on the SAT, while the other prefers & tests better on the ACT. Not terribly different 1460 CR+M (which on the concordance table puts it at a 33 ACT) vs. a 32 ACT. </p>
<p>Not quite sure what you mean about the trend shaping the standardized test landscape though. Both tests have been popular for quite some time, although the ACT has been taken less in the Northeast.</p>
<p>What I meant is that IF most kids who took both tests scored higher on the ACT, then this may lead to a change in our current standardized test landscape. Perhaps this is one of the reasons that ACT is gaining momentum.</p>
<p>It’s hard to tell how most here on CC compare, and how those results would correlate with the general student population. How and how often are the conversion tables calibrated? Both mine so far did better on ACT (33 vs 1380/2060 and 31 vs 1330/1940).</p>
<p>sry, pop, but that is not logical. The concordance tables are developed based on millions of kids taking both tests. “Most” kids do equally well on both. Some kids do better on one or the other.</p>
<p>IMO, the ACT is growing due to several factors: 1) a much more sophisticated app pool – in the old days, no one on the California (CB’s largest market) had even heard of the ACT; 2) Counselors tell kids to take both to see which one is best for them; 3) score choice; 4) some Midwestern states mandate and pay for the ACT as a HS requirement.</p>
<p>Btw: if it WAS true that most kids score higher on the ACT, someone would have leaked that ‘secret’ years ago.</p>
<p>Both my D and S1 did better on ACT.</p>
<p>D got a 34 on ACT (1410/1600 on SATI) - 2 tries for each test. I would consider her equally strong in verbal/Math/Science</p>
<p>S1 got 33, then 35 on ACT (1400/1600 on SATI - one sitting only for NMF) - definitely a Math/Science kid.</p>
<p>S2 took ACT for the 1st time on Saturday - will probably not take SATI unless he needs it for NMF or if PSAT shows he is very strong in verbal section. Got a 790 on Bio SAT II, so I am thinking he will be stronger on ACT since it plays to his strength in science. We will see…</p>
<p>My S scored equally as well on both.</p>
<p>
</p>
<p>Not true. That is, it may be true in individual cases, but over the larger run of cases it’s not true. My D1 always did better on the SAT than on the ACT; among her friends who took both tests, I’d say the split is about 50-50, with half doing slightly to significantly better on the SAT, and half doing slightly to significantly better on the ACT. </p>
<p>In fact, how we determine what it means to do “better” on one test or the other is defined in a way that pretty much eliminates the possibility that students in one group would systematically do better. The ACT-SAT concordance—the chart that is used to compare ACT scores to SAT scores—is based on comparable percentile ranks; so the SAT score that represents the 99th percentile of SAT-takers is matched up with the ACT score that represents the 99th percentile of ACT-takers, and so on. The population of test-takers in each case is about 1.4 million, a large enough sample to be pretty representative. </p>
<p>One factor that might slightly bias the results: In ACT-dominant states, it tends to be mainly high-achieving kids who take the SAT. They take it because many elite schools express their admissions stats in SAT terms (even if they accept the ACT), and the top students in ACT-dominant states want to see how they stack up. (They also need to take SAT IIs for some top schools, but that doesn’t give them a reason to take the SAT I reasoning test). Also, National Merit semifinalists may need to submit SAT scores to qualify as finalists. In contrast, in SAT-dominant states the kids who take the ACT are usually kids who aren’t satisfied with their SAT scores, consequently they’re looking at the ACT as an alternative that will give them a higher score and a better chance at admission. The combined effect of more high-achievers from the ACT-dominant states being in the SAT pool, and more slightly-lower-achievers from the SAT-dominant states being in the ACT pool might be enough to slightly skew the results, with the SAT pool being slightly higher achievers on average. But I doubt the numbers in either group are large enough to influence the final outcome very much. By far the largest numbers of ACT-takers come from ACT-dominant states, and by far the largest numbers of SAT-takers come from SAT-dominant states.</p>
<p>Another factor that might possibly produce a skew: six states now require ALL HS juniors to take the ACT. I’m not aware of any similar requirement for the SAT. This would tend to expand the population of ACT-takers beyond the usual college-bound HS juniors and seniors, driving down the average achievement level of ACT-takers in those states (and note that in all 6 states, the statewide average ACT composite score is slightly below the national average). But again, it’s difficult to say how big this effect is. Some big SAT-dominant states also have a very high level of participation; in New York, for example, 87% of HS grads take the SAT. And any skewing of the ACT pool toward slightly lower average achievement levels as a result of universal participation in some states might be counterbalanced by other demographic differences, e.g., higher proportions of Latino and African-American test-takers (who historically have not tested as well, on average, as white or Asian kids), non-native English speakers, or first-gen college aspirants in some other states. For example, the census bureau estimates that half the nation’s foreign-born population lives in just 4 states: California, New York, Texas, and Florida. Of these states, California, New York, and Texas are all SAT-dominant, while Florida is pretty evenly divided with 65% taking the ACT and 54% taking the SAT. So if standardized test scores, are, on average slightly lower among the foreign born (and I don’t know whether they are or not), then this would tend to skew the SAT pool toward a slightly lower average achievement level. My guess is that if you added up all these sorts of factors, it would be pretty much a wash; 1.4 million in each pool is an awfully large sample.</p>
<p>I think you may HEAR more about CC kids getting higher ACT scores because CC users tend to be heavily concentrated in the Northeast, and the Northeasterners who take the ACT are largely going to be kids who didn’t score as well as they’d have liked on the SAT. The half of this group who score better on the ACT are far more likely to report it on CC than the half who score no better, or do worse on the ACT. The former group will be exuberant, and/or will want to hear confirmation that their ACT score is as equally valid as a comparable SAT score. Those who did the same or worse on the ACT are less likely to want to advertise that fact on CC. Don’t mistake that communications skew for a systematic pattern of test results.</p>
<p>^Very good analysis. </p>
<p>However, if we have a net skew, however small, it will be exploited until we have parity. Shifting ACT or SAT demographics can cause such a skew despite a 1.4 million base.</p>
<p>I’ll jump in here and say that I am in one of the states that requires the ACT for graduation: TN. We live in a top district and the administration is pushing for an average ACT of 24 within the next few years. While there are a few kids who will take the SATs here (for NMSF purposes, or because they are high achievers who are applying out of the area) most just stick with the ACTs. I’ve heard that in some rural areas of the state it is very hard to even take the SATs because they just aren’t offered. </p>
<p>bclintonk: great post :)</p>
<p>I think this is a regional thing that is starting to dissipate. We are in the mid-Atlantic. Students in the 10th/11th grades automatically take the PSATs. The format is therefore more familiar to them, although not necessarily the one that is best suited to their testing style. They take the SAT by ‘default’, only opting for the ACT (often with little or no preparation) simply to see if they do well enough to warrant another go. </p>
<p>Several of the local private schools are now giving the pre-act, PLAN. Based on early assessments in both formats, parents and students are choosing early to place initial energies to prepare for either the SATs or ACTs.</p>
<p>Our son found that statistically he did equally well on the ACTs (as has been suggested in previous posts), however opted to prep and retest a second time on the SAT format.</p>
<p>Our D did “better” on the ACT. How do we know she did better? By the concordance tables, which were developed as bclintonk indicated. If “most” kids did better on the ACT, then the concordance tables would shift.</p>
<p>I am a big advocate of kids taking the SAT and ACT each once, with minimal prep - just enough to understand the format and the rules. Then they can focus on prepping to re-take whichever one they did “better” on.</p>
<p>No significant difference for my kid, but she didn’t prep for the ACT other than looking over a couple of web sites the night before.</p>
<p>My daughter did much better on the SAT. She got a 20 on the ACT, but an 1140 (M/CR) 1800 total on the SAT. Not everyone does better on the ACT.</p>
<p>Mine did about the same but with zero prep for the SAT and only took SAT plus the PSAT. For ACT took the PLAN and ACT two times (before taking the SAT). He had to drive a very long way to take ths SAT. This was a number of years ago and he was interested in a couple CA schools that at the time required SAT. Number two didn’t bother with the SAT and didn’t need to because was not at cut-off on PSAT and no longer required or wasn’t at any schools he was looking at. Nice synopsis bclintock.</p>
<p>I also agree that the concordance tables that I believe were developed in collaboration (between ACT and CB) would shift if indicated.</p>
<p>According to the concordance table, my D had the same success on ACT and SAT the first time, both unprepped. (compared to ACT, the SAT score was skewed with perfect CR and lower math) She liked the ACT better, so she studied from the book, took it again and ended up with a 35. No need to see if she could match that with a comparable SAT score, so I’m not sure if this is a helpful data point. Oh, and we live where very few take the ACT.</p>
<p>The ACT and SAT use somewhat different approaches and these different approaches will favor different groups of students. The SAT, for example, allows more time per question but is less likely to ask you those questions in a straight-forward manner. If you’re a careful, reflective type, the SAT may play to your strengths. If you’re a kid who has trouble with simple questions asked in a convoluted way, then the ACT might be best. The math is far easier on the SAT, but the ACT is better if you’re not a quick writer, because your writing score is separate and does not affect your composite.</p>
<p>Statistically, it’s easier to get a generic 36 ACT than a 2400 SAT, but that’s largely because the ACT rounds up from 35.5, which gives one a lot of various ways to make 2-3 mistakes and still get the maximum composite. For example, you can make a single math mistake and still get a 36 ACT composite, but not a 2400 SAT. </p>
<p>I wonder, however, if ACT is taking steps to make it tougher to max out their composite. The number of 36’s has risen over the last few years from 220 to 300 to 400 to 600 (some of this is due to the large increase in test-takers to 1.5 million). I noted that when my younger son took the ACT this June, he had a ringer question in the math section that required one to know how to use synthetic division to pull the last two imaginary roots. Certainly, my older son never saw such questions, nor have I ever seen such a question in a test-prep book. I would not be surprised if ACT is playing with just one or two extra-hard questions on the total test to push “marginal” 36ers down to 35 and offset the rise that might be due to test-prep classes and repeated test-takers.</p>
<p>Timing of test taking can influence scores. Son got a 35 plus ? on the then new writing test on the ACT spring of junior year in HS, and I have forgotten his SAT I score because he had two 800’s on SAT IIs. He was supposed to study for and retake the Math SAT subject test he didn’t do as well on the fall of senior year but he signed for and retook the SAT I- got a 2400. Would he have gotten a perfect ACT score with repetition the fall of senior year? It wasn’t worth the effort. btw- raw scores do not have to be perfect for top reported scores- unless someone has documented new knowledge, look at the percentiles for the meaning of scores. from recent memory- an 800/800 on the general (not subject tests) GRE could be 99% or 94% depending on if it is the verbal or quantitative score (they do a 990 scoring for the Math subject GRE).</p>
<p>Here’s a tangential question. Would colleges be suspicious of a student from a heavily SAT dominant state, e.g. Massachusetts, turning in only ACT scores? I ask this because I remember reading about this in an article, but forgot where I read it.</p>
<p>
</p>
<p>Ancient history. Look at it from the colleges’ perspective: it is in THEIR best interest to show/publicize only the highest scores of applied/admitted/matriculated. Good for PR, good for alumni relations, good for bond agencies, etc.</p>