Questioning US news SAT data for H and P

<p>The latest ranking states that the middle 50% of SAT ranges at H & P include 1590. This would seem to imply that 25% of their students scored 1600. This is not only inconsistent with the common data set published by Princeton, but a mathematical impossibility given the size of their combined freshman classes, and the comparitively tiny number of students nationwide that the College Board indicates scored 1600. Can anyone shed any light on this seeming inconsistency. </p>

<p>This data if even vaguely accurate would also tend to cast doubt on the generally held CC belief that schools have little interest in the difference between a 1540 and a 1600.<br>
Is this a case of statistics being presented in a misleading way or am I missing something?</p>

<p>Hell yeah they are misleading. They add up the 75th percentile of M and Cr to get that number. So if the 75th percentile in math is 800 and 75th percentile in Cr is 780, USNEWS puts the 75th percentile as 1580. However, often the highest math scorer is not the highest CR scorer and vice-versa. If you want to get into H and P you should be able to understand this. :)</p>

<p>It never occurred to me that these institutions and/or USNWR would stoop to presenting these statisitics in such an outrageously dishonest and deliberately misleading fashion. This method of statistical analysis and presentation would be worthy of a D- at the local community college.</p>

<p>


</p>

<p>FredFred's statement is correct. The error is in how U.S. News (and some other organizations) aggregate and report data gathered by the standardized methodology of the Common Data Set. By the principles of good practice of the National Association for College Admission Counselors, colleges are supposed to report only interquartile ranges, NOT medians, for scores on tests that they use for admission decisions. And scores on the SAT Reasoning Test are often reported by colleges only as interquartile ranges for each section considered on that test (and for the ACT, interquartile ranges for the composite score). To see exactly how Harvard and Princeton report their Common Data Set data, check the College Board College QuickFinder pages on the Web </p>

<p><a href="http://collegesearch.collegeboard.com/search/CollegeDetail.jsp?collegeId=4221&profileId=6%5B/url%5D"&gt;http://collegesearch.collegeboard.com/search/CollegeDetail.jsp?collegeId=4221&profileId=6&lt;/a> </p>

<p><a href="http://collegesearch.collegeboard.com/search/CollegeDetail.jsp?collegeId=1251&profileId=6%5B/url%5D"&gt;http://collegesearch.collegeboard.com/search/CollegeDetail.jsp?collegeId=1251&profileId=6&lt;/a> </p>

<p>As for how many high school students get really high scores on the SAT, that is reported by College Board. </p>

<p><a href="http://www.collegeboard.com/prod_downloads/highered/ra/sat/SAT_percentile_ranks_males_females_total_group_critical_reading.pdf%5B/url%5D"&gt;http://www.collegeboard.com/prod_downloads/highered/ra/sat/SAT_percentile_ranks_males_females_total_group_critical_reading.pdf&lt;/a> </p>

<p>reading </p>

<p><a href="http://www.collegeboard.com/prod_downloads/highered/ra/sat/SAT_percentile_ranks_males_females_total_group_critical_reading.pdf%5B/url%5D"&gt;http://www.collegeboard.com/prod_downloads/highered/ra/sat/SAT_percentile_ranks_males_females_total_group_critical_reading.pdf&lt;/a> </p>

<p>math </p>

<p><a href="http://www.collegeboard.com/prod_downloads/highered/ra/sat/SAT_percentile_ranks_males_females_total_group_mathematics.pdf%5B/url%5D"&gt;http://www.collegeboard.com/prod_downloads/highered/ra/sat/SAT_percentile_ranks_males_females_total_group_mathematics.pdf&lt;/a> </p>

<p>writing </p>

<p><a href="http://www.collegeboard.com/prod_downloads/highered/ra/sat/SAT_percentile_ranks_males_females_total_group_writing.pdf%5B/url%5D"&gt;http://www.collegeboard.com/prod_downloads/highered/ra/sat/SAT_percentile_ranks_males_females_total_group_writing.pdf&lt;/a> </p>

<p>So, yes, U.S. News makes a mistake in how it reports composite scores when those aren't reported by colleges (as they usually aren't). But there is no hiding the fact that most admitted students at Princeton and Harvard have very high admission test scores. </p>

<p>By the way, the statement </p>

<p>


</p>

<p>is incorrect. Check the Common Data Set Initiative definitions </p>

<p><a href="http://www.commondataset.org/%5B/url%5D"&gt;http://www.commondataset.org/&lt;/a> </p>

<p>to see why.</p>

<p>Good for you for picking up on this, holycow.
Still amazing scores. For example, at least 25% of Princeton students in the class of 2010 scored 800 on the Critical Reading section. If perhaps 50% of students are legacies, recruited athletes, URMs, faculty kids, celebrity and political offspring, development cases, etc., it gives you an idea of what it takes to get in on shear brain power.</p>