The SAT Optional Difference
by Laurence Bunin - November 07, 2008
America’s institutions of higher education are on the leading edge of forging the next generation of professionals who will compete in a world unlike that of our parents. The students of today will be competing on a global stage that demands rigor of thought, creativity and performance. In order for them to succeed, we must maintain high educational standards and give students every opportunity to demonstrate their potential. Colleges and universities need every tool at their disposal in order to meet these challenges, and the SAT has a clear role to play in this effort.
College admissions offices face increasingly difficult and complex challenges. Charged with enrolling a highly qualified and diverse class of students, they must walk a fine line, balancing their available human resources with the review and analysis of a growing bulk of student data in an effort to serve the best interests of their respective institutions and students.
The diversity of these institutions is the result of a broad array of mechanisms through which admissions offices accomplish these goals. Some colleges rely on more holistic processes that place greater emphasis on subjective aspects of a students secondary school career. Others lean toward more empirical measures that lend themselves to greater head-to-head comparisons among students applying for seats at any given college. Both systems have their merits and their limitations.
In recent years a small percentage of selective and highly selective institutions have gravitated toward an SAT Optional model, a process in which the submission of SAT scores is left to the discretion of the applicant and is not a formal requirement for admission. Those supporting such a model often cite the virtues of recognizing students and their collegiate potential in terms predicated on factors not related to standardized testing.
As recently as September 2008, the National Association for College Admissions Counseling (NACAC) recommended that colleges give greater consideration to the SAT Optional model in their admissions process, urging them to do so if it is determined that the predictive utility of the test or the admission policies of the institution (such as open access) support that decision and if the institution believes that standardized test results would not be necessary for other reasons such as course placement, advising, or research. (National Association for College Admission Counseling, 2008, p. 7)
Whether going SAT optional is a good thing is a question best answered by the institution, based on its individual mission, goals and research. But it also begs another question: does an SAT Optional policy result in a higher level of academic performance and achievement? Research from Bates College in Lewiston, Maine, which adopted an SAT test optional policy in 1984, did not show significant differences in performance between SAT submitters and non-submitters (20-year Bates College Study, 2004).
But an independent analysis focused on Bowdoin College in Brunswick, Maine shows that the findings for Bates College dont hold true for all schools. Dr. Howard Wainer, professor of statistics at the University of Pennsylvanias Wharton School, examined student data from Bowdoin College, which was among the first of the more selective schools in the nation to adopt an SAT Optional policy in 1969. It accepts approximately 400 freshmen students each year, and, by rough estimates, about three quarters of these students choose to submit their SAT scores with the remaining quarter choosing not to submit them.
During the year included in Dr. Wainers research, the entering class of 1999, there were 273 students enrolled at Bowdoin who submitted their SAT scores with 106 deciding otherwise. A review of first year grades revealed a grade point average of slightly better than 3.1 among students submitting SAT scores, but a GPA of less than 2.9 for students who did not submit them.
The difference in first year grades amounted to about one standard deviation worse for students not submitting SAT scores compared to those who did, and Wainer concluded, If the admissions office were using other variables to make up for the missing SAT scores, those variables did not contain enough information to prevent them from admitting a class that was academically inferior to the rest. (Wainer, 2008)
But the relationship between first year grades and whether a student chose to submit an SAT score is only partially revealing. Wainer then examined the SAT scores of students who did submit them with the scores of students choosing to not submit their scores. The Bowdoin College data showed that the mean score for students not submitting their SAT results was more than nine percent lower than the scores of students who did submit their SAT results in applying for admission.
While Dr. Wainers examination of the Bowdoin data is not necessarily definitive -- other studies on SAT optional schools exist -- it is highly instructive. Dr. Wainers research not only supports the predictive value of the SAT but also the efficacy of having more information available to admissions officers rather than less. It also raises a legitimate question about the utility of the SAT Optional model in terms of determining those students with the best potential for academic success in the rigorous environs of a classical four-year collegiate career. As Dr. Wainer observed, Not having this information does not improve the academic performance of Bowdoins entering class on the contrary, it diminishes it. (Wainer, 2008)
The data also reveal information that reflects on the institution itself and how it may be perceived within the academic community and by those preparing to enter or invest in it. Dr. Wainers analysis shows that, among Bowdoin students declaring their SAT scores in the admissions process, the mean score was 1,323 on the old SAT scale of 1,600. However, the mean for students not submitting their SAT scores was 1,201.
SAT scores are also one of the many factors looked at in ranking the quality of a school and the education it provides. Although Dr.Wainer clearly notes that Bowdoin began using an SAT Optional model before US News & World Report began its widely circulated rankings of colleges, he observes that, a school can game its rankings by allowing their lowest scoring students to not be included in the average. (Wainer, 2008)
The aforementioned NACAC report rejects the notion of using SAT scores in any manner for which they were not intended. Just like any test, it has its limitations in what it can measure. For more than a century, the SAT has provided a fair standard by which students can be broadly evaluated for their collegiate readiness. Over time, SAT scores have been misused in myriad ways never foreseenas false indicators of an institutions financial health, as income-generators for short-term, commercial test preparation companies, and as determining factors of employee eligibility. Indeed, the September 2008 NACAC report, encourages US News and World Report to eliminate test scores as a measure of institutional quality, and follows up on its recommendation by observing that using test scores in such a manner creates undue pressure on admission offices to pursue increasingly high test scores. (National Association for College Admission Counseling, 2008, p. 9)
Nothing here is intended to suggest that the SAT is the end-all be-all in college admissions. As public schools scramble to meet government imposed educational standards, SAT scores are valuable in combating grade inflation and provide an empirical measure to blend in with other important student data, including high school GPA and class ranking, extracurricular and community involvement, student essays, socio-economic circumstances and many other factors that are taken into consideration during the admissions process. Ultimately, the best predictor of college freshman year performance is the combination of the SAT and high school grades; individually they are both strong predictors, but because they measure slightly different aspects of a student's achievement, research shows that together they are the strongest predictor.
Similarly, different colleges have different imperatives, and the volume and variety of data considered in the application process must match a given schools admissions standards. While Bowdoins thoughtful and holistic approach to admissions has its merits, the data from Bowdoin do provide a glimpse into some possible impacts of SAT Optional admissions. In this example, the predictive value of the SAT is clearly demonstrated, particularly when compared with other admissions criteria not associated with standardized testing. More to the point, overall academic performance declines, a prospect that should give pause to all educators.
Americas institutions of higher education are on the leading edge of forging the next generation of professionals who will compete in a world unlike that of our parents. The students of today will be competing on a global stage that demands rigor of thought, creativity and performance. In order for them to succeed, we must maintain high educational standards and give students every opportunity to demonstrate their potential. Colleges and universities need every tool at their disposal in order to meet these challenges, and the SAT has a clear role to play in this effort.
National Association for College Admissions (2008, September). Report of the Commission on the Use of Standardized Tests in Undergraduate Admission. Arlington, VA: Author.
20-year Bates College study of optional SATs finds no difference. (2004, October). Bates Now. Retrieved October 24, 2008 from http://www.bates.edu/x58748.xml.
Wainer, H. (2008, October 14). Dont ask, dont tell: The new rule of the SAT and college admissions. Message posted to Criterias Employee Testing Blog, archived at http://blog.criteriacorp.com/blog/ bid/6927/Don-t-Ask-Don-t-Tell-The-New-Rules-of-the-SAT-and-College-Admissions