Data errors could mean ‘slight change’ in GWU ranking

The chief analyst behind the influential U.S. News & World Report college rankings says the revelation that George Washington University inflated a key measure of the academic credentials of incoming freshmen could, at most, yield “a slight change” in the school’s ranking.

GWU ranks 51st on the most recent U.S. News list of national universities. On Thursday, the university disclosed that it had overstated the percentage of freshmen in fall 2011 who were in the top 10 percent of their high school classes.

Instead of 78 percent, as GWU first reported for that measure, university officials said the share was actually 58 percent. That is one of many pieces of data fed into the U.S. News ratings formula.

“If it does change GWU’s current Best Colleges 2013 edition ranking . . . it would be a slight change,” Robert Morse, data research director for U.S. News, wrote in an e-mail Friday afternoon. “However, we are still trying to carefully make that actual determination.”

The GWU disclosure brought renewed attention to questions about whether universities are providing the public with accurate admissions data in an era when rankings and guidebooks based on that data have gained remarkable influence with college-bound students.

Earlier this year, Claremont ­McKenna College in California and Emory University in Georgia acknowledged that school officials had inflated SAT scores of incoming freshmen in key reports.

GWU said an auditor has concluded that its SAT score reports are sound. But the D.C. university acknowledged that for more than a decade, it has reported class rank information in a way that inflated the academic credentials of incoming students.

The flaw is related to a fundamental problem with the statistic: A growing number of high schools don’t report class rank information to universities. GWU, for instance, bases the corrected 58 percent figure on data obtained for 38 percent of its 2011 freshmen.

For the other 62 percent, there was no class rank data.

The flaw in GWU’s method, officials said, was that the university estimated how many of those students were likely to have been in the top 10 percent of their classes. Then, they said, it mistakenly combined those estimates with documented class rank figures.

It is unclear exactly how much GWU’s class rank data was inflated over the years as a result of the flawed procedure.

On Friday, GWU provided a 15-year summary of how it reported the share of incoming freshmen in the top decile of their high school classes.

For its 1997 freshmen, GWU reported the share as 45 percent, a percentage that remained stable for a few years. For 2002 freshmen, the share jumped to 61 percent. For 2006, it was 65 percent. For 2010, it was 74 percent.

University officials said they have not calculated what the actual percentage should have been during those years. But they said that they are strengthening administrative oversight of data collection and reporting and that the information will be routinely audited in coming years.

Steven Knapp, president of GWU since August 2007, said Thursday that he deeply regrets “this error” and that “corrective action has been taken.” Officials said Knapp was unavailable for comment Friday.

For most of the past decade, GWU has ranked in the low 50s in the U.S. News list. In 2008, it ranked 54th. It inched up to 50th in 2011 before falling back a notch this year.

Critics of the U.S. News rankings say universities should not give the list credibility — a point they want to reinforce after this year’s disclosures about Claremont McKenna, Emory and GWU.

“I’m hoping this set of incidents will shed some light that these rankings don’t accurately measure what parents should care most about, which is the intellectual quality of the program and the level of student interest and engagement,” said John R. Kroger, president of Reed College in Oregon.

Reed College for years has refused to supply information for the U.S. News rankings. Nonetheless, it is ranked 75th this year on the U.S. News list of national liberal arts colleges.

Asked about the U.S. News rankings, Forrest Maltzman, senior vice provost for academic affairs at GWU since June 2011, said, “Rankings are one indicator among many that help parents and prospective students decide which higher education institution is right for them.”

He added that the university had no evidence that anyone involved in data reporting at GWU understood how much the methodology had distorted the percentage of students reported to be in the top 10 percent of their high school classes. “And we are not even sure what if any impact this had on our ranking,” Maltzman said.

 
Read what others are saying