Saturday, December 14

Opinion | Don’t Ditch Standardized Tests. Fix Them.

According to The New York Post’s analysis of New York State Education Department data, “Nearly 200,000 students — or one out of five — refused to sit for the state’s standardized reading and math exams for grades 3-8 administered in the spring” of 2023.

That number surprised me. There’s certainly some precedent for it, but I thought that the educational havoc wreaked by the Covid pandemic might have dampened the popularity of the opt-out movement — the tide of parents who’ve chosen to exempt their kids from state standardized testing. Apparently not.

Opt-out proponents argue, among other things, that “one-size-fits-all tests punish and discourage students who are already vulnerable” and “the tests themselves become the focus of education.” But after the major disruptions of 2020-22, I figured that even test-skeptical parents might reconsider the value of getting a straightforward accounting of learning loss that compared the progress of kids across schools and districts — to know whether their children are still playing catch-up post-pandemic.

I also thought that damning revelations in recent years about “balanced literacy” — a method focused on “developing a love of books and ensuring students understand the meaning of stories,” as the Times education reporter Sarah Mervosh described it — which was shown to be less effective than phonics (“systematic, sound-it-out instruction”), would make parents realize that standardized testing is an important part of developing the best curriculum possible.

Without testing, it would have been harder for the public to discover that balanced literacy doesn’t work very well. As Mervosh reported last year in an overview of the “revolt” against balanced literacy, “The push for reform picked up in 2019, when national reading scores showed significant improvement in just two places: Mississippi and Washington, D.C. Both had required more phonics.”

Opposition to standardized testing isn’t new, but it is especially prominent in New York, which has been described as a “national epicenter” of the opt-out movement. As The Times reported in 2015, resistance to standardized exams “began to snowball after 2013, the first year that the Common Core academic standards became the basis for judging student performance in New York.” In the past decade, an opt-out rate around 20 percent hasn’t been uncommon.

But despite reforms in New York and elsewhere, the broader conversation about standardized testing has stalled.

As Freddie deBoer, the author of “The Cult of Smart: How Our Broken Education System Perpetuates Social Injustice,” told me, “This is a situation that drives me crazy because it presumes that the two alternatives are hours of high-stakes testing for every student every year or no testing at all.”

It doesn’t have to be this way, he said, because there’s a third option: fixing the tests — state tests, in particular — to be more useful and effective. DeBoer said the National Assessment of Educational Progress, a test that all students don’t take every year, is often thought of as “the gold standard.” It is administered annually to a stratified, nationally representative sample of kids — typically around 2,500 per state, according to the National Center for Education Statistics.

Additionally, there are a handful of places currently working on testing reforms that address many of the issues that parents have with state-administered tests, including concerns that they’re too long, don’t really capture the depth of what students know and the results come in too late — sometimes not until summer break or the following school year — to provide usable information to classroom teachers.

Allie Pearce, a K-12 policy analyst at the left-leaning Center for American Progress, said, “Florida and Texas both recently launched through-year assessment pilot programs.” Instead of one long test taken in the spring of the school year, with results that don’t come in until that school year is over, through-year testing provides “three testing opportunities throughout the year” and then schools “make data available to educators,” sometimes as quickly as within a week, Pearce told me. That way, parents and educators have “near-immediate information on how students are doing.”

I talked to Iris Tian, deputy commissioner of analytics, assessment and reporting at the Texas Education Agency, about her state’s through-year assessment pilot, which is currently in its second year. Before students even sharpen a pencil, she said, “every question gets reviewed and approved by a group of current Texas teachers. We actually field-test each of the questions” to make sure it’s not biased. The assessment creators also get continual feedback from the schools to make sure that the information teachers and principals are getting from the tests is actually useful to them.

In the pilot, the three tests take place in the fall, winter and spring, and part of their research involves figuring out how short they can make the test while still being able to provide a valid statistical picture of where students are, Tian explained. They’re also redesigning the test so that it more closely mirrors “what’s happening in classrooms throughout the year” and is supportive of classroom instruction, rather than forcing teachers to teach to the test, a frequent complaint parents and teachers have about standardized testing.

One way the test is being modified? Fewer multiple-choice questions, Tian said. “We all know if all you’re doing in classrooms day in and day out during the school year is giving multiple choice questions, that’s not how kids learn.”

Having quality information about how America’s children are learning is critical, particularly since the educational gaps between the haves and the have-nots were exacerbated by the pandemic. As Tom Kane and Sean Reardon explained in a guest essay for Opinion in May:

In 2019, the typical student in the poorest 10 percent of districts scored one and a half years behind the national average for his or her year — and almost four years behind students in the richest 10 percent of districts — in both math and reading.

By 2022, the typical student in the poorest districts had lost three-quarters of a year in math, more than double the decline of students in the richest districts. The declines in reading scores were half as large as in math and were similarly much larger in poor districts than rich districts. The pandemic left students in low-income and predominantly minority communities even further behind their peers in richer, whiter districts than they were.

Without standardized testing, we won’t know where to put the most resources, or what the contours of the problems students face even look like. Getting rid of widespread assessments won’t help the most vulnerable children, it will only leave us without knowledge about how best to support them.