<p>I’ve looked at this one pretty hard, and have a few observations/theories…</p>
<p>1st, I agree with the observation that a school’s mid 50% ACT is usually lower than the 50% SAT, using the concordance tables for conversion…especially the lower end seems out of wack.</p>
<ul>
<li><p>reported SAT is superscored, whereas ACT is not. Therefore, when converting ACT to SAT, I usually add 30 to 40 points to the SAT concordance. [Concordance study done by comparing ONE ACT sitting to ONE SAT sitting.]</p></li>
<li><p>as mentioned above, SAT is reported & part of USNews ranking, ACT is not (generally)…cynical view is that schools use only ACT scores for some of their (lower scoring) hooked candidates, and therefore do not include SAT scores for these folks in their reported numbers. [Check out Brown’s 25% ACT score…28!]</p></li>
<li><p>finally, all of the comparison between SAT & ACT is done with those pesky concordance tables. I find it interesting that if one plays around with the U Cal model inputing various SAT and ACT scenarios (remember, U Cal does not superscore SAT), the SAT equivalent of any particular ACT score always seems to be roughly 30-50 pts higher than the concordance tables would have you believe.</p></li>
</ul>