Editorial Columns

Don't Throw Out the SAT with the Bath Water

If there is one thing we all learned from the last presidential election, it's that predicting outcomes is not an exact science. Predicting who might be successful in any venture, such as the quest for a college degree, would be a tough call without reliable, proven tools.

In the college admissions process, the SAT and its cousin, the ACT, are tools that have been relied on for decades as partial barometers of students' likelihood of success in college. When University of California President Richard Atkinson last week recommended in a speech that his nine-campus system drop its use of the SAT, his words elicited a gasp, and even some cheers, from within the American higher education community. The University of California is one of the nation's largest and most highly regarded public university systems. Thus, his speech will undoubtedly turn up the heat on a long-standing debate over the fairness of the use of standardized tests.

But students should not put their pencils away just yet. There are many, like me, who believe in the reliability and validity of such standardized tests, and I am troubled by President Atkinson's suggestion. Standardized tests have proven extremely useful in helping colleges compare students by providing common yardsticks across students from high schools where curricula, transcripts, and grade point averages can be highly variable.

There are two fundamental problems with standardized tests such as the SAT. First, such tests are often used for the wrong purposes, for example when institutions of higher education use the test to predict something the tests were not designed for in the first place. Second, such tests are often overweighted. In other words, too much emphasis is put on the results. The challenge, then, is to use the tests properly.We must not throw the baby out with the bath water.

The SAT and ACT, staples in the admissions process for more than 80 percent of four-year colleges and universities in this country, have been used for decades to aid in the selection of potential students. Last year, according to the College Board, which sponsors the SAT, a record number 2 million students across the nation took the test. But it is just one of the many variables that colleges and universities have at their disposal when making admissions decisions.

Every year, institutions receive applications from students who possess vastly different high school experiences that are not easily measured and compared. It is an imperfect science, but scores of research studies verify that the SAT adds to our ability to assess prospective students. Using the SAT as the primary criterion for college admission is not justified, but its judicious use in conjunction with the high school course selection, the rigors of the high school curriculum, high school grade point average or class rank, assessment of special talents, and other life experiences is quite valuable.

In the science of measurement, there are two key indicators of the accuracy of an assessment instrument like the SAT. These concepts are called "validity" and "reliability." In simple terms, validity refers to whether or not we are measuring what we intend to measure; is our measure really helping to predict college success or something entirely separate or irrelevant? Reliability refers to how well we are measuring what we intend to measure; is our measure consistent with each question contributing to a meaningful test score, does the test stand up over time, and do the different parts of the test work together properly to give the measure meaning.

There is no question that the SAT is both reliable and valid. In study after study, researchers have shown that a combination of SAT scores and other indicators provide a better prediction of early success in college than the use of the other indicators alone. This holds true for both males and females and for all racial and ethnic groups.

It is my opinion that the SAT still carries value for colleges and universities. Over the years, it has provided national standards that can be relied on. In turn, it has encouraged higher achievement. To abandon such tools simply robs us of one important variable in making increasingly difficult decisions about the readiness of our students for the rigors of our educational programs.