<p>Yes, Colgate (on their website) includes the top scorers (by using accepted students) that actually go to Dartmouth and Williams instead (not that they know that yet when they’re applying) and Bowdoin doesn’t include its lowest scorers by making the SAT optional and completely distorting the median. </p>
<p>Middlebury has used this tactic too, although now at least includes the avg SAT score for the entire class post facto, which when I saw a year or two ago was 1349 (the exact average as Colgate that year). I’m not convinced, based on the range, that Midd’s 25-75% also includes the entire class. </p>
<p>At least sites like princetonreview, collegeboard, collegedata, etc report stats for the enrolled class at Colgate, although will likely always be 1-2 years behind the most recent class. I think it would be fine for Colgate to report accepted students averages (who is invited to the party) as long as they also report enrolled students averages (who actually attends the party). I’ve seen Bucknell report both simultaneously and thinks it’s a fair representation and gives candidates the most information of what it takes to be accepted and who actually attends.</p>