Pages

Wednesday, February 3, 2016

USED Supports Unicorn Testing (With an Irony Saddle)

Acting Pretend Secretary of Education John King has offered further guidance as a follow-up to last year's Testing Action Plan, and it provides a slightly clearer picture of the imaginary tests that the department wants to see.

Here are the characteristics of the Big Testing Unicorn that King wants to see:

Worth taking: By "worth taking," King means aligned to the actual classroom, and requiring "the same kind of complex work students do in an effective classroom and the real world, and provide timely, actionable feedback." There are several things to parse here, not the least of which is "timely, actionable feedback" for whom, and for what purpose? Is King's ideal test a formative assessment, and if so, is the implication that it shouldn't be used for actions such as grading at all?

"Worth taking" is one of those chummy phrases that sounds like it means something until you are pinned between the rubber and the road trying to figure out what it means exactly. In my own classroom, I certainly have standards for whether or not an assessment is worth giving, but that decision rests heavily on my particular students, the particular subject matter, and the particular place we are in our journey, all of which also connects to how heavily weighted the grade is and if, in fact, there will be a grade at all.

But King's vision of a test aligned to both classroom and the real world is a bit mysterious and not very helpful.

High quality: This means we hit the full range of standards and "elicits complex student demonstrations of knowledge" and is supposed to measure both achievement and growth. That is a huge challenge, since complex constellations of skills and knowledge are not always easily comparable to each other. Your basketball-playing child got better at foul shots and dribbling, but worse at passing and footwork. She scores more points but is worse at teamwork. Is she a better player or not?

Time-limited: "States and districts must determine how to best balance instructional time and the need for high-quality assessments by considering whether each assessment serves a unique, essential role in ensuring all students are learning."

So, wait. The purpose of an assessment is to ensure that all students learn? How exactly does a test ensure learning? It can measure it, somewhat. But ensure it?  Do you guys still not get that testing is not teaching?

This appears to say, "Don't let testing eat up too much instructional time." Sure. Of course, really good testing eats up almost no instructional time at all. On this point, the Competency Based Learning folks are correct.

Fair: The assessments are supposed to "provide fair measures of what all students, including students with disabilities and English learners, are learning." So this uber-test will accurately assess all levels of ability, from the very basement to the educational penthouse. King doesn't have any idea of how to do this, but he does throw the word "robust" in here.

Fully transparent to students and parents: King lists every form of transparency except the one that matters-- showing exact item by item results that include te question, the answers, and an explanation of why the test manufacturer believes their answer is the correct one. What KIng wants to make transparent is the testing PR-- reasons for the test, source of the mandate for the test, broad ungranulated reports of results, what parents can do even though we won't tell them exactly how their child's test went.

BS Tests currently provide almost no useful information, primarily because the testing system is organized around protecting the intellectual property rights of the test manufacturers. Until we address that, King's call for transparency is empty nonsense.

Just one of multiple measures: No single assessment should decide anything important. I look forward to the feds telling some states that they are not allowed to hold third graders back because of results on the BS reading test.

Tied to improved learning: "In a well-designed testing strategy, assessment outcomes should be used not only to identify what students know, but also to inform and guide additional teaching, supports, and interventions." No kidding. You know what my unattainable unicorn is? A world in which powerful amateurs don't make a big deal out of telling me what I already know as if they just discovered it themselves.

And your saddle of irony: Every working teacher reading this or the original letter has had exactly the same thought-- BS Tests like the PARCC and SBA and all the rest of them absolutely fail this list. The BS Tests don't measure the full range of standards, don't require complex, higher-order responses, suck up far too much time, cannot measure the full range of student ability, are supremely opaque, are given way too much weight as single measures, and are useless as tools for improving instruction. They are, in fact, not worth taking at all. Under this test action plan, they should be the first to go.

More swell ideas.

The letter comes with a five-page PS, ideas from the feds about how to improve your testing picture, or at least ways to score money from the department for that alleged purpose.

You could audit your state tests. You could come up with cool data-management systems, because bad, useless data is always magically transformed when you run it through computer systems. You might train teachers more in "assessment literacy," because we am dummies who need to learn how to squint at the ugly tests in order to see their beauty. You could increase transparency, but you won't. You could increase the reliability and validity of the tests-- or at least check and see if they have any at all to start with.

Or you could just take a whole bunch of testing materials and smack yourself over the head with them. Any of these seem like viable options for running your own personal state-level unicorn farm.

No comments:

Post a Comment