-- Steve Brown
I attended a Bob Bailey/Parvene Farhoody seminar this weekend. The seminar was entitled Out of the Lab and Into the Field: Performance Dogs in the Real World and covered how to apply scientific training methods to dogs that need to perform in real-world scenarios. The concepts of operant conditioning, stimulus control, generalization, when to shift criterion, fluency vs. proficiency, distractions, etc. were the main focus of the lectures. Randy Hare and Lucy Newton gave a great demo with several dogs applying these concepts to scent detection work.
But I'm not going to review the material Bob and Parvene covered this weekend. This post is about what I was left feeling and thinking after this weekend, the questions I'm left with and how I'll possibly proceed.
Am I creative enough to come up with training plans tailored to each animal, specific and detailed enough to elicit only the approximations of the end behavior I'm looking for in order to get the animal their in the shortest amount of time? Bob talked about training being a craft for most of history and not a science. In my other hobbies (sewing, baking, gardening, etc.) I am naturally a craftsperson, I can follow a pattern or recipe very well. I can use traditional skills and create something fairly impressive, but I am not really an artist. I very rarely create a truly unique piece of art or anything one of a kind. In dog training, I am very good at following a prescribed plan, the more detail the better. But I'm not sure I can think creatively and critically enough to come up with such detail on my own.
Am I able to take enough risks ("take a flyer" as bob would say) and set the animal up for something just a little more complex than the last time? I am a very anxious person who doesn't like change. When I train, I tend to start out where we left off last time, or even worse a few steps behind where we left off, thus always setting us back at least a little. I tend to stagnate at a certain level for too long, not wanting to progress too fast but end up not progressing at all. Some days I am so paralyzed by the fear of not knowing exactly what the next step should be or not wanting to make mistakes that I don't train at all. I know in my mind that it is worse to stall training totally rather than just get on with it and make mistakes. Yet I have a hard time clearly defining the next steps and following through when it is not exactly spelled out.
Am I observant enough to pick up the tiny shifts in behavior required to really move forward? I find it extremely difficult to both observe myself, my body language, cues, reward delivery, etc. AND still focus on the animal's behavior and responses well enough to deliver properly timed rewards and (importantly) NOT reward incorrect behavior. It's a lot to keep track of all at the same time.
Bob often says something akin to "Anyone can train any dog to do anything using any method given enough time." Do all methods work? Yes, but not all methods work with the same speed and accuracy. This isn't my career, I don't have to produce a 100% fluent working dog in 8 weeks to make a living. This is my hobby, if I'm not the greatest at it no one will starve. I can take my time learning and maybe not ever be great, but my dogs and I should absolutely be enjoying it. If we aren't there is no reason to continue doing it.
But that's no excuse to settle for sub-par training methods. I know it is much kinder to the dog to use methods that are very clear and work relatively quickly so the dogs can have success and earn many rewards in a short time. It isn't fun to see a dog confused and stressed because they don't know what is expected or how they can earn reinforcement. Will I ever find myself scientifically training at 100%, tracking data, keeping myself cool and calculated during trials instead of whooping it up and hugging my dogs? I don't think so, but I will certainly utilize the parts that make sense to me and incorporate them into my training.
What I took away (besides about 15 hours worth of behavior theory lectures):
- To effectively teach, you have to be ready to change your OWN behavior!
- "What do I have" - "What do I want" = "What I need to train"
- Get the desired behavior, then add environmental changes quickly. After two correct responses, move on or change something, don't stall and slow down learning. Adding distractions is really adding discrimination to the learning process, which enhances learning.
- Before creating a training plan, clearly and objectively define the target behavior including what it will look like and what the final cue will be.
- In regards to conditioned reinforcement (such as a clicker) "When in doubt, leave it out!" A lot of times a clicker simply muddies the water and removes the primary reinforcement from the behavior by one unnecessary step.
- Once fluent, the animal should exhibit the same latency, magnitude and intensity in response to the cue as it does to the primary reinforcer.
- You will make mistakes. Mistakes will slow learning down. But don't let mistakes keep you from trying new things or working to be a better trainer.
- Do not take any training technique or advice as gospel. Really critically analyze what you are being told, scrutinize it and decide for yourself if it passes the test of good science/advantageous for your situation.