Inside the Testing Burrito

There is more than a bit of populist truth in the phrase, "Never watch your burrito being made," especially when it comes to bureaucracies and education (all flies and vermin). I see that the NYT has a prominent article on the new SAT test format which includes revamped reading and math sections and a much-discussed new essay section. All this hits home with your truly, since I used to supervise SAT tests at a local community college and currently work in an educational testing/assessment office. So in my down time today I cruised over to Collegeboard.com and looked around the research reports page. More commentary on this will come, but you might want to glance at the summary document "The Research Behind the New SAT" (RS-11, January 2005), especially near the end under the heading "Predictive Validity of the New SAT." You will find the following assessment:

While the scores on the new SAT writing section were slightly better than high school grades in predicting first-year college grades (.46 versus .43), the SAT writing scores were slightly worse than high school grades in predicting English Composition course grades (.32 versus .35).

In the literature, we say these r values have a low to moderate degree of correlation. I'm sure that will be of comfort to college applicants this coming spring.