Questioning the SAT Writing Test

K-12 Testing

The same day the College Board held its news conference to release details of the “new” SAT, its website posted a second bulletin: “SAT II Writing Scores Delayed.” The explanation, it seemed, is that some 20,000 additional students had taken the exam at the June, 2002, administration, overloading essay grading capacity.


The problem would be little more than a historical footnote if the College Board had not just trumpeted plans to increase essay scoring volume at least eight-fold by adding a “Writing” section to the SAT I battery. Currently the SAT II Writing Test is taken by about a quarter million students in each high school graduating class. Most are applying to the fewer than five dozen colleges which require SAT IIs. But more than two million SAT I exams are administered annually. This means that, beginning in March 2005, every major Saturday SAT I Writing Test administration will generate more essays than are now graded in an entire year.


The College Board confidently claims that it can handle the tremendous workload increase by setting up a national network of high school and college faculty graders. Handwritten essays will be scanned and transmitted to readers across the country. Each essay will be graded on a scale of 1 to 6 by two individuals. If scores differ by more than two points, a third reader will resolve the difference.


But neither the technology nor scoring network yet exists. Some observers question the company’s capacity to handle a task of this magnitude, especially since the launch of its website was months late and millions of dollars over budget (see Examiner, Summer 2000). Others are concerned that the difficulty of training thousands of new part-time workers to grade consistently will result in “subjectivity” than can significantly alter reported scores.


Beyond the logistical problems, academic critics fear that the short essay will encourage high schools to focus on formulaic writing. Already test preparation companies are gearing up to drill their high-paying customers on how to package five-to-seven paragraph “essays” that will receive high scores, based on their SAT II Writing Test coaching experience.


Civil rights organizations raise additional, fundamental issues. Will the added test create further impediments for groups that have historically been underrepresented in higher education? The correlation between SAT II Writing Test scores and first year grades for African Americans is only .31, lower than even the weak predictive value of the current SAT I. Can a brief essay drafted under tight time constraints in a high-pressure setting accurately assess the capacity to handle college-level written work for someone whose first language is not English? What about accommodations for students with handwriting disabilities or those accustomed to composing on a computer?


The answers to such questions will be the ultimate measure of the SAT II Writing Test.