Missionlocal

SF school closure survey befuddled data scientists, lawyers and everyone else

A.Smith43 min ago
When the San Francisco Unified School District wants to get parents' attention, it can. We receive multiple emails in multiple languages. Text messages buzz on our phones. Landlines ring and robocalls drone on about an important pending announcement. Public school parent Ed Parillon compares it to the episode of "The Office" in which Ryan creates a startup called WUPHF that links everyone's means of communication, resulting in all the fax machines and phones and printers going off at once.

He has a point.

Last weekend, when the district abruptly delayed its scheduled school closures announcement , none of this happened. Superintendent Matt Wayne put a statement on the district's website, which is a bit different than simultaneously sending out a text, email, robocall and WUPHF. Well into last week, teachers complained that nobody had sent them an official district communique announcing the abrupt delay of the closure announcement — and, needless to say, parents didn't get an email either.

"When the SFUSD wants to send a message, it's generally irritating and over the top," says Parillon. "But the one time there's a message about something parents really want to know, they did nothing of the sort."

He has a point here, too.

A piddling percentage of San Franciscans have kids, and no city in California has a higher percentage of its children in private schools — so strange and terrible things are allowed to fester in the school district for years.

And fester they do — until the excrement hits the air conditioning. The district's yearslong inability to escape near-fiscal insolvency , a disastrous, $40 million-plus poured into a payroll system that made teachers' lives hell and, finally, the inability to formulate a coherent plan to close schools on the agreed-upon timeline led to frustrations bubbling over among current and former members of the school board.

It's the delay in the closures plan that seems to have really torn it. It's not just that a deadline was missed and families and school staff were put on pins and needles for weeks — though that ain't great. Rather, it's that school board members described the long-in-the-making school closure plan's methodology and the strategy to implement it as so lacking that a delay was deemed necessary.

Perhaps more than a mere delay: Last week Board President Matt Alexander and Vice President Lisa Weissman-Ward quietly reached out to the mayor's office for help, and a municipal Seal Team Six was assembled to aid the troubled district . Wayne, for now, kept his job — but gained a group of city execs to look over his shoulder while he does it. This was deemed preferable by the school board to sacking Wayne on the spot at a Sunday emergency meeting, as there is no SFUSD Brock Purdy to leap off the bench and take over .

So both the timing and the extent of the pending school closures are to be determined. That's a hard conversation for the next couple of days. But — there is a list out there. We're told it's between 10 to 14 schools. There have been all kinds of cloak-and-dagger insinuations about how the list will be disseminated piecemeal, as the district readies staff and parents and kids for the bad news.

Thus far, the district has done anything it can to avoid having forthright and straightforward discussions with the families and staff at the schools it deems are struggling and figure out how to move forward. The district has also done anything it can to avoid simply showing its hand and letting concerned parties know what's what — and having arguments on the merits.

We do mean anything. If you're a public school parent — and you check your email diligently and have the time and inclination to respond to district missives — you may have filled out a survey regarding school closures.

Straightforward it was not : Parents were instructed to imagine they had 12 coins, and made to disseminate them in baskets marked "equity," "access" and "excellence." Separate and apart from pitting things everyone should want against each other, it is markedly unclear how such data about anodyne concepts could be satisfactorily distilled into a formula to close brick-and-mortar schools.

But we are informed that district representatives were seriously planning to tell aggrieved parents that their kids' school was closing because of tabulations based on imaginary baskets of hearts, moons, stars and clovers. Perhaps a wizard could make sense of this, but we were unable to locate any wizards who doubled as public school parents.

We did, however, find lawyers and Ph.D data scientists — but they said the survey confused the hell out of them, too.

"I am an attorney who works in software copyright issues. I spend a lot of time thinking through 'how do we slice and dice intellectual property rights? ... structuring and dicing them is pretty complex and deliberate. That is what I do all day," said public elementary school parent Luis Villa.

But Villa was no match for the baskets of equity, excellence and access.

"It was just very confusing. This felt like a logic puzzle. It was completely opaque," he said. "I have some sympathy in that they are trying to build a thing that will not be gamed by parents whose sole concern is 'I gotta save my kid's school.' They want something more general. But it's not useful to just confuse everyone."

Villa, a Cuban American, is a native English speaker. He graduated from Duke and earned his J.D. from Columbia. As a lawyer, he specializes in parsing arcane texts. And the SFUSD survey flummoxed him. As he was struggling to fill it out, he wondered how the monolingual Spanish speaking parents of his child's school classmates were faring.

The answer: Not great.

Laura Padilla is on her kids' school's English Learning Advisory Committee, and volunteered to help monolingual Spanish-speaking parents with the surveys. When she told them to choose between "access" and "equity" and "excellence," they were confused — "they kept telling me that all of these are important." She was then asked to define "excellence" — which is harder than you might think. Who doesn't want "excellence"?

This turned out to be a tortuous process. Padilla says she could only walk two parents through a survey in a half-hour.

The "excellence" situation is a problem Ed Parillon warned the district about all the way back in May. Defining "excellence" isn't only difficult — it's inseparably tied in to other factors.

Parillon, who serves on the SFUSD District Advisory Committee, wrote to the district that "the measures are of dubious value and largely a reflection of family incomes, like test scores, but they also have nothing to say about why a school should or should not stay open. ... test scores surely travel with students and families."

The survey, Parillon concludes, was "essentially manufacturing consent for closing schools."

So the criteria here were opaque, confusing and questionably suitable. That's a problem. But another problem was that, as you'd expect, the parents who'd take the time to fill out an online-only, voluntary survey skewed disproportionately white, educated, and financially stable.

The district weighted for that, giving more emphasis to responses from Latinx parents like Villa and Padilla or Black parents like Parillon. But this doesn't really work — because while the results were corrected for race, they were not corrected for income or educational attainment. This created a homogenizing effect for Black and Latinx participants.

"Look, I'm a Black parent," says Parillon, who used to develop credit-risk models and now works in affordable development. "But there is a certain type of parent who has the access and time to fill out your stupid survey. You are not getting a real cross-section."

You don't need a Ph.D data scientist to tell you that. But we found one — who's also an SFUSD parent of two young children. And he told us that.

"The self-selection of the survey cannot be representative," said Philippe Marchand, who earned his Ph.D from UC Berkeley in environmental science and has spent years as a statistician and data scientist and teacher.

"Correcting the factor of race cannot correct the self-selection of the sample and the overrepresented higher level of education and also income. It's just whoever shows up. They shouldn't make it look more scientific than it really is."

Marchand says he hopes the district doesn't try to use this survey to rationalize its actions. "The process," he says, "was really unfortunately designed."

Well, you don't need a Ph.D data scientist to tell you that, either. But he did.

And he — and all the rest of us — are on pins and needles, waiting for the next email in multiple languages, text messages buzzing on our phones or landlines ringing with robocalls. WUPHF.

0 Comments
0