Evidence-Based Knowledge: The Fool’s Gold standard of behavior analysis.

fools-gold-solar-powerEvidence Based Knowledge
In behavioral science there is a term that defines their gold standard for knowledge – evidence based. They use this term to discount anecdotal information and give the stamp of approval to specific concepts that are believed without question. New information may only be accepted after rigorous examination designed to produce incontrovertible evidence. Ideally this process will yield knowledge that can be trusted implicitly. In reality it creates an orthodoxy that is immune to examination or advancement. That is because the evidence is filtered by existing ideology. Unless you prove things their way, it’s not accepted. Their evidence-based knowledge can and does exclude anything they don’t want to acknowledge. So far, this gold standard of knowledge has produced a consistent outcome. It protects and serves the gate-keepers of behavioral science. It does not advance the human understanding of behavior in the real world. Instead of living up to its promise of creating a body of dependable knowledge it bars the entree of valuable information.

Minutia Exploded to the Universal:
The basis of virtually all behavior analytic research is the examination of the rate of response of a single behavior in one of two primitive species – rats and pigeons. The specific behavior is lever pressing for rats and key-pecking for pigeons. This activity is conducted in an “operant chamber” – a small cube devoid of anything other than the lever/keys and a feeder tray. Though the term operant describes any behavior created and maintained by its consequences, lever pressing and key pecking are the only ones available to the respective species. One does not examine how rats scurry or pigeons fly in an operant chamber. To quote the great behaviorist, Ogden R. Lindsley, one must remember that the box is not there to keep the animal in, it is there to keep the world out. And so it has been. With this miniscule examination of behavior, behavior analysts pretend to know how to rule vast populations of humans. Many of them promote the concept that behavior analysis presents solutions to world-wide problems based on counting how often a rat presses a lever. Knowledge leap-frogs from the miniscule to the universal in a single bound – all evidenced-based, of course.

The Only Evidence Worth Considering: Whatever they say it is.
This obsession with rate-of-response in an austere environment permeates behavior analysis. The process of gathering evidence revolves around this one minute view of behavior. The rats do not naturally press levers. That is the single “opera
skinner_boxnt” they learn. They must be taught to press a lever to create evidence. As there is nothing else in the box that functions the rat eventually hits the lever by accident. That produces food. The animal is kept about 20% below its normal body weight to make it supremely interested in food. That’s not normal. Rats are browsing, grazing animals that need a regular source of food. Wait a second, that’s an advantage! This facilitates the designs of the experimenter. The animals are taken out of the box primarily to make them hungry again and so the exhausted grad student can hit Starbucks on the way home.   275px-Schedule_of_reinforcementOnce the animal starts churning out evidence it is recorded over time. This is called a cumulative record and is the foundation of behavior analytic research. In essence, this is the core of their investigations of behavior. If you cannot present information in the form of a graph or charvarint it is discounted as not being evidence-based. That is the dagger in the heart of scientific investigation. While claiming to push investigations into behavior, behavior analysts exclude anything that contradicts their current understanding – an understanding built on a myopic view of the topic.

The Poster Child of Evidence-Based Behavioral Science:
Once Upon a time, B.F. Skinner and his research colleague, Ogden R. Lindsley set up an operant-chamber-style experiment to see if Beagles could determine pregnancy by sniffing the urine of women. Always on a shoe-string budget, they used an old refrigerator turned on its back as their “ope
rant chamber”. They concluded that the air-tight seal of a refrigerator would confine the scent and exclude extraneous smells. Inside the box was a single lever and a small niche. The cup of urine would be placed in front of the dog’s nose and if the urine came from someone pregnant, the dog was supposed to press the lever. The pressing of the lever produced “evidence” that could then be turned into charts and graphs. The experimental process would be identical to all the other evidence within behavior analysis and would be validated in apples to apples comparisons. That’s not what happened.

As soon as the dog figured out the game, they would, to quote Og Lindsley, ‘Bang the heck out of the lever with both paws and teeth.” Anyone who has ever seen an Airedale go after gophers has seen this behavior. Both front feet and the mouth are used to rip huge hunks of sod away to expose the crafty but elusive gopher. Skinner didn’t know dogs. He never considered instinct important enough to study because it complicated his one-dimensional assumption that all behavior is based on “antecedent events.” In his world, all is a matter of repeating single behaviors, over and over and over because they paid off in the past. Why would a beagle get all excited about simply making the right choice in a simple experiment, right? The short answer is because that is what beagles do. They are hunters. Once they know the scent of the prey, be it rabbit, raccoon or possum, their single-minded pursuit unleashes their instinctive passions. They will go through over or under any obstacle to get to the target, bugling all the way. Skinner metaphorically and unwittingly let slip the dogs of war.

To stop this beagle nonsense Skinner had Lindsley make a sleeve that forced the dog to use a single leg to press the lever. This was simply another short-sighted and naïve decision. According to Lindsley, the dogs almost dislocated their shoulders trying to get one leg, the other leg and both legs into the sleeve, while biting at the edges of the metal tube. All in all, it didn’t go well. Two things were painfully obvious – Skinner didn’t really understand situations where more than a single behavior was possible and he didn’t know anything about beagles. Beagles are highly variable. Skinner picked them for the same reason he picked rats and pigeons for his research – they were plentiful, small and cheap. Skinner displayed his limitations by trying to limit important information about behavior. He didn’t look at the animals as a source of knowledge unless he could shove them into his tiny one-dimensional world. He considered their behavior an affront to his data. That caused him to throw out huge hunks of knowledge that might have allowed him to know something of real behavior. The beagles presented an incredible amount of evidence that was shut out of his evidence-based knowledge. That is the legacy of behavior analysis and evidence-based knowledge. To paraphrase the old Southern aphorism, it’s not that the dog can’t hunt, it’s that they won’t let him out of the box.

Note: I knew Og Lindsley for the last 12 years of his life. We were 042214_1849_ClickerTrai1.jpggood friends and I am honored that he held me in special regards as a colleague. The tale of the beagles is one of the many that he told that I valued as evidence of a full understanding of behavior. Og was one of the few within behavior analysis who was never limited by myopia. This is mildly ironic because he wore “Coke bottle” glasses and had poor physical vision, uncorrected. His mental vision was so far seeing that the world lost a great mind when he died. If you wish to learn more, Google “free operant” conditioning and tack his name onto it. He was long a champion of broadening the scope of behavior analytic research.

2 thoughts on “Evidence-Based Knowledge: The Fool’s Gold standard of behavior analysis.

  1. Interesting article could you expound what would be a more suitable method of observing and recording behaviors and all the other social and emotional Interaction that are natural attributes of humans

  2. John, First one would have to examine and catalog phylogenetic behavior common to the target species. That has never been done with humans. Skinner picked rate of response because it could be counted – ending in the culmination of that art by Lindsley’s Precision Teaching and the use of a standard celeration chart. However, rate of response is a low priority factor in almost all behaviors outside a 3rd world country sweat shop. How you do a behavior compared to a defined standard is far more important. In essence, Skinner’s world is a hot-dog eating contest.

    Another problem is that while there is a ‘learning theory’, there is no ‘doing theory’ – but the purpose of learning is to do things. It is a transitional phase that is no longer active once a person knows a skill. I don’t say the ABC’s any more. They were a vehicle to teach me the alphabet – which was a vehicle to learn words, syntax and fluency. HOW is speak or write is the important thing, not how often or how many words I write – yet that does not exist within Skinnerian parameters and is therefore not ‘evidence’.

    In this day of computers, the beagles could have been tracked by a multitude of factors that were cast off by Skinner. That experiment was in the early 1950’s. How difficult is it to record that the dogs were incapable of actually pressing a lever with a single paw? You simply write it down and record it as ‘evidence’. If further investigations yield similar or identical results, that is valid evidence. The quest to discover the human genome was a matter of cataloging sequences, not by some contrived analysis, but by reporting what actually exists. That is the key to developing sophisticated understanding of behavior.

    As for the latter part of your question, one would have to explain the functional relationship of emotional interaction before determining whether it is valuable or not. Are emotions the engineer or simply passengers on the train? If the are collateral byproducts of various forms of physiological arousal, then they are, of themselves, not causative and not as important as understanding what triggers them. For instance, if I hit my thumb with a hammer, my emotion is meaningless in correcting the problem.

Leave a Reply

Your email address will not be published. Required fields are marked *