<p>@Data10 You are correct. My point was that Bowdoin was in effect selecting students from the pool of non-submitters that had SAT scores close to those of SAT of submitters. If this is the case, then why is Bowdoin test optional, these same students may have been selected if they had submitted their test results because they were within the bottom 25th percentile of admit-tees? If this is the case, then what’s the benefit to the non-submitters who have high HSGPA, excel in other factors but have really low SAT scores relative to Bowdoin.</p>
<p>My guess is that going test optional means that it will receive more applicants from those with high HSGPA and substantially low test scores relative to the overall profile of Bowdoin students. These students have little or no chance of being admitted to Bowdoin, but Bowdoin welcomes these students to apply.</p>
<p>As to your statement " Test scores add relatively little to the prediction of academic success beyond other sections of the application" this is consistent to what has already been stated that after a certain level of achievement on the SAT, applicants are pretty similar in ability and SAT predictability has little value. </p>
<p>I highlight the correlation from the graph because if an applicant has a high HSGPA and high SAT score of that is 2 standard deviations from average those students are fully qualified to do well at any institution from the most prestigious to the least prestigious university. </p>
<p>As @2018RiceParent has mentioned there are many low SES students with high SAT that do not apply to top schools. Given the data from the graph, the SAT can be a leveling factor for many of these low SES students to get into top schools.</p>