Amazon, The Algorithm, and the Future of Education

Intriguing piece in the New York Times yesterday, looking at Amazon's bookstore (and business in general) and how it has become an unholy mess. 

It is framed by a lawsuit being brought by an author, John C. Boland, who has found his own work listed at hundreds of dollars with a false, much earlier, publication date. This, it turns out, is just a tip of the proverbial iceberg. The online market is overrun with third party sellers, and has not shown much interest in policing resulting way-open marketplace. Reporter David Streitfeld rattles off the list of unhappy people:

There are sellers like Mr. Boland, who say they are suffering from the Wild West atmosphere on the site; regulators, who are taking a closer look at Amazon’s power; unhappy warehouse employees, who would like a better deal; and lawmakers, who want Amazon to disclose more about its third-party sellers. There are also the devious sellers themselves, whom Amazon says it is having a hard time eradicating.

Amazon has defended itself from Boland's lawsuit in court, but Streitfeld shows example after example of how the algorithm fails customers, ranking books in categories they don't belong and worse, elevating crap. I don't mean another badly written Dan Brown novel, but things like a knockoff Dave Grohl biography written in Almost English and yet sold "side by side" with the actual Grohl memoir. Meanwhile, the algorithm also removes all sorts of books for reasons known only to the computer. 

Streitfeld talks to Stacy Mitchell, author of Amazon's Toll Road: How the Tech Giant Funds Its Monopoly Empire by Exploiting Small Businesses. 

Amazon “doesn’t care if this third-party stuff is a chaotic free-for-all,” she added. “In fact, it’s better for Amazon if legitimate businesses don’t stand a chance. In the same way Amazon wants to turn all work into gig jobs, it wants to turn running a business into a gig job. That way it can walk off with all the spoils.”

Or consider this:

“In some ways Amazon doesn’t really want to be a retailer,” said Juozas Kaziukenas of Marketplace Pulse, an e-commerce consultant. “It doesn’t want to do curation or offer human interaction,” two of the essential qualities of retail for centuries.

Offering tens of millions of items to hundreds of millions of customers prevents any human touch — but opens up a lot of space for advertising, and for confusion and duplicity. This might be good for Amazon’s competitors in physical bookstores, which have a much smaller and more tightly controlled stock. But it does not bode well for e-commerce.

Why am I writing about this here? Because the future that some folks envision for education is education as e-commerce. Parents get their voucher, aka Education Savings Awesome Freedom Account debit card, and then go shopping for vendors to provide various bits and pieces of a full education. 

One common feature of the various ESA programs being rolled out or expanded across many states is that, like Amazon, the state provides little or no oversight or is even actually forbidden from exercising oversight of these various edu-vendors. It is the parents' problem to sort through the marketplace, to figure out which vendors are the real deal and which are not. In effect, vouchers turn running a school into a gig job done by parents, who will find themselves bombarded with the educational equivalent of fake biographies in Almost English and robocalls from the nice lady worried about your car warrantee. And for the bazillionth time, no, that algorithm being marketed as AI is not going to do the job for you, nor is it going to successfully personalize your child's education. 

Amazon is a trillion-dollar company that  has been unable to make all this work. There is no reason to believe that a privatized education market place will do any better.

(Incidentally, between bookshop.org and alibris.com, I find pretty much every book I'm looking for)

PA: Board Activism Versus Board Business

Earlier this week the New York Times ran a piece by Campbell Robertson contrasting the light and heat surrounding school boards with the actual problems crying for attention. The piece opens with a fairly stark example from Doylestown, PA:

Early in the November school board meeting, a few of the departing members made farewell remarks, talking of things that they believed still need addressing: more special education programs, mental health initiatives, a program for high school students to take college classes. There was a long list, but over the past two years other things had gotten in the way.

When the meeting opened up to public comments, there was an indication of what those other things might be. Parents and other residents took turns standing before the board, speaking about Zionism, Maoism, slavery, freedom, the Holocaust, critical race theory, the illegality of mask requirements, supposed Jewish ties to organized crime and the viral falsehood that transgender students were raping people in bathrooms.

“I fight here week after week,” one woman said, “to ensure that my children will never be subject to having their freedom taken from them.”

"Well, at least I don't have to wear a mask" 

There's this thing that happens, often, with one-issue school board candidates. They run for their particular issue (often a local concern like "save our elementary football program") but then they get on the board and discover that school board work is largely about very unglamorous nuts and bolts decisions. But now we're seeing districts that are being ground to a halt by people who don't care about the nuts and bolts--they just want their righteous--and sometimes imaginary--cause to be championed, at length.

We'll know better in the next couple of months. Some analysts believe that most of the board candidates who ran on liberty and anti-mask and anti-CRT did not win. 

But I live in Trump Country, and several local districts are feeling the pinch. Keystone School District was one of many that saw a contentious and successful campaign by anti-mask, anti-vax, grumpy parents candidates whose stated objectives include things like "put myself between our government and our kids." That in district's whose board has already spent a lot of angry time in struggles with mask mandates, to the point that they've had trouble filling the superintendent spot.

If you really want to see this sort of trouble in action, let me take you down the road a little past the Keystone District to Redbank Valley Schools, where an incumbent was unseated by a young local. Mitch Blose is a RVHS grad and volunteer coach. The election was not as contentious as some (pro tip-- in PA board candidates can cross-file with both parties, so trying to track party affiliation is not always helpful), and in the end Blose won.

His first act as a new member was to shut down board business.

The member-elect attended the board meeting as a member of the public, and refused to wear a mask. So the board held an executive session (required in PA for any personnel or student issues) and then quit for the night. This isn't remotely the first such occurrence-- back in September the Oil City School District board shut down a meeting because a couple of freedom fighters, including one student, refused to mask up for the meeting. Similar tales can be heard from around the country.

The Redbank Valley incident gets some extra detail because nobody thought to turn the livestream off right away, so Blose's arguments with board members was broadcast live. Nobody wanted to defend masking, but one board member pointed out that it looks bad when you make the kids do it and you don't. But another gets closer to the heart of the issue.

“Mr. Blose, I don’t want to wear this mask,” Reddinger answered. “I don’t want to see anyone else wear the mask, but at the same time, I am obligated to this board, sitting in this chair. I am obligated to the taxpayers, to the lawsuits."

As is infrequently noted, mask mandates are related to liability issues. If the state has mandated masks, and you don't follow the mandate, you are an expensive lawsuit waiting to happen. The first time someone can show they got sick in your facility, it will be costly. Consequently, many school district lawyers have been having quiet Come To Jesus meetings with their boards. 

In Pennsylvania, the state mandate for school masking is set to end in January, at which point local districts will have to relitigate the whole thing on their own. Expect lots and lots of spirited public debate, a reshuffling of enrollment in and out of cyber schools, and in general a whole lot of time spent on things other than the actual regular operation of the school district and dealing with the various challenges of doing the work. Bus drivers. Bus routes. Budgets. Program costs. Special ed. Lunch prices. Whatever specific challenges have cropped up in the district. And there's an unmeasurable secondary effect--how many people who would have been really useful board members said, "Well, I don't want to wade into this mess," and stayed at home instead? I'd like to be more optimistic, but I see some rough days ahead. 

Chiefs For Change Recruiting Big Brother

What a twisty road it has been for Chiefs for Change.

They were supposed to be part of the big web of education reforminess that would usher Jeb Bush into the White House, and they've been flailing ever since. They've just suggested a cool new way to expand the surveillance state, but before we look at that, let me crib from my own previous work (here and here) to explain how we arrived at our current situation.

CFC was originally spun off of Jeb's Foundation for Excellence in Education (FEE), a group that lobbied hard for Common Core, school A-F ratings, test-based evaluation, and mountains of money thrown at charter schools. FEE started up CFC because they thought that the Council of Chief State School Officers (CCSSO), the group that holds the Common Core copyright and was the figurehead guiding force behind the core's creation-- that group wasn't aggressively reformy enough for the Jebster.

Initially, the group was to be a new nexus of reform, but they were immediately beset by problems. And I'm not counting the naming problem-- did they think that change would never come, or once the change was the status quo, were they going to just disband? I mean, if your brand is that you favor change, does that mean you just keep trying to change the change that you just implemented? Do you ever say, "Well, hell, no-- we don't worked hard to install that policy and we surely don't want to change that!" I'm just saying-- doesn't seem like a very well thought out name.

At any rate, by the time Jeb's 2013 Reformster Convention rolled around (incidentally, the 2016 National Education Summit is scheduled for end of November in DC-- I wonder how that's going to go), the Chiefs were already in rather a mess. Chief Tony Bennett had already had to change jobs because of that whole lying and cheating thing. And Jennifer Berkshire provides a great account of Rahm Emmanuel's speech, a weathervane moment that showed the Winds of Change no longer at CFC's back.

Since those not-so-halcyon days, CFC has decided to implement a little mission creep. In 2015 they dropped their connection to FEE, which was more than financial, but also structural and organizational. They also decided to change their definition of "chief." Previously that had meant a state-level education chief, but they had already developed a problem in that department, in that many of their big names (like Chris Barbic, Kevin Huffman, et al) were no longer actual leaders of state education programs-- six of the seventeen members in 2016 were "formers." And so "chief" was redefined to include superintendents of school systems. 

That expansion continued, and since the group included so many Teach for America products anyway, they cribbed from the TFA playbook to redefine themselves as a group promoting diversity. That has included bringing in chiefs from various charter and privatization advocacy groups. These days CFC includes ten formers (one of whom is listed as "Chief in Residence"), superintendents from Ector County, Wichita and Akron, and the executive director of the Broad Center, the Yale-based version of the old Broad Academy Fake Grad School for Superintendents.

CFC has continued a string of not-very-awesome projects. There was the Student Growth Simulator, which turns out to be a calculator that does subtraction augmented by cool graphics--and the damned thing is still up. Then in 2019 they made a call for an end to "toxic rhetoric" on school choice, which turned out to mean--well...

That is why today we are calling on policymakers across the nation to end the destructive debates over public charter schools. Proposed caps and moratoriums allow policymakers to abdicate their responsibility to thoughtfully regulate new and innovative public school options: like banning cars rather than mandating seatbelts. They are a false solution to a solvable problem.

  

Yeah, not good old fashioned tone policing, but an actual call for people who were saying the wrong thing to just shut up.

So now, CFC has some thoughts about data collection. Specifically, they're in favor of more of it.

They've teamed up with Data Quality Campaign. DQC was set up in 2005 with ten reformster partners:

Achieve, Inc. (www.achieve.org)
Alliance for Excellent Education (www.all4ed.org)
Council of Chief State School Officers (www.ccsso.org)
The Education Trust (www.edtrust.org)
National Center for Educational Accountability (www.nc4ea.org or www.just4kids.org)
National Center for Higher Education Management Systems (www.nchems.org)
National Governors Association Center for Best Practices (www.nga.org/center)
Schools Interoperability Framework Association (www.sifinfo.org)
Standard & Poor's School Evaluation Services (www.schoolmatters.com)
State Higher Education Executive Officers (www.sheeo.org)

DQC has always been all about hard-core data mining of students, agitating for, among other things, more holes in FERPA. They continue to push for testing data to be collected.

The "report" that CFC is pushing sounds an alarm:

State and local K–12 leaders need access to actionable information about the postsecondary and workforce outcomes of former students to make decisions that will prepare current students for success.

"K-12 leaders are hungry," says the report, "for a more complete, longer-term understanding of former students' postsecondary and career journeys."

And they are not entirely wrong. My old school used to send out follow-up surveys to try to keep tabs on graduates. Heck, every time a former student visited I asked them about the good and bad of how I'd prepared them (or not) for their life. There is no doubt that having information about the trajectory of students' lives after graduation would be helpful for schools-- though fresh data from recent grads doesn't necessarily capture the long-term payoff of education, and longer-term data that is decades old might not be all that actionable.

However, it's one thing to ask for this kind of data, and quite another thing to just take it.

The list of data that CFC/DQC wants to have is large: industry data, apprenticeship data, post-secondary enrollment, completion and course data (both in and out of state), financial aid data, veteran/military data, wage data, much of it disaggregated by demographic details and by institutions. CFC says that "leaders want" this data student level, longitudinal, de-identified, in context, timely, easy to use, and "paired with support for analysis and use." 

Currently, they say, "the data is not easy to find or use." So they propose a "cross-agency data governing body"-- a state office of data overlordship. A big statewide data infrastructure. User-friendly data tools, sort of like Big Brother with a convenient dashboard. Oh, and funding for all of it.

This is a huge data grab, reminiscent of the old cradle to career pipeline which also required constant data collection and crunching. Beside being a hugely intrusive extension of the surveillance state, it's also the information age equivalent of a land grab, an attempt to swoop up and take possession of a valuable before anyone catches on ("We'll be happy to take that ugly black goo seeping out of the land off your hands! It'll make things better for you, we promise!")

These attempts to launch vast, intrusive data-mining operations are not new, and they aren't likely to stop any time soon. I don't know how much juice CFC has left to actually accomplish large projects like this these days. But it's worth staying alert for any time that Big Data waves its tentacles again. And no matter how much Chiefs for Change is hard up for members, they should not be recruiting Big Brother.

TN: Moms For Liberty Lose--On A Technicality

Tennessee has one of the more punitive gag laws of the recent spate. It comes with the prospect of financial penalties for the school district and punishment for teachers all the way up to firing and loss of license. The state's law lists fourteen forbidden concepts. It was arguably an easy win for education conservatives who need to distract the public from all manner of education shenanigans, from a persistent failure to fund schools (even though the money is just sitting there) to a tendency to give away the store to privatizers and sweetheart deals with chums.

But once the law was passed, Tennessee's edition of the Moms For Liberty jumped in to provide the first court test by filing an 11-page complaint against the Wit and Wisdom book series. 

The complaint centers on four particular anchor texts, including "Martin Luther King Jr and the March on Washington, " "Ruby Bridges Goes To School : My True Story," "The Story of Ruby Bridges," and "Separate is Never Equal." It objects to the use of certain images and argues that certain themes, like "white people are4 bad" are prevalent. 

Lots of folks have picked out specific examples from the list of complaints, like the use of an actual photograph of children being firehosed, or Bridge's description of the white mob screaming at her and the signs they held. And looking at the various examples, I'm struck that mostly they are simply depictions of things that happened. If your desire is to have history taught strictly as facts, then these details they object to are doing just that. White folks held up those signs. That's a thing that happened.

I'm willing to buy an argument that elementary school is not a place for sophisticated explanations of systemic racism and discussions of the same texts that are brought up in college CRT courses. Despite the insistent dichotomization of this issue, there's a full continuum of possible positions from, on one end, requiring little white kids to admitted they are privileged oppressors and, on the other end, insisting that racism ended in 1865 and the only way to deal with the issue is pretend it doesn't exist. The M4L complaint seems to hew closer to the latter end. I don't often cite Eduwonk (Bellwether), but Andy Rotherham has been reading the books. The Bridges text includes this quote:

“Some people did not want a black child to go to the white school”

I guess you could argue that if you’re one of those people this might make you feel bad except it, you know, happened. It’s literally a recitation of the history of what happened to Ruby Bridges (and by extension to other kids in other communities). If you want to argue that you shouldn’t be telling little kids they’re complicit in white supremacy you will get a lot of support across racial lines. If you want to argue that you shouldn’t tell kids that not long ago schools were segregated, here’s why, here’s what changed, then you’re living in the past.

He also notes a quote that M4L doesn't mention. Bridges finishes her account “Now black children and white children can go to the same schools. I like to visit schools. I tell my story to children. I tell children that black people and white people can be friends. And most important I tell children to be kind to each other."

Bridges story and her journey always seem extra powerful to me because she is only a couple of years older than I am, and she always elicits that jolt you get when you look sideways across your own generation and think, "Holy shit, what a gulf there is between our experiences of the same country at the same time." 

Picking out some of the complaint's worst details does skip part of M4L's point, which is that these units are spread out over a great deal of time. If I understand their complaint, it's not just the quality of the materials, but the relentlessness of them. I don't sympathize. I can ask myself the question, "Would it bother me to have my four-year-olds experience this material?" It would not. I would be far more interested in how this or any material is actually used by the actual teacher in the actual classroom.

At any rate, the M4L complaint failed--but not on the merits. Instead, the state found that their complaint didn't land within the right time frame. Expect to see it refiled.

The Coming Pre-K Religion Battle

State-funded pre-schools are a spotty lot, ranging from well-funded to non-existent. The gaps are filled in by a variety of providers. My own small town is a good example; here you have a choice of a state-subsidized program, a small program run by the YMCA, and a program run by a local church. 

That's not at all unusual. People look for someone they can trust with the care of their littles, and for many that means churches. For churches, a pre-school or daycare makes a perfect "outreach" ministry, and they usually have physical facilities readily available, as well as a pool of people willing to work for volunteer-level wages.

When it comes to pre-school, there wall between church and state is more like a shower curtain, and it's about to be tested.

There are many, many things that the government can screw up with an initiative to create universal pre-K, from measuring results badly to simply trying to turn pre-K into the new first grade. And there really isn't anything in the track record of either party to suggest that screwing up will not be done.

But aside from that, the expansion of pre-K is going to unleash another fight over religion in schools. The current giant pile of money for pre-K and child care comes with a nondiscrimination requirement, and if we've learned anything about religious schools, it's that some of them are certain that discriminating against some people is a critical part of their religious mission. 

So conservative religious groups are already pushing back against the nondiscrimination clauses, because they might have to pass on that federal money if  they prefer to continue "to teach religious content, convene all-boys or all-girls programs, or give preferences in hiring or admissions to people of their religion." Most analysts say that the Build Back Better Act, in which this is buried, would not squelch actual religious instruction.

There are other challenges to small providers in the bill, including a requirement that they have seven years to have lead teachers with at least a bachelor's degree, and a requirement for wage levels. Also, this bill works by having states opt into receiving federal funds, which means that states can also say no. So even if Build Back Better passes the Senate, the fights about pre-K will move to the state level. Expect arguments about religious liberty to be part of that fight.

Charter Scandal Collection Now Available

Years ago there was a charter scandal website that eventually started to gather dust. In more recent years, the Network for Public Education has been collecting those various scandals under the tag #AnotherDayAnotherCharterScandal.

Now those various items are collected on the NPE website in a (partially) searchable collection. You can search by state or by category, as well as searching with terms of your own. 

The searchable portion of the collection goes back through 2019; there are more items collected, but not yet indexed on the database. If you suspect there are problems with your local charter folks, or you just want to browse the size and scope of charter issues, this will be a useful tool for you. 

Check it out right here.

A Reminder About The Uselessness Of Test Scores

 As we move through the latest stage of the pandemic in schools, we still get a lot of noise about how we Really Need to get those Big Standardized Test scores collected and crunched, because only then can we address Learning Loss or Pandemic Stumble or general Falling Behind. 

In doing so, we once again make the same old mistake of trying to use Big Standardized Test scores as a measure of future success (at its most extreme in the "students will suffer with years of lost earnings" think pieces).There is no particular reason to believe this is true. 

Let me remind you of this old graph.

In other words, a rich kid who drops out of high school has as much a chance of success as a poor kid who graduates from college. 

There are plenty of theories about why this is so. A Georgetown study concluded that early tests scores are less predictive of future success than socio-economic status. Those researchers point to an idea that echoes the issue of social capital that Robert Putnam explores in Our Kids-- that wealthier families have connections that both help locate opportunities for children (My kid really likes ponies, and I know a guy who runs a stable) as well as providing a safety net. As the Georgetown report puts it

When students from affluent families stumble, they have a softer landing and assistance getting back on track, while those in adverse environments are more likely to land on rocky ground and never recover.

The lead author of the report told CNBC

People with talent often don’t succeed. What we found in this study is that people with talent that come from disadvantaged households don’t do as well as people with very little talent from advantaged households.

The Georgetown report, like most such studies, is using test scores as a proxy for talent or smartitude. So what we're seeing here repeatedly is that tests are a lousy predictor of future earnings, life outcomes, etc. Which means that if we are concerned about those future outcomes for students, we need to look for better predictors.

There is a lot of legitimate concern right now over the fallout from pandemic. But obsessing over BS Test scores and throwing all our energy into trying to lift those scores is not the answer. The scariest part of that Georgetown report is in the last part of the sentence-- "those in adverse environments are more likely to land on rocky ground and never recover." If it's not too late to keep students from landing on rocky ground, we should try to prevent, and for those who have already landed, we should be helping them get back up. Hammering them to prep for the Test so they can Raise Those Scores is not the way.