Wednesday, April 29, 2015

Merryl Tisch Shows How Test Supporters Get It Wrong

New York Chancellor of the Board of Regents Meryl Tisch stopped by All In with Chris Hayes to avoid answering some pointed questions about high stakes testing and the opt out movement in New York. She had the additional disadvantage of sitting beside Diane Ravitch, who did answer questions and made Tisch look even slipperier by comparison, but I think Tisch's appearance is a quick, capsuled look at what promoters of high stakes testing get wrong.

After opening with some background (Atlanta convictions, rising parent opposition, left-and-right wing hatred for Big Standardized Tests), Hayes notes that New York's opt out numbers are huger than ever and turns to Tisch.
Hayes: When you see the reports of opting out, including the high strong numbers from some areas, do you think "People are crazy" or "We are doing something wrong"?

Tisch does not think people are crazy (phew!) and believes that people "should act in what they perceive to be the best interests of their children." That's an important construction because like many BS Test fans, Tisch is also a charter school fan, which puts her in the awkward position of believing that parents should opt out of public school, but not public school testing. Choice is only okay sometimes.

So why the push back? "Perhaps we have not been clear enough in describing the intent of the test." So, opt out is a PR problem, because if people disagree with us, it could only be because they don't understand how right we are. So what is the intent of the test?

"The intent of the test is to give a snapshot of performance and allow parents to know where their children are at any given point in their educational career as compared to their peers."

And there's your first problem, because that doesn't even make sense. "Snapshot" and "at any given point in time" do not go together. I can't see how my child is doing at any point in time because I only have a snapshot from one particular point in time.

Tisch moves on immediately to asserting that income inequality is directly tied to the achievement gap (which is actually the BS Test score gap) for our poor students, and she starts waving the Wait Let Me Speak hand at Hayes because he is completely ready to call her on that piece of baloney, so she squeezes in that poor students can't make more money unless they have access to high quality education. Hayes calls her on her correlation-causation fallacy, but I'd like to call her on her fallacious equating of high quality education and high stakes snapshot testing. What does taking the BS Test have to do with access to high quality education?

The reformster answer (which Tisch doesn't get to) is that BS Test results allow us to target the students who are struggling. The problem here is that 1) we already know where they are and 2) after years of targeting them with BS Testing, we have yet to actually get them additional resources to help them do better.

Ravitch gets her turn and uses it to point out that tests are not vaccinations and these tests are not useful because the results provide no useful information. "There is no diagnostic value to the test," somehow prompts Tisch to smirk, like she has caught the help trying to act like they know how caviar really tastes. Hayes notes that Tisch clearly has something she wants to share with the rest of the class, and she unveils Test Purpose #2.

The tests are a diagnostic tool for curriculum and instruction development on the state level, and a way of making sure the taxpayers get their money's worth.

In other words, a completely different purpose for the tests than the one she offered about two minutes earlier. It's now a snapshot of how our children, schools, and systems are doing-- for the taxpayers. So that business about info for the parents was, what-- just spitballing? Because if this is the actual purpose of the test then 1) what's wrong with the NAEP and 2) why is it necessary to test every child every year?

Hayes points out that Tisch gave a non-response to the observation that the tests are not diagnostically useful for students, parents or children, and she insists that she be allowed to insert a non-response to that point. When parents opt out it messes things up. Also, she was in a doctor's office where a parent wanted to compare their child to a growth chart. Like the vaccination analogy, this is bogus for many reasons. I'll just pick one: When I weigh my child, I get a full picture of how much my child weighs, but when my child takes a BS Test and I get just the score, I get only the tiniest sliver of a slice of how well my student is doing in school.

As for the diagnostic value of the tests, Tisch asserts (with her asserty hand waving before her) that school districts report "all the time" that they make decisions about curriculum around the test results. Which certainly proves that schools will teach to the test as best they can, particularly when threatened with punitive responses to the results. This does not prove that either test results or the following curriculum adjustments serve the educational interests of the students. She also says words about how the ability to glean specific info from these tests is really important, which is not remotely the same as proving that it can actually be done.

Hayes asks Ravitch if there's a right way to do test-driven accountability or if it's just the wrong tree at which to be woofing. "Wrong tree," says Ravitch. You can't do the wrong thing the right way. The model is wrong. We are the most overtested nation in the world.

What would Tisch like to say to parents?

Tisch would like parents to understand that this is all the union's fault, and that if teacher evals hadn't been linked to the tests, they would all be testing away happily. Children have just been trapped in a labor dispute between the governor and teacher.

In about six minutes, Tisch manages to showcase a full range of pro-test arguments, all specious.
If the goal is to give parents information about their student, why does the test return so little data? And what difference does it make if other students opt out?

If the goal is to give teachers and schools actionable data to inform instruction, why return so little data, so late?

If the goal is to give taxpayers and policymakers feedback about how the system is doing, testing every child every year is by far the least cost-effective method.

If the goal is to identify and diagnose troubled schools for intervention, why don't bad scores trigger a release of additional resources for the identified school?

And why do pro-testers never, ever provide solid data about how well the tests actually measure any of the things they supposedly measure?

Tisch can blame the opt out movement on the union and politics all she wants; the reality on the ground is that more and more parents have had enough. The BS Test boosters are going to need better talking points.

Originally posted at View from the Cheap Seats


  1. Unfortunately, a big part of the problem is that we as a nation are horribly information-illiterate. We are never taught how to critically evaluate information. We never learn how to assess research design. We receive no guidance on how to determine the relative strength of a particular source of information (I.e., does the source have a particular agenda like our friends at the Gates Foundation?). In fact this past semester none of my students (college) could even explain validity or reliability.

    This results in a communication/ cultural breakdown similar to your point about new coverage. We are bombarded with information, but we've been trained to process that information in sound bites and tweets. Graphs, statistics, and charts fit well in that format. Subtle discourse on the limitations of the data and the distinction between correlation and causation don't. As such, what information are people more likely to ingest?

  2. This comment has been removed by the author.