New SAT percentiles

SAT and ACT Tests & Test Preparation
#1

<p>It is interesting that the percentiles shown on score reports for the most recent SAT test administration were calculated using last-year's data (from the old 1600-based test). In other words, if you received a score of 700 on this year's math test, you were given a percentile of 93%. But in fact, you scored higher than 93% of people who received that score on the old math test, not the new one.</p>

<p>It seems to me that with a harder math section, a modified critical reading section, and the addition of a writing section (fatigue factor), scores can't really be compared across tests. It surprises me that those percentiles were even included, since they may or may not be accurate. When they eventually collect enough data from the new test, percentiles may shift. Students who are disappointed in their scores now may find that they actually performed better (percentile-wise) than they had been led to believe. Of course, they could also shift in the opposite direction.</p>

<p>Any thoughts on this? Am I wrong -- have the new math and CR been statistically equated with the old sections so that the old percentiles still apply?</p>

#2

<p>Fatigue won't play as much of a role as everybody says. The new math section only includes a few minor things the old one didn't, and in any case it's usually geomtry and small mistakes that trip people up, not algebra II stuff. In the case of CR I think the average score went up if anything, because analogies were a section you either knew or you didn't. The percentiles are probably close enough to be accurate estimations, as the test hasn't changed <em>that</em> much.</p>

#3

<p>College Board made sure that the new Critical Reading and Math scores can be compared directly to the old SAT counterparts, meaning that the percentiles should not shift very much at all. Additionally, in the development phase of the exam, the testmakers tested the significance of fatigue in negatively affecting the scores, and found that fatigue does not appreciably affect scores until testing time exceeds about 5 hours (which is the case for exams such as the MCAT -- probably one reason why perfect scores are never achieved in practice there). Even though 3 hours and 45 minutes is a long time, most students will get used to this longer testing time.</p>