<p>I am late to this discussion but I have to point out a glaring inaccuracy on the first page. calmom stated:
</p>
<p>In fact, what the authors actualy stated was the following:
That is, they only say there is “no similar threshold.” Indeed, they go on to say this:
So, yes, some low-SAT students do well in other fields, but there are actually “somewhat higher correlations” in those other fields. </p>
<p>The authors also refer to their earlier paper [Data</a> Mining the University: College GPA Predictions from SAT Scores by Stephen Hsu, James Schombert :: SSRN](<a href=“http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1589792]Data”>http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1589792). In it, they directly address the predictive power of SAT scores:
<br>
There goes the argument for dispensing with the SAT entirely. </p>
<p>Finally, it is humorous to see the vehemence with which people argue that high school GPA is the best predictor of college GPA. (Next you will tell us that SAT math scores are the best predictors of GRE math scores. That is not much of a revelation.) It is a little like saying that today’s high temperature is the best single predictor of tomorrow’s high tremperature. While true, using that single instrument exclusively causes you to miss the events we care most about–the big changes in temperature.</p>