Ravitch was herself a supporter of high-stakes testing. Back when the research pointed that way. This kind of approach was applauded as being best practices and in the interests of the children. And it empowered researchers.
Now, research of no demonstrably higher quality says something else. If those results were being implemented by Duncan he'd be applauded. Because it empowers researchers. Advocates for their own agendas.
A lot of what the researchers say makes it into the classroom. Much of it is twaddle.
In some cases, some body comes along and exposes it as twaddle through a peer-reviewed study.
In most cases, some instructional specialist goes to be trained at a university then comes back with the Ultimate Solution to Being Off Task. USBOT. Teachers look at USBOT and say, "Ah, I've done that for the last 20 years. I was taught to do that but it had a different name then, and my teacher said it had a yet different name when he was learning it 50 years ago."
Or they look at it and roll their eyes. "Of course it's never been tried. Not by a teacher that had his contract renewed, at least. It's a foolish idea." And it's tried, the researcher applauds the innovative and forward-thinking school, and it flops. "You didn't implement it properly. You weren't trained properly. You misunderstood the basic principle." Etc. But 10 years later if you try it that same researcher, never admitting that his idea was idiocy, would decry it as wrong-headed.
At least Ravitch is honest. However, in this case there's a kind of implicit dishonesty: The researchers were all opposed to Duncan. However, had any one of the researchers' ideas been adopted, Duncan would still have been booed by a hefty percentage of the other researchers.
Such is education research. And why many teachers, when they say they have a masters, say it's a *real* masters and not a masters in education.