<p>
</p>
<p>When discussions of these nature appear, the same annual website data and academic studies – particularly the Revealed Preferences Rankings and Espenshade’s regression work – will be repetitively cited because they are, to my knowledge, the only ones that exist that are freely accessible to the public. </p>
<p>
</p>
<p>Hunt had the most accurate conception of why very high scores are retaken in #105. High retakes aren’t necessarily reflections of intellectual vanity, insecurity, perceptions of deficiency in other portions of the application, “having no life,” or whatever else. </p>
<p>It is a fact that higher scores at the very top scoring echelons hold greater admissions potency at the most selective universities than scores that are lower. Data are not even remotely compatible with the notion of a pre-established cutoff at an arbitrary performance level followed by selection contingent on other variables. Forty points can (and often do) make a difference. To state that even ten points cannot possibly effect admissions outcome is to misunderstand the basic purpose of score increments. If each score distinction denoted the same level of performance as between that and another increment, ad infinitum, every score along the continuum would be functionally “equal.” In terms of selective admissions observation, this is often correct when tracing admissions effects at “mid-range” test-score values (for a clearer conception, visit page eight of the RPR (page seven if excluding the title page), which was URL-linked in #91). But it does not hold true at the high end of the scale when admissions effects follow a rapidly escalating sequential or, of course, when making basic performance distinctions. Whether one personally believes that certain scaled disparities represent differences in terms of aptitude, achievement, academic readiness, or whatever metrical value the SAT claims to be determining is irrelevant and shouldn’t be involved in fact-claims concerning selective admissions behavior.</p>