Monday, July 28, 2014

Steering the Narrative (Part 3): The Martin Report

This is Part 3 of a three-part essay. You can read Part 1 here and Part 2 here.

Omitting Key Findings

The N&O’s reporting and editorializing on the Martin Report may be best described as journalistic malfeasance, and executive editor John Drescher is the culprit most blameworthy.

Jim Martin is a former North Carolina governor and U.S. representative with a PhD in Chemistry. With the assistance of Baker Tilly, a national accounting firm, Martin conducted the most extensive investigation of the UNC scandal to date, gathering data as far back as 1994. After he concluded his investigation, he presented his Report to the UNC Board of Governors, on December 20, 2012. In Martin’s presentation, he distilled the Report down to 15 key findings that conjointly support his conclusion (pp. 71 – 74). However, his conclusionthat the AFRI/AFAM scandal was an academics scandal, not an athletics scandalwas not the conclusion Drescher was counting on for his star investigative reporter to win a Pulitzer.

In response, Drescher has coordinated a propaganda campaign to discredit the Report, grossly misrepresenting its findings in an editorial he himself wrote and directing four other N&O writers (with reinforcement from the editorial board) to dismiss the Report in similar fashion. Drescher and his staff’s strategy has been to overemphasize the importance of select findings while cunningly withholding reporting on the majority of the findings.

In Drescher’s editorial, published on February 1, 2013, his manipulation of the facts is so egregious it is not merely a matter of misinterpretation: it is a blatantly dishonest reading of the Martin Report. When one of Martin’s 15 key findings had to be retracted, Drescher seized the opportunity to exaggerate the significance of that finding and thereby call the entire Report into question. In both the title and the first sentence of his editorial, Drescher refers to the finding as a “key finding,” but nowhere in his editorial does he inform readers that the finding was just one of 15 key findings, the other 14 of which are still intact. (Throughout this essay, I will cite aspects of the other 14 findings. Readers interested in examining all 14 findings in their entirety should read pp. 71 – 74 of the Martin Report.)

The retracted finding was that an Athletics official and the Director of the Academic Support Program twice raised concerns, first during a meeting in 2002 and then in 2006, with the Faculty Athletics Committee about the lecture classes Julius Nyang’oro conducted as independent studies. After the Martin Report was released, Dan Kane queried the faculty members on that committee, and he discovered that Martin never interviewed them about those meetings. Moreover, none of the faculty members could recall concerns being raised. Unquestionably, Martin should have interviewed those faculty members to verify the claims about the two meetings, and without verification that finding should not be included in the Report.

However, that finding, again, was only one of 15 that Martin cited to support his conclusion. His other 14 findings remain incontrovertible, though Drescher editorializes as if they do not even exist. 

Drescher's editorial is replete with suggestions that Martin staked his conclusion entirely on that one retracted finding and that the entire Report reflects the questionableness of that finding. For example, after Drescher quotes Martin asserting that “findings and conclusions should be based on evidence, not hearsay and imagination,” Drescher smugly interjects, “If only that were what Martin’s report did.” In light of the unverified anecdotes and quotes Dan Kane has included in his reporting on the UNC scandal (the subject of the first two parts of this essay), Drescher’s criticizing Martin for a singular, ultimately irrelevant misstep is hypocritical in the extreme. What is more, Drescher knows the Martin Report’s 101 pages (including Addendum) contain an overwhelming amount of hard evidence, including voluminous statistical data, that extends well beyond the one retracted finding. Drescher's selective citing of the data demonstrates he has read the report and is intelligent enough to comprehend it; however, he is too committed to the sensationalized narrative to represent the Report's findings with fairness. 

Consider, for example, the following passage from Drescher’s editorial:

At issue were 172 bogus classes within the African and Afro-American Studies department. About 45 percent of the students in those classes were athletes; fewer than 5 percent of UNC students are athletes. Baker Tilly explained that discrepancy by noting that a disproportionate share of athletes are African-American and more likely to take a course in that department. 
That might be a factor. But the evidence also is overwhelming that the academic support staff, which at the time effectively reported to the athletic department, knew these classes did not meet and steered athletes toward them

Drescher’s reporting leaves readers with the impression that Martin merely rationalized athletes’ disproportionate enrollment in the anomalous AFRI and AFAM classes by citing the fact that UNC athletes are disproportionately African American. Drescher, however, neglects to mention an entire section of the Addendum, titled “Summary of ‘Clustering Analysis,’” which thoroughly demonstrates the seriousness with which Martin took the athletes’ disproportionate enrollment (pp. 5  6).

For his analysis of clustering, Martin and Baker Tilly investigated whether athletes’ enrollment in the grouping of anomalous “paper classes” was any higher than in other groupings of classes in which athletes were disproportionately enrolled, both within and outside the AFRI/AFAM department. After identifying several “cluster groupings” of classes across multiple departments, Martin investigated them to assess whether any irregularities in their administration had occurred. He found no irregularities in any cluster groupings outside the grouping of anomalous AFRI/AFAM classes. (The other cluster groupings are referred to as “cleared.”) The Report explains that one reason cluster groupings attracted athletes was because the classes fit into athletes’ exceptionally busy schedules better than did other sections of the same courses.

Most importantly, Martin and Baker Tilly found that athletes’ enrollment in several cleared cluster groupings across multiple departments was consistent with athletes’ enrollment in the grouping of anomalous classes. In the grouping of anomalous AFRI/AFAM classes, athletes’ enrollment rate was 44.9%; in the cleared cluster grouping of AFRI/AFAM classes, athletes’ enrollment rate was actually slightly higher, at 48.9%; and in the cleared cluster groupings outside the AFRI/AFAM department, athletes’ enrollment rates ranged from 44% to 47.7%. In other words, athletes’ enrollment in the grouping of anomalous classes was no more disproportionate than it was in several cleared cluster groupings of classes across multiple departments. All of the cluster groupings, including that of the “paper classes,” appear to have attracted athletes for the same reason: the classes accommodated athletes’ exceptionally busy schedules better than other classes.

Readers would not understand that, however, if Drescher’s editorial (and the N&O's news reports) were their only source of information.

Additionally, the percentage of football players majoring in AFRI or AFAM between 2001 - 2010 never exceeded 13%. Furthermore, only three times during that same timespan did the number of men’s basketball players majoring in AFRI or AFAM exceed two, and the percentage exceeded 25% only once. Yet Drescher avers that the “overwhelming” evidence demonstrates academic counselors “steered” athletes to the anomalous AFRI and AFAM classes. If “steered” in this context means anything other than sometimes recommended, then Drescher’s criteria for evidence do not reflect any epistemology with which I am familiar. Conjecture and insinuation are not typically understood as evidence.

Indeed, the narrative Drescher and his staff have been steering seems to be underscored by a uniquely journalistic epistemology that is defined by a single principle: a statement becomes true if a journalist can convince readers it is true. Journalism as a discipline of verification is clearly not the journalism Drescher practices.

Neglecting Context

As mentioned above, Drescher’s campaign against the Martin Report has been carried out by multiple N&O writers. Not surprisingly, Dan Kane has been at the forefront of the charge.

Altogether, the Martin Report and Addendum totals 101 pages and contains not only the key findings and the data on which those findings are based but also an explanation of the procedures, or methodology, by which Martin and Baker Tilly collected, analyzed, and interpreted the data. (If only Mary Willingham were so thorough.) Yet Kane, an investigative reporter, cites alarmingly scant data from the Report and never even mentions the procedures, allowing detractors to dismiss the data as a “smokescreen” meant to divert the public’s attention from the real scandal. (By the way, Pack Pride’s censure of the Martin Report’s emphasis on quantitative data is rather ironic in light of the fact that NCSU is best known for its Engineering school.)

Notwithstanding Kane’s complacent disregard for the procedures, the Report’s methodology provides context to and justification for the data, and it demonstrates the relation between the data, the interpretive findings, and the conclusion. To illustrate some of the primary procedures, I have included the below diagram from the Report, and I encourage those who want to gain a more comprehensive understanding to read the “Approach for Completion of this Review” section on pages 4 – 5 of the Report. Because this essay is already lengthy, the diagram will have to suffice for now so that we can focus on the data itself.

Again, Kane cites very little data from the Martin Report, selecting only that which he can attempt to explain away. His article written with Andrew Curliss following the release of the Addendum is the clearest example of Kane's selective reporting. For nearly every one of the few data points Kane and Curliss cite about athletes in the anomalous classes, they neglect to cite the corresponding data point about non-athletes. By so doing, they remove the context in which to understand the data, thereby leading readers to sensationalized inferences. Reporting statistics with little to no context has been one of the standard methods by which N&O journalists have steered the narrative.

For example, consider the lede and following sentence in Kane and Curliss’s article:

New data released Friday about a long-running academic fraud scandal at UNC-Chapel Hill show athletes made up nearly half of the enrollments in 172 bogus classes within the African studies department, and also accounted for a just under half of 512 suspect grade changes during that period. 
Their average grade: 3.56, between a B-plus and an A-minus.

To readers unfamiliar with the Martin Report, Kane and Curliss’s opening two sentences may seem like a straightforward account of the findings. However, absent from the findings Kane and Curliss present are several other, related findings that provide context.

First, although Kane and Curliss mention the 172 anomalous AFRI and AFAM classes, they neglect to mention the finding that not all of the anomalous classes had athletes enrolled. In fact, there were more anomalous classes with only non-athletes enrolled (26) than there were anomalous classes with only athletes enrolled (21).

Furthermore, the isolated second sentence leaves readers with the impression that athletes were being graded with exceptional leniency. However, Kane and Curliss neglect to mention that the average grade point for athletes (3.56) was actually slightly lower than for non-athletes (3.63).

Additionally, like Drescher, Kane and Curliss mention that “nearly half of the enrollments” in the anomalous classes were athletes, but the reporters neglect to explain that the 44.9% athlete enrollment rate was no more disproportionate than it was in several cleared cluster groupings across multiple departments.

Kane and Curliss also neglect to report any context for understanding the unauthorized and suspected unauthorized grade changes. Among those irregular grade changes, Martin and Baker Tilly identified some that were made to permanent grades and some that were made to temporary grades. The distinction is important because, as the Report states,

We would expect student-athletes to have more temporary grade changes than non-athlete students, as the demands of an athletic schedule are more likely to create conflicts or delays in student-athletes’ abilities to [complete assignments] at scheduled times.

Therefore, in classes in which irregular grade changes to permanent and temporary grades occurred for athletes and non-athletes alike, we would only suspect preferential treatment for athletes if their permanent grades were irregularly changed at a rate higher than were non-athletes’ permanent grades. Martin and Baker Tilly found the opposite. The 22 irregular grade changes to athletes’ permanent grades constituted only 1.6% of athletes’ enrollments, whereas the 75 irregular grade changes to non-athletes’ permanent grades constituted 3% of non-athletes’ enrollments. In other words, non-athletes were almost twice as likely as athletes to receive an irregular grade change to a permanent grade.

In short, Drescher, Kane, and Curliss (and other N&O writers) have neglected to report multiple findings from the Martin Report, including the following four: (1) athletes enrolled in the “paper classes” no more disproportionately than they enrolled in several other, cleared cluster groupings of classes across multiple departments, (2) not all of the “paper classes” had athletes enrolled, and there were more classes with only non-athletes than there were classes with only athletes, (3) athletes’ average grade point in the “paper classes” was slightly lower than was non-athletes’ average grade point, and (4) non-athletes were nearly twice as likely as athletes to receive an irregular grade change to a permanent grade.

Those are not the findings of an athletics scandal.

In his editorial on the Martin Report, Drescher writes, “Martin and Baker Tilly seemed more determined to absolve the athletic department of blame than to get to the bottom of what went wrong.” If the straw man version of the Martin Report constructed by Drescher and his staff were the real version, then Drescher’s assertion would have merit. However, as Drescher knows, the straw man version is an insidious fabrication propagated by the N&O and parroted by complacent journalists from the national media. N&O writers have executed their campaign against the Martin Report by overemphasizing the importance of the one retracted finding and consistently reporting only select findings while omitting other findings that would provide context. When we examine the Martin Report ourselves and contrast it to the N&O’s straw man version of it, we see that the converse of Drescher’s assertion is more accurate: Drescher and his staff seemed more determined to implicate the Athletics department than to represent the findings of the Martin Report accurately.

Author's note: I had intended this three-part essay to be four parts, but I could not verify the information for the fourth part before having to publish other entries. However, I cover the intended topic—the administrators who knew about the paper classes—in later essays.