The historic campus ofWashington and Lee University in Lexington, Va. (Stephanie Gross/The Washington Post)

Washington and Lee University declared in its official reports last year that 5,972 students applied for admission and 19 percent were accepted. Those numbers helped define the public profile of one of the nation’s most-selective liberal arts schools.

They also were the result of a counting method that worked to benefit the university’s image.

Internal university data obtained by The Washington Post show that more than 1,100 applications for the class of 2016 at the private Virginia school — roughly one out of every six — were never completed. The files were missing required elements such as teacher recommendations or test scores, raising questions about how many of them were seriously considered for admission.

If the incomplete applications had been omitted from the official count, university dean of admissions Bill Hartog acknowledged, the admission rate would have been 24 percent, a 5-percentage-point swing in selectivity.

“No matter how you slice it, we’re among the most competitive schools in the country,” Hartog said. He denied that the university had intended to inflate its numbers. “We don’t even think about that stuff.”

What Washington and Lee does in computing its selectivity does not appear to break the rules for reporting data to the federal government or market analysts such as U.S. News & World Report. Some top schools count only completed applications, and others use Washington and Lee’s method.

But interviews with officials at several colleges suggest that the incomplete share of Washington and Lee’s applicant pool last year was unusually large. And the gap between the reported admission rate and what it would have been using a stricter counting method illuminates a growing debate about the reliability of data schools present about themselves.

“This is part of the great mystery of college admissions,” said Andrew P. Kelly, a higher education analyst at the American Enterprise Institute. “We don’t have much of an idea of how this works internally.” Kelly said pressure on colleges to stand out in a crowded market inevitably raises questions about self-reported admissions statistics.

“If counting them one way helps you, and counting them in the other way doesn’t,” he said, “there’s often an incentive to count it in the way that helps you.”

Hartog said the university’s counting method is common among its peers and follows federal rules. Washington and Lee makes every effort to get applicants to send in all required materials, he said, but the school sometimes admits students who don’t.

Conor Moran, 19, a student whom Washington and Lee apparently counted as an applicant, said he didn’t consider himself one because he never finished the steps required to apply.

Moran graduated from Chantilly High School in Fairfax County in 2012. The data The Post obtained indicated that he submitted an incomplete application to Washington and Lee, the kind of partial application that the university counted in its official total.

But Moran said he dropped his bid for Washington and Lee, instead pouring his energy into a successful application to George Mason University. He said he chose GMU for its tennis team, its financial aid and its scientific research programs. As Moran recalls, he did not send his teacher recommendations or ACT scores to Washington and Lee because he decided he didn’t want to go to the school in Lexington, Va.

“Definitely, I don’t consider myself an applicant at all,” Moran said. He chuckled when he learned that he was probably counted as one anyway.

“It’s statistics at its finest,” Moran said.

The federal standard — also used by U.S. News and others — is to count only those applicants who have fulfilled the requirements to be considered for admission and who have been notified of one of four actions: admission, non-admission, placement on a waiting list or the withdrawal of an application.

But there remains enough subjectivity in what is considered an “actionable” bid for admission that the very definition of an applicant is open to debate.

The University of Virginia, for example, takes a strict view.

“If we don’t have enough information to make an admission decision, meaning a transcript, recommendations, etc., we would not count the application in our numbers,” said Greg W. Roberts, U-Va.’s dean of admission. “We do not count or review applications that are incomplete.” The U-Va. admission rate for 2012 was 30 percent.

Harvey Mudd and Wellesley, two liberal arts colleges in the same elite company as Washington and Lee, also said they keep incomplete applications out of their official totals.

“There are all kinds of reasons why we want to be as precise and accurate as possible,” said Harvey Mudd’s director of admission, Peter Osgood. “There’s no need to candy-coat things or juice things up to make it more attractive.”

But officials with Vassar, Pomona, Bowdoin, Haverford and other prestigious colleges say a modest portion of incomplete applications are routinely included in their totals. Bowdoin spokesman Scott W. Hood said the number of incompletes was “very small” — 49 out of 7,052 — and did not affect the 14 percent admission rate for the Brunswick, Maine, school’s class entering this year.

“We count any application that’s been started by a prospective student, and we’ve been using that method for nearly two decades,” said Eric Sieger, a spokesman for Carleton College in Northfield, Minn.

Georgetown University’s dean of admissions, Charles Deacon, said incompletes can account for as much as 5 percent of the university’s typical reported application total of 20,000. Georgetown’s admission rate for 2012 was 17 percent.

In the past two years there has been a major movement to scrub the statistics colleges report about incoming students.

In April 2012, an investigation for the board of trustees at Claremont McKenna College found that for several years the California college had reported inaccurate data on SAT and ACT scores, high school class rank and application totals. A Claremont McKenna spokeswoman said the college now excludes incomplete applications from its count.

More revelations followed.

Emory University misreported test scores and class rank information. George Washington University misreported class rank statistics. Tulane University’s business school misreported test scores and application numbers for a master’s in business administration program. Bucknell University misreported SAT scores.

In May, Dominican University of California told U.S. News that it had overstated application totals, which led to a correction in its 2011 admission rate — to 71 percent, instead of 54 percent.

The U.S. Naval Academy, which has an unusually lengthy admissions process that includes a one-mile run and push-ups, in 2011 faced criticism for how it counts applicants. By including several thousand incomplete applications, the academy reported an extremely selective admission rate of about 7 percent.

Cmdr. John Schofield, an academy spokesman, said the academy this year modified its count, excluding certain Reserve Officers Training Corps candidates who did not indicate a specific interest in attending Annapolis. That helped shrink its applicant pool, to 17,819 from 20,601, and raised its admission rate to 8 percent.

Some colleges solicit applicants so aggressively that they waive all fees and fill out practically everything in an application form except a student’s signature. That practice, known as a “fast app,” also affects admission rates.

For U.S. News, the admission rate is a small factor in its annual college rankings. But it is displayed prominently on U.S. News Web pages, including sortable charts. Washington and Lee has the 13th-lowest admission rate among schools that U.S. News calls “national liberal arts colleges” in rankings published Sept. 10.

Washington and Lee, roughly 150 miles southwest of the District, is named for two famous Virginians. U.S. President George Washington helped endow the school in the 18th century, and Confederate Gen. Robert E. Lee was its president after the Civil War.

The 2,300-student university takes its honor system so seriously that students are allowed to take final exams on their own schedule, without any proctoring. Students found guilty of one instance of lying, cheating or stealing are kicked out.

Hartog, who has led the admissions office since 1978, said there is “no way” the school would fudge its admissions data.

“I’ve been observing this a long time,” Hartog said. “There are places that are making a mockery of the volume of applications they purport to have. In my view, we’re not. The last thing in the world I’m going to be doing is misleading anybody.”