Personalized Learning by Facebook

If there’s one thing Facebook is great at it’s collecting and using data on “members”. Those skills are why the company is attractive to advertisers. So why not have the same programmers who built that attention-grabbing system create “a powerful tool that could reshape how students learn”?

That tool is called Basecamp, a joint project with the Summit charter school network, and is described in this Post article as a program that “tailors lessons to individual students using software that tracks their progress”. More personalized learning.

And personalized learning systems requires lots and lots of data to do the job.

But it also captures a stream of data, and Bilicki had to sign a consent form for her children to participate, allowing their personal data to be shared with companies such as Facebook and Google. That data, the form said, could include names, email addresses, schoolwork, grades and Internet activity. Summit Basecamp promised to limit its use of the information — barring it from being used, for example, to deliver targeted ads — but Bilicki agonized over whether to sign the form.

Question: if they promise not to use the data to deliver targeted ads, why is it being shared with Facebook and Google? Two of the largest distributors of targeted advertising?

Anyway, currently about 20,000 students in 100 charter and public schools are providing that data as the company is racing to have their product ready by the beginning of the school year next fall. A product that will compete with similar personalized learning systems from dozens of edtech startups.

Although the reporter tries to put a positive face on this story – starting with a headline claiming the software “shows promise” – there are so many things wrong with this project beyond the involvement of Facebook. Like this:

“There’s a lot of hype,” said Joel Reidenberg, a Fordham University law professor who researches student privacy. “In effect, they are experimenting on children.”

Then there’s the fact that the developers have very little evidence of the effectiveness of personalized learning systems.

“We really don’t know that much about personalized learning,” said Monica Bulger, senior researcher at the Data and Society Research Institute in New York.

Which applies to all the other companies on the hunt for venture capital to develop their version of personalize learning.

Not addressed in this story, of course, is whether the curriculum being “personalized”, how it is presented, and the pacing is appropriate for every child. Or if their learning from these systems will be meaningful enough to persist past the spring exams.

But I suppose none of these concerns are important as long as schools can boost test scores, administrators can keep their jobs, and investors are paid their profit.

At least in this case, a billionaire (Zuckerberg) is paying the bills. And Summit is the one organization in the world immune to potential data loss.

“We’re offering this for free to people,” she [Diane Tavenner, chief executive of Summit] said. “If we don’t protect the organization, anyone could sue us for anything — which seems crazy to me.”

I’m convinced.

The New Business of School

In case you haven’t heard, Silicon Valley is getting into the disrupting education business. Several tech entrepreneurs, including Salman Khan himself, are creating schools with the ultimate end goal of earning big profits from learning.

A recent segment of the podcast Note to Self visited the Brooklyn location of one called AltSchool, founded by a former Google executive and backed by $100 million from folks like Mark Zuckerberg and Marc Andreesen. The goal of the company is to “optimize” education by build a “new operating system” for schools, one that seems to lean very heavily on the concept of “personalized” learning.

The descriptions and interviews from the program makes the instructional process sound great, and likely wonderful for the kids involved. But in effect they are guinea pigs, contributing data so the company can take what they learn from these ten or so small elementary and middle schools and scale it up into marketable products that could be used in most “regular” public and charter schools. Plus make projetsons teacherfits worthy of a hundred million dollar investment, of course.

I have many doubts about these experimental schools, but especially about the scaling part.

To start, the kids attending these “micro-schools” come from families that can afford to pay the up-to $30,000 tuition, or qualify for a grant. Public schools spend far less per student, and few parents could afford that kind of money for private. And we already know that small classes (30 to 120 students per AltSchool) with highly motivated students, and parents, led by a well-qualified staff almost always leads to better learning.

But what happens when they try to take the same concepts and put them in a public elementary school with 25 not-specially-selected kids per teacher? As the number of adult-kid interactions decline in the classroom setting and the software they are building (based on collecting lots of data, something else that should raise a few red flags) takes over, the experiences will also change. It’s likely test scores, not to mention student engagement, will decline noticeably.

Go listen to the whole thing and see if you hear something different. You may also want to read this profile of the company and it’s founder in Wired.

However, for now, I have many, many questions – and doubts – about this and other “high tech school” startups, siding with this view from the podcast.

NPR’s education reporter Anya Kamanetz is skeptical of Ventilla’s goal to optimize education for the masses, and she’s concerned about Silicon Valley’s foray into education. “They have a giant promise, which is that the right software system, the right operating system, is going to transform teaching and learning… and, what it ultimately means is that they have shareholders to satisfy.”

Transforming teaching and learning is not necessary compatible with making profits large enough to satisfy those shareholders. Nor should it be.

More Status Quo

So, spending on edtech is growing. Which means the amount of money being invested in edtech companies is also growing as investors anxiously anticipate huge profits from the “education sector”. These days all those edtech startups, and their funding chase, are not only being covered by business news sites, but also by general interest education information sources.

Here are just three examples I’ve stumbled across recently and I think they have one major attribute in common. See if you can spot it.

First is Nearpod, as reported by Tech Crunch, a website that covers the technology industry in general. The company just received an investment of $9.2 million in series A funding for their product that helps “teachers use tech for live instruction”.

Nearpod’s app lets teachers deliver digital lessons to students right on their mobile devices, during class.

First, teachers sign up and select from a smorgasbord of digital lessons on the Nearpod content marketplace. Then they assign a digital lesson to students, who engage with the material in class via student accounts on Nearpod.

The lessons feel like interactive stories, projects or mobile games, typically. But they give teachers a view on students’ mastery or struggles with particular topics.

Then there is Brainly, which is getting “$15 million in a Series B funding round led by Naspers that brings its total funding to $27 million”.

Brainly is a bit like Quora for students, a social network where children and teens come to help one another work through homework problems that are stumping them. “Peer-to-peer learning” is how the company describes it.

Students earn points for the quality of their answers and can eventually climb into the leaderboards for subjects like math, biology, and so on.

Finally, we have Lilwil, “one of the hundreds of projects built overnight at the TechCrunch Disrupt NY Hackathon” and not a company that venture capital people are funding. Yet.

It then applies IBM’s Watson to assess the different personality traits of students based on their work such as openness, conscientiousness, extraversion, agreeableness and emotional range.

Lilwil then presents a personality analysis to teachers, along with suggestions for the best methods for assisting that student. For example, Lilwil could identify higher levels of extraversion and conscientiousness in a student, and determine that they’re best taught through role-playing simulations and roundtable discussions.

For me, the common thread in these hot new products is that none of the technology will be used by students. These are all tools for teachers and administrators. Even Brainly, which sounds like it’s building a community, is nothing more than a help session for kids to help each other work their way through traditional homework assignments assigned by adults.

Schools and districts that spend their scarce edtech funding on these tools are reinforcing a direct instruction model for the classroom, rather than enabling kids to create and communicate. These and many other edtech startups are building products that reinforce the status quo, rather than something “innovative” or “disruptive”.