<p>lacgrad and EveVeee make good points. Against my better judgment I will make a couple more points about selectivity. If a given school’s acceptance rate drops, that may or may not mean that it is more difficult to get accepted. Imagine that a school embarks on a campaign to increase applications. If literally all of the additional applicants are uncompetitive, the set of matriculants will not change. So, one will see a drop in the acceptance rate but no change in the SAT interquartile range. I would argue that the school has not become more selective despite the drop in the acceptance rate. By contrast, if the campaign leads to an increase in the number of competitive applicants, the interquartile range will shift. I would then argue that the school has become more selective. The point is that the interquartile range gives additional information with which to interpret changes in the acceptance rate for a given school. </p>
<p>Now consider comparisons between schools. Imagine that school A has an acceptance rate of 18% while school B has 17%. Looking at SAT scores, school A has an interquartile range of 1350-1550 while school B has a range of 1190-1370. I would bet that the applicant pools are very different and, as a consequence, the typical applicant would find it much harder to get into school A. When schools are in different “markets,” SAT scores may be useful to get a sense of relative selectivity. My conclusion is that no single statistic should be taken too seriously. </p>