Monday, April 23, 2018

One More Bad Personalized Learning Puff Piece

Yet another example of a writer who just took the PR packet he was handed and ran with it.

Adam Thomson has covered a variety of topics for the Financial Times, including real estate in Law Vegas, a review of John Mayer's handling of Grateful Dead music, and an update on the weed industry. I guess that made him just the guy to report on that new Personalized Learning thing that all the kids are talking about these days.

No, journalists-- that is not your job.

Mind you, there is no special rule that one must be part of a special elite corps in order to report on education. But there are a couple of rules that apply for journalism of any sort that are frequently flouted by folks reporting on education, and Thomson's article is a prime example. Following just two rules would have saved this piece:

1) Do your homework.

2) Check your subject's claims.

Thomson gets off to a bad start, claiming that "if you are already an adult" you have probably never seen a mathematics class like the one he's profiling. Then he describes a class where students check in, work at stations, and complete an "exit ticket." In other words, a classroom that runs exactly as millions of classrooms have been running for years.

Thomson is visiting Joel Rose, the guy behind Teach To One, a super-duper tech-based math platform. Rose is a Teach for America alumnus who served three years, then (having beefed up his application) went to law school, then went to work with Edison Education before working on the School of One for NYC schools-- then he launched New Classrooms which now sells Teach To One. Oddly enough, Rose's corporate bio mentions that he started out as a teacher without noting his TFA background.

Anyway, Rose wants us to know that "teacher-led instruction for all kids for all kids at all times" can't be the best way for each child to learn. Then it's time to drop a definition:

Welcome to personalised learning (PL), the fast-emerging face of education that is changing traditional approaches to how we educate young minds by tailoring the content and intensity of study to an individual students' needs, abilities and goals

One might observe that this sort of tailoring is what all good teachers do on a daily basis since forever. But Thomson isn't going to go there. He will talk to Linda Shaw (University of Arizona), a specialist in disabilities and counseling.  Her observation is that

PL has been around for a long time but did not catch on until recently because teachers lacked the tools to tailor material in a highly personalised way.

Yup. Nobody ever did this before, ever. Oh, no, wait-- PL has a long history. I know this from a variety of sources, including the Teach To One website! Now, her point may be that super-duper personalization was never possible until we had-- ta-dah!!-- technology. Thomson goes to the VP of Education at Microsoft for a quote, and Anthony Salcito comes through with "the journey towards personalization has been a goal for education all over the world for the last two decades." Hey, did you know Minecraft can be educational?

Now Thomson will go ahead and note that there has been criticism of PL. Screen-time. Breaking learning into tiny unintegrated bites. Data mining. Hacking and theft of data. These are all problems people apparently talk about, but Thomson will not further examine any of them. He will, via a Rand Corporation analyst, finally acknowledge that PL remains largely undefined, but it's definitely about watching out for students who are behind or ahead. Thomson is not going to follow up on any of this, either, not even to note that tracking has been a thing for decades. But having sort of acknowledged some dissenting views, kind of, Thomson will now return to his infomercial:

Mr Rose of Teach to One remembers his time as a fifth-grade mathematics teacher in Texas, teaching 10-11-year-old students: "I had a classroom full of kids with second-grade math skills and eighth-grade math skills,"  he says. "Then I was given a stack of fifth-grade books and told, 'good luck'."

Yes, when you've only had five weeks of training and you don't stick around a school long enough to figure out how to do your job, this is the sort of thing that's a problem. And because he hasn't uncritically repeated enough baloney yet, Thomson drops the old idea that Rose's students have made "a year-and-a-half's-worth of gains in one school year." Is Thomson going to ask what that even means? Don't be silly.

Thomson finishes with our Rand guy again, who says the research on PL is mixed, and that "widely varying technology products often did not integrate with existing data systems, teachers' efforts to develop personalised lessons were very time consuming, and personalised learning plans were sometimes at odds with what was needed to pass standardised tests." More research is needed.

The final line is another quote from Mr. Pane at Rand-- "It's a nuanced story."

That may be, but there's no nuance in Thomson's story. And there certainly could be. Along with the missed opportunities noted above, I found that Googling Teach To One quickly turned up this story: "Trainwreck: The Teach to One Math Experiment in Mountain View, CA Is a Cautionary Tale About the Perils of Digital Math Education." The post at Open Cultures links back to local coverage in the Mountain View Voice (Mountain View is right in the heart of Silicon Valley), and it includes some "nuance" like this:

Since the program's launch, however, parents at both schools have voiced major concerns that the curriculum is a haphazard mess, jumping between remedial math and overly challenging course content, and that the primary role of the math teacher has been relegated to managing the program rather than to providing direct instruction. Worse yet, some parents say their sixth-grade children have become frustrated and unhappy with math under Teach to One, and are turned off to the subject entirely because of the pilot program.

I found that in roughly thirty seconds, during my lunch half-hour. The same Googling turned up the Bill Gates connection to Teach To One, and it would take little effort to uncover Gates' controversial relationship with ed reform. Presumably Thomson also has access to Google and could have found some nuance on his own.

It's the same experience again and again-- some ed tech reform guru or other announces "We have software that can create luxurious clothes that are visible only to the Very Best People," and entirely too many journalists reply, "You have?!! That's incredible! Tell me more, and I will print whatever you say!!" Journalists are supposed to tell us that the emperor is naked, not provide uncritical listing of the non-existent clothing's detail work. I have no reason to doubt that Thomson is a fine person who is just doing his best to make a living, but dammit guys-- you have got to do a better job than this.


  1. "But there are a couple of rules that apply for journalism of any sort that are frequently flouted by folks reporting on education...."

    Except not just education. Dutifully reporting [sic] whatever some government muckity-muck or corporate big wig tells you without demanding evidence or seeking other sources is pretty much standard "journalism" in all fields today. That's why so many people believe, with still no proof after a year and a half, that "da Russkies stole the election!!!"

  2. There have been two different randomized control studies of Teach to One and both found no significant gains. One is posted on the EdWeek website here: The other is here: we wrote about this in the NPE online learning report