The latest, remarkable misstep of a Cabinet nominee who has misstepped plenty came in answer to a simple question:

"Why do you think their performance is so poor?" asked Senator Patty Murray, D-Wa., in a written question to Betsy DeVos, President Trump's nominee to lead the Education Department.

"Their" refers to virtual schools, of which DeVos has been an outspoken champion. The "poor" refers to a large body of research — study after study after study — that raises serious questions about the quality and efficacy of schools that attempt to educate full-time students through the computer, without traditional access to teachers or classrooms. In response, DeVos wrote:

"High quality virtual charter schools provide valuable options to families, particularly those who live in rural areas where brick-and-mortar schools might not have the capacity to provide the range of courses or other educational experiences for students. Because of this, we must be careful not to brand an entire category of schools as failing students."

Then comes the misstep, first detailed by Columbia University professor Aaron Pallas in this opinion piece for The Hechinger Report, followed by Ben Herold reporting for Education Week.

"The following virtual academies have four-year cohort graduation rates at or above 90 percent," DeVos wrote, listing some apparent success stories:

"Idaho Virtual Academy (IDV A): 90 percentNevada Virtual Academy (NVV A): 100 percentOhio Virtual Academy (OHV A): 92 percentOklahoma Virtual Charter Academy (OVCA): 91 percentTexas Virtual Academy (TXVA): 96 percentUtah Virtual Academy (UTV A): 96 percentWisconsin Virtual Academy (WIV A): 96 percent"

There's just one problem with these numbers, Pallas points out.

They're wrong.

The Nevada Virtual Academy, for example. Its graduation rate for the class of 2015 wasn't 100 percent. It was 63 percent, according to Nevada's own school report card.

Ohio Virtual Academy's 92 percent graduation rate? Try 53 percent.

Utah Virtual Academy's 96 percent rate? Cut it in half.

You get the point.

Where did DeVos get these inflated numbers? Questions to the Trump administration went unanswered, but they appear to have been lifted verbatim from this report by K12 Inc., the for-profit company behind the online schools listed. DeVos herself was once an investor. It would not be the first of her answers to senators that appear to have been borrowed without citation.

K12 Inc. has already responded to the controversy, explaining that its numbers are "the graduation rate of continuously enrolled high school students – those who enrolled in ninth grade and remained enrolled until twelfth grade. This is not the federal graduation rate and our report makes that clear."

If it's not the federal graduation rate, as DeVos suggested it is, what is it?

"It's a fictitious rate," says Russell Rumberger, professor emeritus of education at the University of California, Santa Barbara. "It ignores almost all of the students who transfer out or drop out."

As K12 Inc. acknowledges, the rate it published only accounts for students who persisted, remaining enrolled for four years. The federal rate, on the other hand — known as the four-year cohort graduation rate — requires schools to account for all students who transfer or drop out.

In short, DeVos' numbers implied that few students had dropped out of these virtual schools when the actual rates, reported by states, suggest otherwise.

"I think this was a mistake," says Michael Petrilli, President of the Thomas B. Fordham Institute, a conservative-leaning education think tank. "It's impossible to argue that most online charter schools are high-performing because most are performing abysmally."

Petrilli should know. He once worked for K12 Inc., and, under his leadership, Fordham has studied Ohio's virtual charter schools extensively.

It's not just Ohio, either. Herold was also part of a sweeping Education Week investigation of cyber charters nationwide titled "Rewarding Failure."

K12 Inc. chooses not to use the so-called federal graduation rate because, the company argues, it isn't an accurate measure of success for many virtual schools. "High numbers of credit-deficient students and high mobility rates adversely impact the four-year graduation cohort rate for online schools," according to that 2016 K12 Inc. report.

That's a reasonable argument and worth debating. Virtual schools often cater to vulnerable students — kids who transfer repeatedly or may not intend to stay for four years.

"I agree that the four-year cohort graduation rate is limited for schools that enroll high-risk kids," says Rumberger. "Imagine students who walk in the door reading at a fourth-grade level. It's much more challenging to get that student to graduate in four years, and I don't think schools should have to."

DeVos, however, did not make that argument in her written response to Sen. Murray's question. She did something else entirely.

Instead, DeVos built an argument for virtual charter schools on language apparently taken — without citation — from a report written by a for-profit company with a huge stake in the industry. In the process, DeVos either knowingly or unwittingly mischaracterized the official graduation rates of virtual schools, making them look more successful than they are and making online learning, in general, look like a reliable pathway to student success when research suggests it is anything but.

Copyright 2017 NPR. To see more, visit http://www.npr.org/.