Lou Hirsh, director of admissions at the University of Delaware, shares thoughts on standardized testing and the admissions process. This appeared in the October 5, 2008, issue of The News Journal.

Every year, the University of Delaware Admissions staff gives thousands of admissions presentations to audiences throughout Delaware and across our country. One question is asked more frequently than any other: "What are you looking for in the SAT?"

Lying just beneath the surface of that question is the fear that SAT (or ACT) scores will have a veto power over everything else that the student has accomplished, and at some colleges they probably do. The encouraging news is that at a growing number of colleges they play a less controlling and more complex role in admissions decisions.

I use the word "complex" deliberately. Human beings are complex - especially human beings who are also adolescents. Because test scores are numbers, they give the illusion of being precise measures of what a student is academically prepared to handle in college. In truth, they are far from precise, and we know it. To put it more bluntly, all tests are flawed and cannot come close to capturing the complexity of human accomplishments and intelligence.

If that troubles you, bear in mind that most medical tests are also imprecise and do not necessarily give valid results for all patients. They can point your doctor towards a diagnosis, but invariably a doctor will insist on additional tests and procedures to confirm that diagnosis. Not all patients are alike.

The analogy holds for what we do in college admissions: not all students are alike. For some students, the SAT and ACT are valid measures of their readiness for college, especially when combined with a review of their course selection and grades. For others, however, they can be misleading or give an incomplete picture of the student.

Our challenge is to use responsibly a tool that is flawed, but sometimes helpful. It is not unlike the challenges we face when we look at other aspects of a student's academic record.

For example, some who oppose testing argue that we should focus only on high school grades. In a typical year we draw applications from over 2,800 public and private high schools worldwide. There are huge differences in course content and grading standards among them. An "A" student from a weak high school may be less prepared than an "A" student from a competitive high school.

Some test optional schools require that students submit a graded essay in lieu of the SAT. But why would an essay written on a narrow subject be a more reliable gauge of what a student has learned in four years of high school than a standardized test score?

What we do know is this: if there is one piece of information that is consistently more valuable than any other, it is the rigor of a student's high school course selection. Succeeding in tough, academic classes is a good sign that the student is likely to succeed in college.

That's why we admit students to Delaware with modest SATs when they have taken tough classes and earned outstanding grades in them. Similarly, we deny admission to applicants with very high SATs when their grades are not commensurate with their test scores. Twice in my professional career I have signed letters of denial to students who had perfect 800's on their SATs; we denied them because their high school grades were dreadful.

What role, then, should the tests play? Let's begin by acknowledging that in the majority of cases, they neither help nor hurt our applicants: their course selections are so rigorous, their grades are so strong, their letters of recommendation are so supportive, and their admissions essays are so convincing that it is hard to see them as anything other than terrific additions to our freshman class. Many of them are Delawareans who fully meet the guidelines set forth in the UD Commitment to Delawareans.

There are, however, many applicants for whom the evidence in the file is ambiguous and contradictory, and that is why the majority of colleges still require standardized tests. Let me give you an example.

Several years ago we enrolled a student from an inner-city high school. His grades were a frustrating mix of A's in some courses combined with B's, C's, and even an occasional D in others. His junior and senior year grades were somewhat better than his freshman and sophomore grades. From his counselor we learned the he was raised in a single parent household with a family income that was below the poverty level.

What startled us when we reviewed his application was that he had a combined verbal and math SAT of 1230. That is more than 200 points above the national average for college-bound students. It is probable that no more than 1% of the students nationwide coming from his income level and school environment are able to score that well on a standardized test.

His SATs combined with his improving grades are what encouraged us to take a chance on this first generation college student - a chance that we would not otherwise have taken. After a shaky start at UD, he is now an upperclassman earning mostly B's and even a few A's.

In each applicant pool there are the late bloomers - kids who have only recently started getting good grades. They are disproportionately male, and the standardized tests help us understand whether they have mastered a sufficient amount of their high school course material that we can confidently admit them to UD.

There's another concern. Some high schools have grade inflation and a penchant for designating nearly every course as "honors." When we see a dozen applicants from the same high school, all with "A" averages, all taking "honors" courses, but all scoring well below our averages on the SAT or ACT, we have to conclude that their school is not as rigorous as it would like us to believe.

No single measure works for all students or all high schools. At the University of Delaware, we would no more be "test" optional than we would be "essay" optional or "transcript" optional or "letter-of-recommendation" optional. These are all pieces of information that shed light on our applicants. Individually, each piece is flawed and incomplete in what it tells us. Collectively, however, they tell us a great deal and help us understand what is distinctive about each of our applicants.