The importance of setting out a written training plan in order to collect evidence and set dogs up for optimal success
Originally published in BARKS from the Guild (issue 42 / may 2020)
Guppy, a young male “pit bull” and German Shepherd Dog mix, didn’t know how to sit. Or, more accurately, he didn’t know how to sit on cue. To be considered adoptable by the suburban families that frequented the shelter, this was a pretty important behavior. Guppy was incredibly friendly and goofy and he loved people. So he jumped all over them. 70 lbs of exuberance, tongue and pointy teeth a few inches from your face isn’t exactly what most folks write down under “I’m looking for…” on adoption forms. Guppy could knock down able-bodied twenty-something kennel attendants, let alone seniors or small children.
His trainers decided a sit (on cue) would be the best way to replace the jumping behavior. But the actual training proved very difficult. Trainers tried luring the behavior, but his butt never touched the floor. They tried capturing a sit and noticed that he never seemed to sit on his own.
Shelter workers started coming up with Zebra-scenarios like, “He’s got hip-dysplasia and can’t sit!” or “He’s not very smart,” or “He’s not food motivated.” The shelter veterinarian ruled out any pain or medical cause, and Guppy gobbled up all meals. Was he truly un-intelligent? Even the word “dominance” was thrown into the mix. Guppy was being labeled as “stubborn” and “untrainable”. For a dog already happily homed with a family, being unable to sit on cue might not be the worst fate, but for a shelter dog it could easily mean a long length of stay, at best.
Why do trainers and dogs reach these impasses? When progress plateaus, trainers begin to consider more invasive measures. Dogs are sent to boot camps and given shock or prong collars because “treat training” didn’t work. The problem is rarely motivation, of course, or the effectiveness of positive reinforcement. The problem is the trainer’s inability to set appropriate criteria.
Jean Donaldson explains criteria as your “contract with the dog.” What, exactly, does the dog need to do in order to get paid? Or, in the case of classical conditioning, what exactly needs to happen before the food starts? While some people may have a very good idea of this contract in their mind, a written plan with specific criteria steps is the best way to avoid roadblocks to progress.
For Guppy, the criterion of “nose follows lure over head until rear-end touches floor” was too high. A simple adjustment to easier criteria was required. Guppy could follow a lure so his head was tilted up, so that could be reinforced. Then Guppy could tilt his head up for three seconds, so that was reinforced. Then Guppy could slightly bend his back knees while his head was tilted up, so that was reinforced. In three ten-minute sessions, Guppy was sitting for a food lure. A couple sessions later, the lure was faded out. And a couple sessions after that, Guppy was reliably sitting on a verbal cue. He was adopted.
The ability to set good criteria isn’t just for jumpy, mouthy shelter dogs. It is even more important in cases of fear and aggression. Trainer Kelly Lee, PhD, CTC, is currently fostering an extremely skittish dog named Pancake. In order to set criteria appropriately, a trainer needs to fully understand the parameters of their training plan. She said, “For Pancake, any plan has to include a parameter for session duration, my body’s orientation, and eye contact. He’s not comfortable in the presence of people for prolonged periods, and eye contact spooks him. I’m constantly re-evaluating his plan. With fearful dogs especially, you might feel stuck if you aren’t able to come up with ways to split criteria into small pieces. You might stall-out in training progress.”
No doubt there are professional animal trainers who are adept at shaping behavior and effective behavior modification done by feel. They have an innate sense of when to make things harder or easier for the animal. But that’s rarely transferable to clients.
Professionals who practice behavior modification on fear and aggression cases know progress can be slow. Making progress at all is something to be celebrated. It’s not always easy for guardians to see their dog’s progression or improvement. A written plan of criteria can serve as a roadmap and a barometer.
“When I train with a plan I’m able to show a client, ‘Look, two sessions ago we were on Step 2 and now we’re on Step 12. Even though your dog can’t be picked up yet, we are on our way there,” said trainer Kylie Reed, CTC. This can help immensely in client satisfaction and patience with the process.
PPG’s Pet Rescue Resource has a page completely devoted to criteria changes called “Training With A Plan.”
As reward-based trainers, we are constantly touting our use of “science”. Of course, the science of classical and operant conditioning is working, and for everyone, whether a trainer is aware of it or not and whether they are force-free or not. “It's true whether or not you believe in it," goes the famous Neil deGrasse Tyson quotation.
It’s often troubling to me that reward-based trainers don’t utilize scientific principles, namely controlling for variables and making decisions based on evidence (e.g. Did the dog make criteria on the trial?) Without a written plan of incremental criteria changes, how can a trainer collect this evidence? And how will they know what to change when the dog is not improving? So often I see guardians giving up on positive reinforcement because “it’s not working”. The science works just fine, it’s the pace that the guardian is frustrated with, and training without a plan can make the process even more inefficient.
Trainer Maria Karunungan, CTC, PhD recounted a recent consultation with a dog guardian. The client was a first-grade teacher. Dr. Karunungan said, “She was having a really hard time understanding why her dog was barking and lunging at strangers. She said he wasn’t like this as a puppy. At one point she was worried about how long it would take for her dog to be able to pass strangers without barking and lunging. She said, ‘I’ve been doing the treat thing and he still sometimes barks and lunges,’ so we had to talk about criteria. She needed to understand what, exactly, her dog could handle, today. I asked her if she would expect her first graders to do multiplication and long division and she looked at me like I had three heads. I said, ‘This is what is really going on with your dog. We might be expecting more from him than he is able to do today.’ Her criteria was too high. Eventually her dog will get to that point, but written plans of incremental criteria changes helps humans temper their expectations of their dog getting from A to Z in one day.”
コメント