Friday, January 8, 2016

Metrics and Behaviorism

Mark Garrison is a helluva guy. He lists scholar, activist, evaluator and artist as his businesses, and lists two books and three albums among his achievements. He has clearly put a whole lot of thought into the current state of education policy, with a particular sharp eye for the problems of Trying To Measure Stuff. Garrison says many things I agree with, but he says it all smarter.

In "Metric Morality," Garrison addresses the issue of just what can be measured, and why the current educational measurements are doomed to fail. I love this stuff, because it cuts to why the whole concept of educational metrics is, as he puts it, "fraudulent."

"They confound properties of individuals, individual schools and individual school systems with the relations those individuals, individual schools and individual school systems have with their social contexts." In other words, they talk about qualities like "wetness" as being properties of the individual and not as they are related to whether one is standing inside, under an umbrella, or outside naked in a monsoon.

They "follow the long discredited practice of defining the object of measurement 'operationally'; that is, things and phenomena are defined by how they are 'measured'." Garrison's example is perfect-- the definition of "intelligence" is the ability to do well on an intelligence test. Yay, tautology.

They use the "flawed definition" of measurement which is the process of assigning numbers according to some rule. This, Garrison says, ties directly to the notion that everything that exists must exist in some amount. "This would mean, for example, that we accept the proposition that humans exist in their degree of human-ness. Some of us are more human than others. Thankfully, the testers will select the chosen ones!"

They confuse ranking with measurement. Garrison seems to attribute this to confusion; I'm inclined to think reformsters do it on purpose. As noted a gazzillion times-- the fact that a school falls in that fatal bottom 5% does not mean it's a bad school-- just that it is ranked below the other 95%. The best rock band in Uraguay may still be a lousy rock band. Rank is not measurement.

Also, cut scores are baloney, and nobody even pretends to understand what "validity" and "reliability" mean -- or used to mean-- in the testing world.

Garrison considers the way in which leaders like Elia of New York have equated allegiance to the testing regime with morality and ethic. He might as easily noted Pennsylvania, where our test administration instructions are called "Ethical Standards of Test Administration," as if our role in overseeing the holy test is one founded on ethical principles rather than compliance to power and money.

But this takes him right to our old buddy, behaviorism.

In "The Behaviorist Origin of Close Reading," Garrison travels back to the twenties and the roots of the original close reading. Garrison sides with the critics who assert that close reading founding father I. A. Richards was directly tied to the behaviorist work of John Watson.

Some of us were learning all about behaviorism in teacher school back in the seventies, but it's not a popular term these days, so we can talk about personalized learning and game-based learning without ever noting the similarity to a big, computerized Skinner box. And just as sure as TSWBAT means "the student will be able to display an observable behavior that will be the entire basis for measuring the educational results," behaviorism has made a comeback.

Common Core reading is all about observable skills, and what we cannot see or measure does not count. One of Garrison's very best lines is actually a subheading from this essay:

Behaviorism: Yearning for Skill Without Consciousness

Garrison explains the disregard for consciousness in behaviorism:

A key tenet is this: Behaviorists have in common disregard for or denial of human consciousness. Because consciousness is not something one “does”, it is not “observable”; its existence or importance is denied in favor of fixing attention on behavior itself. 

Or as David Coleman put it, "Nobody gives a shit what you think or feel." Or, he might have added, what you know. Just what you can do.

The behaviorists didn't much care for words like "consciousness" or "intent." They can't be seen or measured, and as I heard repeatedly in more than one college course, you don't need them to explain human behavior. Somewhere in my house, I still have my copy of Skinner's Beyond Freedom and Dignity(1971) in which Skinner makes the case that our devotion to these tired old notions of free will and individual moral agency are keeping us from using scientific techniques like operant conditioning to modify behavior so that we can have a happier and more orderly society. Skinner's idea was a form of determinism, the notion that human behavior was just a collection of behavioral tics created by conditioning, and with the proper conditioning, we could get people to have the proper tics.

"Skill" is a nice term for a particular tic, a tic that humans are trained to perform when given the appropriate stimulus.

So. Is any of this sounding familiar?

Close Reading, says Garrison, adopted much of the behaviorist creed. First, treat the text as a behavior. It is, in fact, a perfect example, since the page has no consciousness or intent, and we are instructed to deliberately ignore the author's intent. The reader's response is a reaction to the behavior-- the behavior that the behavior elicits. In this model, there is nothing to analyze (and certainly nothing to understand, since understanding is all about consciousness and intent and moral agency) except how exactly the behavior elicited the response. It's simply figuring out what about the hot stove made you holler and yank your hand away; there is nothing more to reading Shakespeare or Morrison than setting your hand on a stove and seeing what you do next.

As Garrison puts it:

For Richards, “all mental events — including literature — occur in the course of processes of adaptation somewhere between stimulus and response”. Thus we have the basis for a method that renders the skill of reading necessarily devoid of consciousness.

My friends and I were not fans of behaviorism in college. It has its initial charm-- the surprise of a simple and clear explanation for much messiness of human existence-- but it's just so dehumanizing, cold, and ethically empty. It certainly has its place; there is much human behavior that becomes more comprehensible through the behaviorist lens. Ultimately it's too inhuman and inhumane, while suffering from a serious "if your only tool is a hammer, everything looks like a nail" problem. I could never understand why behaviorists didn't just say, "We can explain some aspects of human behavior, but beyond certain points we don't have enough tools in our box." (And don't even get me started on Walden Two, Skinner's 1948 big middle finger to Thoreau and the very concept of a human soul).

Garrison has done a good job of illuminating the connection between bad metrics, bad behaviorism, and the hollow futility of some current reformster ideals. And he affirms what many folks had already figured out for themselves-- that much of the current reformster regime feels counter-intuitive and anti-human because it is.


  1. TSWBAT to a little more than they could do yesterday!

  2. I've thought from the beginning that reformers treat the process of teaching like Skinner's idea of the mind as a "black box" that we can't see inside, therefore don't understand, therefore should just ignore in favor of observable actions that can be measured. But this is because the reformers aren't teachers and don't understand the (complex) process of teaching and learning, so for them it IS like a "black (incomprehensible) box".

    (I also think Miami University of Oxford must have been more in the fore-front of thinking than I realized when I was there, because in my 1972-73 Ed Psych class we learned about Skinner, but it was made clear that the idea of behaviorism was limited in scope and passé in education theory, and that Jerome Bruner and cognitive learning theory was the way to go. Bruner has always served me well.)

    I think Garrison is very perceptive when he says, "Behaviorism instituted a “new” sense of morality rooted in the experimenter’s control of the environment to induce the 'right behaviors'...In this framework, tests do not produce understanding...metrics thus became a favored means to establish control." So standardized tests are used to try to "modify" teacher behavior, and "No Excuses" charters use behavior modification on their students. But all this is because of simplistic, linear, unimaginative thinking, lack of understanding of cognitive psychology and the complex process of teaching and learning, and ignorance of the importance of the social context.

  3. Following another tangent (which is the way my mind often works), you mentioned "free will". I actually personally dislike this term, especially since I've come to the conclusion that it's a belief in "free will" that causes some people to blame the poor for being poor. They reason that people have free will, therefore the only thing that has caused them to be poor is either "poor choices", or "choosing" to be lazy, or criminal, etc., without taking into account the context and impact of any external determining influences.

    You could say that the argument of free will vs. determinism is like that of nature vs. nurture, both having influence of varying degrees, but to me free will vs. determinism is a false dichotomy because I don't think either concept exists in reality in the way that we think about them. I go with Spinoza in believing there really is no such thing as "free will"; we just think there is because we don't understand the causes of our decisions. It doesn't mean that anything is pre-determined, but using this framing puts a moral element into the issue that isn't helpful because the focus should always be on understanding the complex interplay of factors involved.

  4. Two things:
    (I apologize for still being "anonymous," but after several attempts to get on board one of the other platforms, I've given in to my Luddite frustration.)

    1. This piece is expressed so well. We are bombarded with all-or-nothing arguments all the time, and it's good to see some nuance in our discourse.

    2. I don't know if this is "behaviorism," but I have found great value in the practice of analyzing student tasks (or skills) by breaking them down into component parts. It gives me great insight into what students may be struggling with, and thus able to address their missteps or misconceptions specifically. Diagnostics are a great use of this sort of objectivism, as long as they are used as such.

    1. I do that too, and I think it's so helpful. I don't think it has anything to do with behaviorism, though.

      I forgot to mention, I think Peter doesn't like performance objectives much, but I always found an adapted form of them helpful. Just doing an activity isn't an objective, and if you just say you want them to "understand" or "appreciate" something, how do you know if they are? I DO want them to appreciate and understand things, and I can usually tell by their reaction, but I don't write that in my objectives; but especially where skills are involved (and there are a lot of different skills involved in teaching foreign language) I try to write as many objectives as I can where for assessment I want them to be able to do things that will show me as many things as possible that they can do with what they've learned, usually in a similar way to things we've practiced in class. But I don't think of that as behaviorism either, it's more complex than that and it's cognitive skills, not behavior.

  5. Peter, you lucky guy, you mentioned the best rock band in Uruguay. I learned this song on guitar and used to sing it to my kids when they were made them college-ready, I think:

    And Los Shakers are still shakin' in the 21st century...