<html><head><meta http-equiv="content-type" content="text/html; charset=utf-8"></head><body dir="auto">On May 30, 2019, at 3:47 PM, William Flynn Wallace <<a href="mailto:foozler83@gmail.com">foozler83@gmail.com</a>> wrote:<br><div dir="ltr"><br></div><blockquote type="cite"><div dir="ltr"><div dir="ltr"><div class="gmail_default" style="font-family:comic sans ms,sans-serif;font-size:small;color:#000000">You may have read of graduate schools dropping the use of the Graduate Record Exam as a part of admission.</div><div class="gmail_default" style="font-family:comic sans ms,sans-serif;font-size:small;color:#000000"><br></div><div class="gmail_default" style="font-family:comic sans ms,sans-serif;font-size:small;color:#000000">While it is true that the correlations between GRE performance and graduate school performance are small (recent Science article), it does not mean that what is measure by the GRE is irrelevant to grad school work.</div><div class="gmail_default" style="font-family:comic sans ms,sans-serif;font-size:small;color:#000000"><br></div><div class="gmail_default" style="font-family:comic sans ms,sans-serif;font-size:small;color:#000000">I think was is going on, and always has gone on, is that you are dealing with a restriction in range. Imagine if you were to give the GRE to a random sample of people and then measure their ability to do grad school work. The correlation would be very high. But people wanting to go to grad school are mostly pretty smart - restricted range.</div><div class="gmail_default" style="font-family:comic sans ms,sans-serif;font-size:small;color:#000000"><br></div><div class="gmail_default" style="font-family:comic sans ms,sans-serif;font-size:small;color:#000000">No correlation does not mean no importance of the traits being measured.</div><div class="gmail_default" style="font-family:comic sans ms,sans-serif;font-size:small;color:#000000"><br></div><div class="gmail_default" style="font-family:comic sans ms,sans-serif;font-size:small;color:#000000">Give the restricted range, I cannot think that any test, present or future, would be a good predictor. The tests are not flawed; the samples are.</div></div></div></blockquote><br><div>I was going to ask, but then your last paragraph answered my question — on whether you think there are other tests that better predict grad school performance.</div></body></html>