CHAPEL HILL, N.C. (AP) -- North Carolina's top academic officer said Friday that reading specialist Mary Willingham's findings on athlete literacy levels are flawed, though she stands firm that her research is ''100 percent correct.''
In a meeting of the school's faculty council, provost James W. Dean Jr. said there is ''absolutely no basis whatsoever in the dataset to make any claim about the literacy of our student-athletes.''
In a CNN story last week, Willingham said her research of 183 football or basketball players at UNC from 2004-12 found 60 percent reading at fourth- to eighth-grade levels and roughly 10 percent below a third-grade level. Dean said the findings were based largely on standardized scores in a 10-minute timed vocabulary test that isn't an appropriate way to measure literacy levels.
Dean called it ''flawed analysis'' and said it was unfair to use the data to say students can't read. Dean said the school will have outside consultants review their analysis of Willingham's data.
''I'm not playing some kind of trick with this data,'' Dean said. ''This isn't marginal. This is really quite central to the claims that have been made. And I take no joy in this. I think it's very sad, actually.''
The school has stopped Willingham's research until she receives approval from the school's Institutional Review Board (IRB), saying she can't continue to use data with information that could identify the subjects.
In an email to The Associated Press, Willingham said she is scheduled to go before the review board next week to go through the approval process.
''My data is 100 percent correct,'' Willingham said. ''In addition, I worked with the overwhelming majority of the students in the data set on reading and writing skills between 2004 and 2010.
''It's interesting that my IRB was pulled and I was told that I could not talk about it until it was resolved, meanwhile the provost is allowed to discuss the findings. That is what is truly erroneous about all of this - and at a research university - wow. At UNC we protect our brand at all costs.''
Faculty had a chance to ask questions of Dean, chancellor Carol Folt and admissions director Stephen Farmer during the council meeting after Dean's presentation.
Jay Smith, a history professor critical of how big-time sports affects UNC and other universities, defended Willingham's efforts during the session.
''She's spent 40- and 50-hour weeks for seven years teaching athletes how to read and write. Now that's a simple fact,'' Smith said. ''That is her experience. ... But I've heard her stories. I know how deeply it impacted her. The suggestion that somehow she is picking on athletes, diminishing athletes, denigrating athletes is just absurd. I just want to get that on the record.''
Hassan Melehy, a French professor, called Dean's rebuttal to Willingham's research ''a very convincing presentation'' during the Q&A session. But he warned against being tempted to dismiss ''all of the allegations of problems with the athletic program,'' mentioning years of fraud in an academic department that featured classes with significant athlete enrollments.
''There has been a pattern of denial here,'' Melehy said. ''To me, today, I don't see denial but it could be easily perceived as a pattern of denial. All effort has to be made to resist that - the temptation to use this as a pretext to insist that there's nothing wrong.''
Willingham provided her research data to the university earlier this week. By Thursday, the school had released an analysis from its admissions office saying that more than 97 percent of 1,377 first-year student-athletes admitted through special-talent policies between 2004-12 met testing thresholds used in the CNN report - a 400 on the SAT critical reading or writing test, or a 16 on the ACT - about reading levels for first-year students nationally.
''We're working really, really hard to make sure that every student who we accept here - whether they're an athlete or not - can be successful,'' Dean said after Friday's council meeting. ''I don't think we're there yet. This was not really about there's no problem. This was just about is this data, does it mean what people have said that it means? Unfortunately the answer is no.''
Follow Aaron Beard on Twitter at http://www.twitter.com/aaronbeardap