This week’s seminar focuses on evaluating the academic performance of groups of Williams students.

Any SAT experts out there who can educate us about the impact of Score Choice?

Designed to reduce student stress and improve the test-day experience, Score Choice is a score-reporting feature that gives students the option to choose the SAT scores by test date and SAT Subject Test scores by individual test that they send to colleges, in accordance with each institution’s individual score-use practice. This allows students to put their best foot forward on test day by giving them more flexibility in score reporting. Score Choice is optional, and if students choose not to use it, all scores will be sent automatically.

Apparently, there are a bewildering menu of options (pdf) for how colleges can now use SAT scores. Williams (pdf) uses “Highest Section – Version 1”:

Williams College has indicated that it considers your highest section
scores across all SAT test dates that you submit. Feel free to send any
additional scores that you want to make available to Williams College.


1) Is that a reasonable choice? How does it compare to our peers? Looking at the list briefly, Amherst does the same but Yale requires that students send in all their scores.

2) The whole Score Choice scheme seems extremely sleazy to me. (Contrary opinions welcome.) ETS loves it because it encourages multiple test dates, thereby maximizing revenue. Colleges love it because it allows them to report higher SAT averages for their incoming class. Coaches love it because they can just tell desirable recruits to take the SAT many, many times. Just one lucky sitting will allow the recruit to meet the college’s minimum cut-offs.

And note that the Williams option allows the College to pick the highest math and critical reasoning scores even if they are from different sittings of the test. So, if you get 700 CR and 600 M in May and then 600 CR and 700 M in October, then Williams will report you as a 700 + 700 = 1400 student even though you never got that score.

3) Isn’t it obvious the Score Choice benefits rich and well-advised students at the expense of lower income and poor-advised ones? I am not sure what the magnitude of that benefit is, but I don’t like it.

4) Williams could easily fight the system by telling applicants that we require them to submit all SAT scores and that we use the average. This minimizes the benefits of taking the test multiple times.

What does all this have to do with our seminar topic on the academic performance of groups of Williams students? Simple: In order to study academic performance, we (sometimes) need to control for things like SAT scores and high school grades. To the extent that some categories of students (rich kids from prep schools) make extensive use of Score Choice and other categories don’t, our whole analysis is (potentially) screwed up! Very annoying.

Again, this is a topic that I don’t know much about. Informed commentary welcome!

Print  •  Email