March 22, 2005

Attacking the PSAT

Not content with picking on the SAT, UC officials are now going after the PSAT as well:

...officials at the UC system have a new target: what they perceive to be the National Merit Scholarship Programís overdependence on the the SATís cousin, the Preliminary SAT. A forthcoming article in National Crosstalk reports that a faculty committee at the university has recommended that the systemís campuses stop awarding National Merit Scholarships, and that a longtime UC administrator at is asking the College Board to break its ties to the program.

At the core of the universityís objections is the belief that by using studentsí scores on the PSAT examination as a strict cutoff for whether they qualify as National Merit Semifinalists, the merit scholarship program discriminates against black, Hispanic and American Indian students and students from low-income families who, on average, score significantly lower on standardized tests than do their white, Asian American and more-privileged peers.

Classic misunderstanding of test bias. The results of using PSAT score as a cutoff don't agree with the politically-correct version of how UC officials think the world should work; thus, the test is allegedly unfair.

There is nothing wrong with asking, as some UC officials and former officials are, if the PSAT is a valid test in this situation. It's true that it was not designed (to my knowledge) to be used as a means to obtaining a scholarship. However, it measures basically the same skills as the SAT, and thus it is an early measure of students who should be on the college prep track, and who could most likely use the money. (Full disclosure: I was a National Merit Scholar.)

College Board officials could not be reached for comment on Sunday. But Wayne Camara, the boardís vice president for research and development, told Crosstalk that the process by which the National Merit Scholarship Program and some individual colleges winnow the 16,000 semifinalists down to the 8,200 students who actually receive National Merit Scholarships each year takes factors other than PSAT scores into account. ďThe practice that National Merit is following is very consistent with the requirement that they use multiple sources of information in making a high-stakes decision,Ē Camara told Crosstalk.

Camara also said in the interview that because the PSAT has been shown to be valid in predicting studentsí SAT scores, the SATís validity in predicting student performance extends to the PSAT.

It's nice of the Colloge Board to point that out, but the heart of the complaint here is that not enough of the under-represented groups get scholarships when the PSAT is used for selection. It's disappointing to see the test being attacked in this situation, rather than a school system in which minority students are overwhelmingly short-changed. We can certainly discuss predictive validity and cutoff ranges, but those attacking the test would do well to step back and ask themselves just why so few minority students (and poor students, allegedly) are able to score at the high percentiles required for scholarship selection.

Posted by kswygert at March 22, 2005 11:17 AM