Home >  Local News

Sunday, September 08, 2002

By Bill Schackner, Post-Gazette Staff Writer

The University of Pittsburgh points to the average SAT score of its incoming freshmen each year as a sign of the school's growing selectivity.

That's because year after year the score rises, sometimes dramatically.

But there's something about that statistic that Pitt doesn't advertise as loudly: Not every student is counted.

Pitt excludes test scores of students it classifies as "special access" -- athletes and students enrolled in three programs serving "underrepresented populations," including African-American and Hispanic students, on its main Oakland campus.

This year, the practice meant that Pitt excluded roughly one in 10 freshmen, or 305 students, and in years past it has meant as many as 448 students were excluded. By dropping them from the tally, Pitt in effect inflates its SAT average.

It's no secret that colleges competing for top students use admissions data to cast themselves in a favorable light. And experts say an unknown number of these schools across the nation have chosen to exclude scores from some lower-performing groups, ranging from athletes to international students, before calculating their SAT average.

Still, some Pitt faculty and other groups said they had no idea the region's largest public university had been excluding scores for decades.

Pitt's SAT average would have risen just as quickly over the last decade if all the scores had been counted, but the average score itself would have been lower than the school now reports.

Excluding certain groups gave Pitt a 1220 average this fall, 19 points higher than if all the students were counted. (A perfect score on the SAT is 1600.) In years past, the practice has raised Pitt's SAT average by as much as 33 points.

Pitt administrators said they believed that the practice, in use since the 1960s, was sound. It's not an attempt to deceive, they said, but rather an attempt to give guidance counselors, families and others a more realistic picture of the competitive environment applicants to Pitt will face.

"There is certainly no attempt here to consistently or systematically put out a number that doesn't represent a fair reality within the university," said Vice Provost Robert Pack.

"These groups are taken out, not because they have the lower scores," he said. "They are groups that are taken out because they are admitted on a set of criteria that places much less emphasis on prior levels of academic attainment."

Pack said he doubted anyone would choose a school based on 19 extra SAT points.

"I don't think we have done any of this to make the university look good," agreed Betsy Porter, director of admissions and financial aid.

Pitt officials declined to release year-by-year comparison figures for the SAT average it uses vs. an average for all incoming students. Officials also would not identify how many "special access" students come from each of the four programs: athletics, the University Challenge for Excellence Program, Engineering Impact and Engineering Prep.

Pack said more important is the fact that the number of students excluded from the SAT average has declined from 18 percent of the freshmen class a decade ago to 10 percent this year.

The National Association for College Admission Counseling, to which Pitt belongs, publishes a code of good practices that argues against omitting any freshmen scores. Member schools agree to "report first on all first-year admitted or enrolled students, or both, including special subgroups, e.g. athletes, non-native speakers," the guideline states.

The association knows that some schools prefer to do otherwise but ultimately concluded that for prospective students, "more information is better," said Martin Wilder, an association vice president who also is vice president for enrollment at Mary Washington College in Fredericksburg, Va.

As long as a school provides a score for all freshmen, it also can provide a number in which subgroups are broken out, the guideline states.

For generations of SAT-takers, the averages released by colleges have been a way to gauge the odds of being accepted. But the averages also have become a barometer of prestige for schools increasingly locked in a fight for top students. Pitt is no exception.

The school's SAT level, and how it has risen from 1139 since 1995, is among measures of academic ascent regularly mentioned by Chancellor Mark Nordenberg, who took office that year. School officials also regularly talk about the rising percentage of students eligible for Pitt's honors college and how student applications have doubled.

The SAT performance was even cited in 1999 as one reason Pitt trustees gave Nordenberg a $35,000 performance bonus that year.

"The image of greater selectivity appeals to many important constituencies -- your board of trustees, your alumni, your faculty, your prospective students and your local media," said Robert Schaeffer, an official with the National Center for Fair and Open Testing, an advocacy group that monitors standardized exam practices. "It becomes a self-fulfilling prophecy. It signals to prospective students, their parents and guidance counselors that this is an elite institution."

Schaeffer said Pitt's practice "is certainly an example of test score misuse. They are calculating their average in a way to create a misleading impression of the institution," he said.

A random sampling of other schools found almost none that say they exclude parts of their freshmen class when issuing test scores.

Virginia Tech, which competes against Pitt in the Big East, said its public releases about SAT performance are based on all freshmen, from international students to linemen on the football team. Officials at West Virginia University, Arizona State University, UCLA and the University of Washington -- all schools with major athletic programs -- insisted they do likewise.

Virginia Tech spokesman Larry Hincker said he has yet to hear a reason for excluding selected SAT scores that makes sense. He believes the public would be equally indignant with Virginia Tech no matter if it was athletes being omitted or disadvantaged students.

"If you are going to report an average SAT for your student body, then it should be for everyone," he said. "Once you start saying, 'Janie had a tough life, we're not going to count her score,' then you start down a slippery slope. Who else aren't you going to count?"

"I don't know why you'd want to risk the supposition that you're gaming other numbers," he said. "It isn't worth the attack on your credibility."

Ohio State University uses the ACT and reports scores of "all incoming freshmen -- athletes, underrepresented minorities, everything," spokeswoman Amy Murray said.

Penn State University said it includes athletes and other freshmen, with the exception of international students and adult learners who are not required to retake the exam upon entering.

Some widely read college rankings, including the one done by U.S. News & World Report, choose not to publish average SAT scores. U.S. News solicits scores for all freshmen and determines the range of scores for the middle 50 percent of the class.

Of the 1,450 campuses it approaches each year for data, about 10 percent to 15 percent leave out part of their freshmen scores. The magazine de-emphasizes the SAT result for those schools when deciding how the campus should be ranked.

Pitt officials said they provide U.S. News with scores for all their freshmen because the magazine requires them for its rankings.

Faculty members had varied reactions to Pitt's SAT reporting.

Math professor Tom Metzger said he's been at campus meetings on education policy and heard Porter make it clear that the figure does not include all incoming students, and Metzger said he does not object to that "as long as they're consistent."

But Jim Holland, who served three times as University Senate president and has taught at Pitt since the 1960s, said the exclusion "is total news to me. I assumed it was just a plain average. There's never been a qualification that I've heard."

Faculty member John Baker said he also was unaware of the practice, and feels that all freshmen, athletes included, should be counted if they took the test.

"I think it's more honest," he said.

Pack said Pitt officials believe a measurement of a "middle 50" range -- excluding the lowest quarter and highest quarter of all test scores -- is preferable to using an average, but he said Pitt has been unable to resist families, faculty members and others who wanted the simpler measure as a benchmark. "We caved in," he said.