21 December 2018

Better Behaviorism for K-12 Schools, Part 1 of 4

There is a contingency of folks in the alternative education world that are die-hard anti-behaviorists. One source of this ire is a book-length case against “behaviorism” as a tool for school and classroom management called Punished By Rewards by Alfie Kohn. Kohn makes some good points about ineffective school management practices, but from my perspective he misrepresents and/or misunderstands behaviorism. 
I consider myself to be a behaviorist since I have participated in the training of both rats in scientific experiments and pack llamas in the real world. Having lived on a llama ranch for years and worked with them on pack trips into the wilds of Oregon and Washington (though I have never owned one) I have directly observed extremely skilled animal trainers who were not behavioral scientists. But I also have a co-author credit on a published research paper in a well respected peer-reviewed behaviorist publication. I mention my minimal behaviorist credentials because for decades of my adulthood I have participated in alternative school communities which often have a decidedly anti-behaviorist attitude. But what I discovered in the process of being formally trained in the theory and practice of scientific behaviorism is that there are significant assumptions made by practicing behavioral scientists that are ignored in the applications of behaviorism in mainstream K-12 schools.
This was brought home to me by an interesting challenge posed in a class I was taking from Reed College emeritus professor Allen Neuringer, who was acquainted with B.F. Skinner himself (more on Skinner in part 2). He had us read a chapter of Alfie Kohn's book and at the same time a chapter from Karen Pryor's book Don't Shoot the Dog. Both are excellent books and I highly recommend them. Pryor believes that behaviorism is the solution to most of the world's problems while Kohn believes that it is a heinous crime against humanity. They both oversell their positions and, for contrast, I have overstated theirs. 
The question Professor Neuringer posed was, “Who's right?” They are firmly entrenched on opposite sides of the question of whether behaviorism works or not. So whom should we believe? My answer is that they are both right and we should believe both of them. But that only makes sense if you accept my contention that while they both use the term “behaviorism” they are not, in fact, talking about the same thing. Almost everything that Pryor references as effective practice in behaviorism is confined to dealing with people and other animals one-on-one, while almost everything that Kohn references asineffective behavioral practice is confined to dealing with groups. What I suggest is that Alfie Kohn was not talking about behaviorism, at least not as it is practiced by scientists. He was talking about a version of “behaviorism” used in mainstream K-12 schools that may sometimes borrow the name but fails to bring with it key assumptions. If those assumptions were brought along then they would guide teachers and administrators to different means of managing the behaviors of children and, if I am correct, then much of Kohn's criticisms would no longer be valid.
In my experience of behavioral research there are some interesting things that are essentially assumed. My co-author credit is for a 2014 article in the Journal For The Experimental Analysis of Behavior, a leading behavioral science journal. The article is about a set of experiments co-created by graduate student Lavinia Tan and Professor Tim Hackenberg and conducted with the participation of a class of undergraduates, including myself. I am going to quote for you one paragraph of that paper which is interesting because of how it can illuminate certain assumptions behind behaviorist research. 
Ten male Long Evans rats served as subjects, five in each phase of the experiment, which were conducted separately, several months apart. Rats were approximately four months old at the start of the experiment, and pair-housed. Note that due to the odd number of rats per phase, one subject in each group of five was housed with a sixth rat not used in this experiment. Colony rooms were programmed on a 12 hour light/dark cycle. Food was restricted 22 hours before experimental sessions.
First thing to notice is our subjects. “Long Evans” refers to a specially bred strain of rats that are supposed to be ideally suited to behavioral research. Notice that they were housed in pairs so that they always had access to a playmate. Notice that we restricted their access to food as experimental sessions approached. When your work depends on a rat eating food there is no point in doing the experiment if he is not hungry. Another way that experimenters can manipulate the effects of reinforcement is to take note of each rat’s individual preference for different flavored pellets, though that was not relevant in our experiment.
Our experiment involved just the observation of feeding itself, but if the experiment was about specific behaviors that rats are not inherently inclined to do, then “shaping” would be done to ensure that the rats are at least minimally capable of doing something like the desired behavior. That term “shaping” refers to reinforcing certain spontaneous behaviors to encourage them to act in the ways that the experimenter needs them to behave. For instance, rats are not inherently inclined to press levers for food. They have to go through a process of behavioral shaping in order to be able to reliably exhibit that behavior when they are hungry. While the process starts with spontaneous behavior it can lead to quite extraordinary feats.
You can see how elaborate animal training can become by checking out YouTube. I have collected a few choice examples for your convenience on my YouTube channel in a playlist under the name Behavioral Training. The first one is of B.F. Skinner himself demonstrating shaping on a pigeon. If you don't know who Skinner is there is more about him in Part 2.
One final thing to note is that behavioral research has been carried out almost exclusively with individual animals, not groups. The experiment that we conducted was innovative because it dealt with five rats simultaneously in one large box. We did not even attempt to train any specific behaviors, we were interested in how they would distribute their eating behaviors across two sources of food without training. Our box dispensed food at certain times at two different spots regardless of behavior.
So in behavioral science articles, they use a well-established shorthand to encode assumptions about breeding, feeding, and preliminary training plus they normally deal with only one animal at a time. That makes complete sense for behavioral scientists because they all presumably know what their colleagues are talking about and the frameworks they use are focused on the behavior of individuals. They are psychologists, not sociologists. But when behaviorism gets translated into schooling, then we should question whether or not those assumptions still hold true and, if not, raise doubts about whether the result should properly be called “behaviorism.” Despite this fact, many behaviorists count the applications in schools as a vindication of their work because those applications are measurably effective. The latest example I'm aware of is in the book The Nurture Effect: How the Science of Human Behavior Can Improve Our Lives & Our World by Anthony Biglan. 
The use of behaviorism in schools is also notoriously controversial. Newer programs avoid using the term “behaviorist” for fear of political backlash. If they use the term “behavior” they usually attach the word “positive” to frame their program favorably and avoid associations with some very cruel animal experiments from the past that do not sit well with the general public. So this brings us back to the question: Does it work or doesn't it? If the behaviorists are taking credit for applications in schools and have data to support their case, doesn't that mean it works? But then there is the controversy over these programs; if they really work then why the controversy?
Part 2 is called “Psychology is a Proper Science.” It is posted here.

No comments: