The Personalization Paradox

Remember when standardization was king of the K-12 world.

We’ve all heard the stories. The one boasting that on any given day, the Education Minister in France knew what was being taught in each school in the country.  If only we could be more like France. The one reminiscing wistfully about the time when someone travelling from coast to coast in the US could stop in high school science classes along the way and never miss a lesson.

If we could just get standardization right, what a wonderful world it would be.

Those halcyon days of imagining schools as factories, tightly-managed cogs in the military industrial complex, are behind us now – at least on paper. I’m not really sure what goes on in schools these days, but I’m certain that there is no more boasting about or longing for the quiet efficiency that comes with standardization.

Not that there’s necessarily anything wrong with that.  In fact, it’s probably a good thing.

My goal here today is not to bemoan personalization or cry over the loss of standardization in public education or even in assessment (as distinct from standardization in some aspects of testing).

The truth of the matter is that the notion that public education in the United States was standardized was always equal parts myth and aspiration.  The myth – that schools were factories churning out widgets on a conveyor belt (note: the term widgetwas famously used in reference to teachers not students).

Yes, for much of the 20^th century, public education was based on the factory model, but it wasn’t a factory that produced only one product. At its best, it was a system in which the comprehensive high school prepared students for one of a variety of postsecondary choices available to them. At its worst, it was a system designed to early on separate students, place them onto irreversible tracks, segregating them, and cutting off options to particular groups and individuals.

In response to that worst case scenario, in the last quarter of the century, standardization became the aspiration.

After Goals 2000, the Core 40 and Algebra II for All, our last great grasp at standardization and the efficiency of a national curriculum and a national assessment based on national standards was the Common Core and the Race to the Top Assessments.

But even if adequate upfront effort had been devoted to designing curriculum and preparing teachers to implement the CCSS, the pendulum had already begun to swing away from standardization, away from one-size-fits-all education.

Even if the roads taken by the two assessment consortia had not diverged so sharply, I’m confident that those walking through that yellow wood that in the spring of 2015 still would have bypassed the consortia’s’ standardized high school college-and-career readiness tests in favor of personal pathways.

In public education, the pendulum is now paused on personalization.

This Time It’s Personal

The future of education and of learning is personalization.

But what aspects of public education are to be personalized? Outcomes? Inputs, including instruction? Pathways to either the same or different destinations? All or some combination of inputs and outcomes?

And what level of personalization is ideal, is feasible, is the goal?

There doesn’t appear to be a clear answer yet to that question. We’re not quite sure where this push for personalization is headed, but there does seem to be widespread agreement that top-down, standardized tests and testing is an impediment to getting there.

Test is a four-letter word, and I guess, standardized, coming in at 12 letters, is three times as bad.

Standardized tests, like Aaron Burr, have become the villain in our history. That designation may or may not be deserved, but that’s not the point.

Yes, standardized tests and testing need to adapt to meet the changing demands of public schools. We knew that 35 years ago.

But the idea that we can attain personalized public education without the standardized collection of standardized data is ludicrous.

You wanted actionable information, right?

Be Careful What You Wish For

The irony is that a serious, legitimate move toward personalization in public education (and not simply performative hand-waving) will require more standardized data than we have ever collected in the past, likely even more than we can imagine collecting now.

The need for data is inversely proportional to the level of personalization.

The simple fact is that our historical model of public education didn’t require much decision-making, and therefore, didn’t require much data.

When schools were designed for one group of students with one outcome in mind and curriculum was determined by a committee of seven or ten, we didn’t need much data at all.

When we added a few more outcomes, we needed a little more data.

When we started to worry about whether all groups of students were achieving desired outcomes, we again needed a little more data.

When we started to worry about whether all groups of students had equal opportunity to achieve the desired outcomes, once again, a little more data.

In contrast, when we start thinking about personalized outcomes and inputs for individual students within complex contexts, the need for data grows exponentially.

Personalization requires decision-making on an almost continuous basis, decision-making which requires some level of generalization based on data – lots and lots of standardized data, standardized norm-referenced data.

Personalization without data is at best, wild-ass guessing, and more often than not, absolute chaos.

Where do we fit in?

For those of us involved in assessment to help avoid the chaos our task will be to provide the high-quality data needed to support personalization.

We know how to do norm-referenced data.

Of course, the definition of the norm groups (or referent groups, if you prefer) will be more finely-grained than ever before, which will make collecting and analyzing data even more complicated. (read Mislevy, 2018)

And the data will be much more complex and multidimensional than responses to multiple-choice items and writing prompts (read von Davier et al., 2021)

Plus, that data will have to be relevant in real-time (i.e., not based on end-of-year norms) and available in real-time. (we don’t have a book for this one yet)

And to make ourselves useful, we will have to get much better in the ways that we communicate with stakeholders, providing them with the data and information they need and not simply the data we have.

Is it going to be challenging? Sure.

But it’s going to be much so more fun than figuring out whether kids are proficient on grade-level standards.

So, let’s not get too wrapped up in trying to determine every more precisely how well each and every individual third-grade student can read and apply mathematics. Personalization is about so much more than that.

Image by Jensen Art Co from Pixabay