<p>Comparing the equivalencies in percentiles versus the standard conversion table, a score of X in the ACT, compared to that score converted through percentiles to the SAT, is significantly higher than the score of the standard table conversions. The “lower” ACT scores (by standard table conversion) seem to be acknowledged by the colleges’ admissions.</p>
<p>So the colleges don’t use the standard table conversions, but instead look at percentile from what it looks like?</p>
<p>Good question, but you can’t compare by percentiles, because different populations take each test. The only way to do those concordance tables that makes any sense is to look at the population of kids that took BOTH tests and compare the scores.</p>
<p>For the schools you list, look at the percentages of kids who submitted SAT scores vs. ACT scores. Like at Pomona, 93% submitted SAT scores and 26% submitted ACT scores. That means 19% submitted both, and their highest score of the two tests was probably used for admission. The discrepancy in the ranges is more a reflection of the differences in the populations using and submitting either or both tests than admissions practices. Furthermore, those are the score ranges for the middle 50% of enrolled students. They’re merely descriptive of the freshman class, not statements of admission criteria. There are so many other factors that go into the admissions process; the kids with the lower ACT scores must’ve had other things going for them, which likely include, among other things, higher SAT scores. It’s also possible the kids with the higher ACT scores, or those with the lower SAT scores, decided to go somewhere else.</p>
<p>Mrs. Ferguson, just to clarify, are you saying that the conversion tables (not percentiles) from SAT to ACT (and vice versa) are pretty accurate? As in a 31 on ACT is similar to a 1380 SAT?
Thanks!</p>
<p>I really have no idea. If they are based on those who take both tests, like I think they are, then I’d think they’d be pretty accurate for THAT population. The sample is self-selected, so you have to ask why they’re taking both tests. Did they not do as well as they’d hoped on the first one? Things like that skew the data. But the colleges that develop these tables have a LOT of data at their disposal, and have been relying on that data for many years, so they must be fairly reliable. And if the school you’re applying to thinks two scores are equivalent, then for their intended purpose, they are.</p>
<p>Do colleges use these conversion charts? Because it’s sort of weird that some of them say a 31 = 2040, while some say a 31 = 2100-2150. That’s a big difference. Will colleges treat them out of 36 or would they convert it to SAT and look at it that way?</p>
<p>OP, your comparison is statistically invalid! You are comparing a composite ACT/whole SAT conversion chart to charts giving you percentiles for INDIVIDUAL sections of the SAT. Your math or reading section puts you in the 93% range, but if you are 93% in both, you are much higher than 93% when you combine them. This is because it is much harder to get to the 93%mark in two different test sections rather than just one. Most kids that hit 93% in math WILL NOT hit 93% in reading, and vice versa. The WHOLE TEST percentile comparison between the ACT and SAT will match up with the first conversion charts you link to.</p>
<p>the best thing to do is to check with the school your applying to. For example, i had an interview with U of Miami and the admission offcer told me the school only really care about the reading and math section of ACT. That way they can compare M+V for SAT to ACT</p>
<p>I, too, will also start doing some detective work as I am convinced that there is a pretty interesting story here that doesn’t get told out loud. Whether it is based upon the school’s tracking the admitted students from the past and correlating them over their four years, who knows? I do know that many kids get a boost with their internal correlation tables.</p>
<p>Madame B, thank you. Everyone start pumping those college people visitng your schools.</p>