By Joshua Benton
You probably didn’t notice, but Texas schools just celebrated a big holiday.
I doubt anyone brought cupcakes to class, but Oct. 27 looms large in principals’ offices and the halls of administration buildings.
That’s because the last Friday in October is New Kids Stop Mattering Day – the day after which any new students enrolling at your school won’t be counted in next spring’s TAKS scores.
It’s a holiday that makes life easier for teachers and principals wishing for higher test scores. But it also hurts thousands of Texas kids.
Jennifer Booher-Jennings is a Columbia graduate student whose research I’ve written about before. She studies how poorly constructed testing systems can leave some kids without the attention they deserve.
Last year I wrote in this space about her study of a Texas elementary school, where teachers gave enormous help to kids at risk of falling just a few points short of passing TAKS. That’s good.
But that extra help came at the expense of weaker kids – kids who might not pass even with more tutoring and teacher time invested. They were being written off as hopeless – at age 8.
That’s bad. It’s bad because it ignores what would be best for kids – helping the weakest at least as much those on the bubble – and instead does what’s best for the adults. Namely, it boosts a school’s passing rate by going after only the low-hanging fruit.
While working on that study, she noticed something I’ve seen often in Texas: Teachers were very aware of whether kids’ TAKS scores would count against them. They called kids “accountables” and “unaccountables.”
In Texas, the “unaccountables” come primarily from two groups. First are kids in special education – some get counted, but many others don’t. Second are kids who switch schools during the year – the ones who arrive after New Kids Stop Mattering Day.
One teacher at the school Ms. Booher-Jennings studied had mandatory tutoring for all the “accountables” three days a week. The 10 “unaccountables” in her class? They were, in her words, “put on the back burner.”
Here’s a transcript from one grade-level planning meeting Ms. Booher-Jennings sat in on:
Teacher 1: “This kid is not accountable. Do I even need to worry about him?” Teacher 2: “No … don’t worry about him.”
If teachers know that some kids can be safely ignored – given all the test pressures they already deal with – some are going to redirect their attention elsewhere.
She and a colleague, Andrew Beveridge, have authored a new study that looks at the impact of excluding students from the accountability system.
The numbers excluded for special education are bigger. But in some cases, the late-arrival totals are significant, too. For example, in 2005, Dallas’ Pease Elementary excluded 20 percent of its kids because of their arrivals after the October deadline. So did City Park Elementary.
And in many cases, the kids being excluded for arriving late are disproportionately poor and minority. At Seagoville Elementary, 30.8 percent of black students weren’t counted. At Daniel Webster Elementary, 23.8 percent of Hispanics were eliminated.
All those exclusions have an impact on test scores. In a paper to be released in a few months, Ms. Booher-Jennings and Mr. Beveridge analyzed test scores for all Houston ISD schools. They found that if every Houston student had been tested – and all their scores counted – the district’s performance would have plunged.
Specifically, 37.7 percent of Houston schools would have fallen to a lower state rating because of lower reading scores. Nearly 28 percent of schools would have fallen because of math scores.
I don’t think all teachers give more attention to the “accountables.” And I don’t think the ones who do are evil, or even ill-intentioned. They’re just responding rationally to a system that has the incentives wrong.
Think about your own job. Imagine your boss told you that you were going to be evaluated only on your work with 80 percent of your clients, not on the other 20. Can you be sure you wouldn’t respond accordingly?
There are legitimate fairness issues about counting all kids. There are certainly some kids – the severely mentally disabled, for instance – who shouldn’t be counted. And it may not be fair for a school to be blamed for the poor performance of a child who enrolls 48 hours before test day.
But the current system leaves too much room for kids to be ignored.
This spring, Texas officials tried to gain even more wiggle room – asking the federal government for permission to ignore any kid who wasn’t enrolled in the same school district for two consecutive years. That would have eliminated another 10 percent of kids from the ranks of the counted. The feds said no.
This isn’t an attack on testing. If anything, it’s a validation. It’s proof that testing can be a powerful tool to improve learning. But that means a testing system should count as many kids as it can – not ignore the weakest.
Here’s the question. Is Texas’ testing system designed to give struggling kids the attention they need? Or is it designed to make the adults look good?