The Atlantic’s Caitlin Flanagan came to her most recent assignment with moral and factual certitude, starting with the headline: “The University of California Is Lying to Us.”

In Flanagan’s crosshairs was the UC system’s decision to phase out the SAT and ACT in admissions and scholarship decisions. That choice responds to a legal settlement with students who argued that requiring standardized tests discriminated against poor students of color.

“It is a move so widely hailed by the administrators and faculty that you know someone’s getting hustled, and in this case the marks are the state’s low-income Black and Latino students — the very ones whom the new policy is supposed to help,” writes Flanagan, who also argues that the decision will “probably hurt thousands of Asian American teenagers” as well.

On Thursday afternoon, UC blasted the Atlantic in a letter alleging “unfounded assertions and misleading generalizations” in the piece. UC also said that though it had cooperated in the fact-checking phase of the story, the magazine “elected to ignore the information we provided.”

“It is completely untrue that we ‘elected to ignore’ information from the UC,” Atlantic spokesperson Anna Bross writes in an email. “They were given every opportunity to clarify these points pre-publication during the fact-checking process, but didn’t.”

In bashing the system’s decision on standardized tests, Flanagan cites the findings of a 225-page study from a task force on standardized testing empaneled at the request of former UC president Janet Napolitano. The report didn’t find evidence that UC’s use of standardized testing “played a major role in worsening the effects of disparities already present among applicants.”

Following an inventory of the reasons for ditching the tests, Flanagan roars: “These are not facts. They are assumptions, all of them flawed or flat-out incorrect.” With that, Flanagan launches into a multipronged defense of the status quo ante at UC. But several of those prongs have attracted pointed factual challenge from Zachary Bleemer, a postdoctoral fellow at Harvard University’s Opportunity Insights, and Sarah Reber, a UCLA associate professor and an economic studies fellow at the Brookings Institution.

For starters, Flanagan insists that the California system had developed an innovation to avoid an often alleged pitfall of standardized testing. “One of the ways the UC system found to work around the state’s ban on affirmative action was to evaluate test scores ‘in local context.’ You didn’t need to be a top test taker in California to be UC-eligible. You just needed to be a top test taker within your own school,” she writes, in what could be only a reference to the system’s clumsily monikered “Eligibility in the Local Context” (ELC).

Trouble is, the ELC doesn’t work the way Flanagan describes it. “[S]ince the ELC began in 2001, it has only evaluated students on the basis of high school grades, not test scores,” Bleemer noted in a letter that he posted this week to Twitter. The UC letter reaches this conclusion: “[T]he author’s application of ‘local context’ is wrong.”

Another Flanagan argument:

The biggest barrier to getting into the University of California is not the SAT; it is, again, the GPA. Because students at underfunded schools have such limited access to college counseling, they often assume that if they want to go to the UC, they should keep an eagle eye on their GPA. What many don’t know is that, to be eligible, they must complete a series of 15 college prep classes called the A-G requirements. Good grades in other classes don’t count. (And—shockingly—some high schools don’t even offer all the A-G requirements.)

To understand why this paragraph is misleading, a brief tutorial is required: As Reber notes in this explainer on UC admissions, the system offers “guarantees” to students if they qualify for one of two pathways — the ELC (discussed above), which premises eligibility on certain course and GPA thresholds; and the “statewide index,” which, until the recent decision, set an eligibility requirement based on GPA and test scores. Those systemwide “guarantees,” however, guarantee admission into only the least selective school in the system, UC Merced.

“[M]any applications from eligible students will not be competitive for admission, particularly at the more selective campuses,” writes Reber. Students wishing to attend UC Berkeley or UCLA, for example, need to show much more than mere eligibility for the pathways cited above. And it’s here that Flanagan’s point about noncore classes comes into play: Of course applicants to competitive UC campuses need to show achievement in “other classes.” As Bleemer wrote in his letter to the Atlantic, “This statement confuses UC eligibility — the minimum requirements needed to be admitted to UC’s least-selective campus — and UC admission. Good grades do not count for eligibility, but they do count for admission to every UC campus.” See this rundown of UC application review factors.

And here’s Flanagan’s slam-dunk-sounding passage in favor of standardized testing:

There was a loophole these students could use—a Hail Mary option for smart kids who, for whatever reason, didn’t do well in high school or did well but not in the A-G classes—and it involved test scores. In 2018, about 22,000 students “tested in” to the UC, meaning their high SAT scores compensated for low GPA. Almost half of those students were low-income, and more than a quarter were Black, Latino, or Native American. The UC has now taken this lifeline away.

The situation isn’t quite that simple. When Flanagan says that in 2018 about 22,000 students “tested in” to UC, what she’s really saying is that they qualified for the “statewide” guarantee discussed above. It relies on a mix of GPA and test scores. And when Flanagan suggests that “SAT scores compensated for low GPA,” she leaves out that the reverse could be the case for many applicants — i.e., that good GPA numbers compensated for low SAT scores.

As for the characterization of standardized scores as a “lifeline,” the data offer a conflicting portrayal. More than 8,000 of that group, notes Bleemer, were rejected by every UC school to which they applied. Since they qualified for the “statewide” guarantee, however, they were assured admission to the least selective option, UC Merced. How many chose that option? No more than 168.

So Flanagan is positing a “lifeline” that the students themselves don’t value.

As the UC letter states, “These students did not ‘test in’ to UC.”

Earlier this week, the Atlantic published one correction pursuant to the factual pushback that cropped up on social media. The story had incorrectly stated that the system “could waive course-load requirements for applicants with high SAT or ACT scores.”

With that gesture, the Atlantic appeared poised to wrap up its amendments to the story. Following publication of the UC letter, however, the Atlantic made further adjustments. “We corrected one error in fact on Tuesday evening,” writes Bross. “We’ve continued to review this closely and in a few places where we stand by our original reporting, we added clarifying language to better explain the issue and prevent additional misreading.”

That’s a very Atlantic way of blaming the reader for factual inaccuracies.

Here’s how the italicized walk-back note now reads:

This article originally stated that the UC could waive course-load requirements for applicants with high SAT or ACT scores. In fact, high test scores could be used to compensate for low GPA, not to waive the course requirements entirely. The article has also been updated to clarify how test scores are evaluated to avoid confusion with the UC’s Eligibility in the Local Context program, and to clarify that the task force report identified the number of students who were guaranteed admission on the strength of their test scores, not the number of these students who ultimately enrolled.

What’s missing here? The word “correction.”

The UC letter and other rebuttals make clear that Flanagan bootstrapped complex admissions data and procedures into a hot take that cooled upon inspection. There might be strong arguments in favor of keeping the SAT in the UC admissions process, but don’t go looking for them in this piece.