Sunday, June 28, 2015

Threading Cake Testing Needles (Part III)

I've been enjoying a dialogue centering on some postings by Nate Bowling. Bowling is more reformy than I, but his work is interesting because A) he is an actual teacher and B) he does not claim that he didn't know how to do his job until Common Core. There aren't that many actual classroom teachers out there articulating the reformster case well, but Bowling seems to be continuing a sincere and teacherly search for some answers and solutions around the issues of testing, and I have appreciated watching part of his journey.

You can find earlier pieces of this conversation here and here and here and here.

Part III is up and while we're still torturing metaphors, at least we're still eating cake and threading needles and not eating cake with needles in it.

Bowling opens with observations about the effects of polarization, in particular how it becomes an obstacle to useful discussion. I have some ideas about how that polarization became welded onto the education debates (when you attack teachers and the work they've invested themselves in, it's hard for them not to take it personally), but I absolutely agree that any time you assume that people who disagree with you must be either evil or stupid (or both), you're not going to accomplish anything worthwhile. As much as hugely disagree with many folks on the other sides of the Great Education Debate, I have found that almost every last one of them is amenable to civilized conversation. It's like they're actual humans, or something!

Bowling presents that view as the context for trying to further clarify the issues that he laid out in the first part of the series. His thoughts are worth looking at, even if I disagree with some of them.

Problem 1: Standardized testing comes with huge costs of both money and instructional time, and it gives no real useful information in return. I think that's dead-on.

Problem 2: Testing as a civil rights issue. Reading Bowling's explanation of this issue (which leans on writing by Chris Stewart) suggests to me that maybe what we're talking about here is a new audience for student progress reports.

I've often mocked the notion that either parents or teachers need standardized test results (which are hugely limited in scope, in depth and in detail) to know how their students are doing. But the civil rights testing argument seems to include the notion that communities and leaders need hard data about school failure in order to create political and social pressure to right wrongs and close gaps.

There are problems with using tests for that purpose. One is that the tests are still bad measures. Bowling writes

One of the most frequently raised arguments against testing that I come across is that testing is not an accurate measure of "the whole child," or their “real worth.” I agree, but no one (no one worth listening to anyway) is arguing that it is. 

I disagree with his second sentence. Every time someone makes a statement about student achievement or teacher effectiveness or whether a school is swell or not, all they are talking about is test scores. "Student achievement" as currently used literally means nothing but "student test scores," and so test scores have become a proxy for every kind of measure that can be imagined. And that can't help but narrow the view of what schools are supposed to do.

Nor is there much useful data. Bowling notes that "Tammi got a C+" isn't for some folks-- but it has more granularity than the Big Standardized Test reports which just tell us which of four possible grades that student earned.

I could look past that, maybe a little, if low test scores were used to prove that Lowscore City Schools were not getting sufficient support and resources from state and federal government. But that's not how the story plays out. Instead, we see two things happen over and over again.

One is that the state sweeps in and cancels democracy for the community. Instead of coming in, sitting down with community leaders, and finding out what resources they need to support their local vision for their own community, from Newark to Philly to Chicago to Detroit, over and over, the state comes in and says, "Clearly you brown/black/poor folks can't be trusted to run things, so we're going to suspend democracy, silence your voices, and tell you what you should have."

The other is the building of tiny lifeboats. In the name of rescuing students from failing schools, charter systems are created that allow a small percentage of students to escape the failing schools. Meanwhile, all the other students are still in the troubled schools-- which are now getting fewer resources rather than assistance.

When the citizens and students of Newark are in the streets repeatedly-- and fruitlessly-- demanding to have a voice in their own community's schools, that doesn't look like a civil rights win to me.

Now, should we have some means of keeping relentless and forceful pressure on politicians to make sure that all communities are well-served and absolutely unignorable? We should. I don't believe for a milisecond that politicians do not know which the communities need assistance, but if we need to be able to generate charts and graphs to hold their feet to the fire, then let's play that game. But the current wave of test-based accountability-- which we've been trying for over a decade-- is failing to do the job. We have never really had; a system for generating data for the audience of politicians and policymakers, and we need to go back to the drawing board to come up with the right instrument for that task.

Bowling goes on to offer solutions, with the caveat that these are Washington State-based ideas, and your mileage may vary.

No test scores in teaching evals. Well, yes. I'm not sure what will finally kill this, since there is not a small continent's worth of VAM debunkery out there. I suspect that this won't crumble until we have enough local stories of how Beloved Mrs. Teachswell, known by one and all to be wonderful in the classroom, has been judged Terrible by the state evaluation system. Right now the system is so crazypants that folks literally refuse to believe me when I explain that the shop teachers evaluation is partly based on the test scores of students he's never even met; it's so bizarre that they are sure "that can't be right."

Eliminate redundant exams and shorten existing ones. And, though this implied by the rest of his paragraph, be damn sure you can explain why failing Exam X should, all by itself, keep a child from graduating or moving on to the next grade.

I mean come on, the test to determine whether you are proficient at any single grade level should not be longer than a Bar Exam. We can create assessments, linked to the CCSS, NGSS, (insert your own SS) that indicate where a student is on a continuum from way below grade level.These assessments don’t have to be insanely expensive, overly complicated and should be able to be completed in one or two class periods, rather than the five days (2xs) it took to administer each SBAC (math and ELA), at many schools this year.

I'd go further. Why does it need to be a test?

Shift power from testing companies to educators.

"If tomorrow I was given the power, I’d commission a group of teacher leaders to create the exams for my state. I would shift the duty of designing state exams from unknown figures at various testing companies to noted and notable educators." Yeah, I'll back that. The problem is money. All of reform is a shift of power from educators to people who would like to make money from education. How we push Pearson back out of the BS Test manufacturing market when on any given day we are in our classrooms and they are lobbying in capitals is a mystery to me.

I'm not saying we shouldn't try. But the whole premise of BS Testing is that the grade given by a classroom teacher can't be trusted, and I don't know how we can possibly turn that political tide. Lord knows we're trying to get the message across-- but very few people with actual power are listening.

One other advantage of teacher-created testing? Students would be more likely to take the testing seriously. And teachers might get something we could actually use. It really would be a vast improvement in many ways, but I don't know how we sell it.

I appreciate Bowling's resolve to see this conversation through and to examine the positions honestly. One other problem with polarization is that it can give you blinders-- you only allow certain conclusions to be reachable, and that, of course, colors how you view everything. It takes some nerve and patience to track the ideas without trying to force your way to a particular conclusion, and I appreciate that Bowling appears to be doing so. Thanks for the cake, Nate.

1 comment:

  1. In my district, a series of tests has been created to "prove" that social studies teachers are "proficient." The tests are TERRIBLE. There is no quality control, and the test questions are poorly written, with lousy maps and charts, asking pointless questions. Furthermore, some of the answers are wrong, and there are so many grammar errors that I would be embarrassed to give them to my students. Teacher created standardized tests are STILL standardized tests. They are not the panacea some people seem to thing they are.