<p>Whistle Pig You wrote:</p>
<p>"Just some clarification: There is no significant correlation between the standardized test scores, etc. required for admission and potential for successful academic performance..."</p>
<p>I have asked you to substantiate this statement but you did not choose to do so. Yet you challenge me to disprove your assertion... this is not fair. </p>
<p>I have done about 30 minutes of research and I cannot validate your assertion. Your best case perhaps comes from fairtest.org. And even they say that the SAT is weakly correlated with success. (A correlation coefficient of r=.22, or r=.27 depending on the study) The do not use the term insignificant. </p>
<p>In spite of jokes about "Military Intelligence" the military is not stupid. They know what they are doing. According to the book: "The Naval Academy Candidate Handbook: How to Prepare, How to Get In, How to Survive Second Edition". The Naval Academy uses a scoring system called the "Whole Person Multiple." The most important selection factor is Class Rank, next most important is Math SAT score, then the Verbal SAT Score. </p>
<p>Q: If SAT scores are not valuable in predicting success why does it rate so highly in the whole person scoring system? If you think the Naval Academy is going by gut feel you would be wrong. The Academies periodically reexamine the makeup and weight given to each component of the scoring system with all the military academies in the country there is no shortage of students willing to do thesis papers on this subject and others. I submit that the Academies would not be using this if their internal studies did not validate it. And they will stop using it when studies determine that it is prudent to do so.</p>
<p>Here is a link to a bunch of published studies that shows the type of things that have been studied. It is probably a small fraction of what it actually out there. I only mention this to show that the military is not bashful about studying itself. <a href="http://www.stormingmedia.us/search.html?q=naval+academy+sat&search.x=0&search.y=0%5B/url%5D">http://www.stormingmedia.us/search.html?q=naval+academy+sat&search.x=0&search.y=0</a></p>
<p>One interesting study I skimmed said that when a school admits a broad sprectrum of students (broad spectrum from an SAT score perspective) the SAT is a better predictor of success. This kind of makes sense, if you are MIT and you only take people who score 1500 on the SAT or better, then SAT would not be a very good predictor (because there will always be people who finish last), but if you are a school that admits all types of scorers: students who score highly and lowly and everthing in between the test is a better predictor at those schools. </p>
<p>One final note. I worked at a company that used multiple linear regression to study what factors determine who is more likely to renew their magazine subscription. We would get horrible regression scores among our test groups. Scores of r=.04 vs. r=.02 to a particular trait. I was taught in school that scores such as this were basically useless as predictors, yet experienced proved time and time again that we would make more money if we sent renewal offers to the popluation that scored r=.04 vs. the r=.02 population.</p>
<p>One more final note. Since the Naval Academy uses Class rank as it's number 1 metric in it's Whole person multiple it can provide an effective counter weight to the SAT for those that do not do well on the standardized tests. Every college that uses SAT in admission agrees to a set of standards put out by the collegeboard that states in part they they will not use SAT the sole determination in admissions. I think the SAT score has a place in the admissions process.</p>