Thursday, April 15, 2010

Choosing reinforcement

The wikipedia entry for reinforcement is as follows:

Reinforcement is a term in operant conditioning and behavior analysis for the delivery of a stimulus, (immediately or shortly) after a response, that results in an increase in the future rate or probability of that response. The response strength is assessed by measuring frequency, duration, latency, accuracy, and/or persistence of the response after reinforcement stops. Experimental behavior analysts measured the of rate of behaviors as a primary demonstration of learning and performance with non-humans.
A reinforcer is the stimulus, event, or situation whose presentation is dependent upon a response.

I think I spent more than my share of undergraduate efforts studying and writing about reinforcement - interestingly enough this was during my dressage queen days and before my dog trainer days. Oh how things change but stay the same huh? I'm not going to get into BF Skinner's positive and negative reinforcement cues - not in this post anyway - stand by for something like that in a future post.

The topic of reinforcement is a timely one that we should all consider. It's really tempting for handlers and trainers to place value in objects that we provide to our dogs. The temptation to find a really super powerful treat to bring to shows, the toy that we see on a shelf or in a catalog that we just have to have - only to proudly present it to our dogs as reinforcement and have them give us that look as if to say "yeah, that's oh-kay...but what else ya got?". I've fallen "victim" to this thinking as well - there was a time when on the eve of every show I'd go to Papa Franks (in Winooski, VT) to pickup a to-go order of their AMAZING meatballs. Not for myself - but for Murphy. I love Papa Franks meatballs - Murphy likes them - so ergo they must be the highest value reinforcement possible for Murphy...right? Well maybe...maybe not.

Murphy is all about the food. It doesn't matter what that food is or the power of that food. By power I'm describing value. If I were a normal dog I would expect a hot dog to be more powerful (valuable/tempting) than an oyster cracker or a carrot. Murphy is one of those dogs who would rather have five pieces of carrot rather than one piece of hot dog. Teller on the other hand, appreciates food and certainly doesn't turn food treats down - but in the grand scheme of things he tends to work for the sake of work. Hot dogs are fine with him - but all things being equal he likes pocket lint too. He's similar with toys - he likes to tug - would prefer to retrieve but his favorite reinforcement is physical attention, thumping** and verbal praise.

This comes to mind because I recently found a sheepskin tug that Teller had received as a puppy gift. It was (is) gorgeous - but Teller was too tiny for this massive tuggie when we received it as an 9 week old puppy, Murphy doesn't particularly like to tug so the tuggie was tucked away in a box until I found it a few weeks ago. When I rediscovered the fleece tuggie I was excited for a new toy - a new reinforcement. Teller has a couple of sheepskin tuggies from clean run (just like the one at the left) that he loves - so surely he'll like the really thick sheepskin braid right? I brought it to the school while I was training and he was clearly interested in the toy as I tucked it under my arm for later use. After breaking off a particularly fantastic effort on a short sequence I shook the fleece and invited Teller to tug with me. He came at it with gusto, but the verdict was clearly "well, I guess..if you really want me to - but I don't like this as well as I like my frisbee". Clearly he didn't enjoy the tuggie as much as I did. The bottom-line is that we don't get to choose what our dogs find reinforcing.

This is all the more challenging when we have goals of competing with our dogs - those food and toy oriented dogs need to "get off the monkey" so to speak and we've got to work all that much harder to make the work of the sport and playing that sport with us reinforcing enough for the 25-60 seconds (agility) or the 5-10 minutes (for obedience) that we're in the ring competeing. Of course reinforcement (or lack of) does not equal proofing and lack of reinforcement with a correct performance does not necessarily mean that a dog truly understands his job in those circumstances - all great topics for a future blog posts.


** Before anyone calls animal control on me (again), Thumping is relatively solid pats about the dog's neck, shoulder and ribcage. Pat, pat, pat, butt scritch, pat, pat, praise, pat, pat, praise. It's that deep pat that tends to echo a bit - for whatever reason that's Teller's absolute favorite thing - and I like that I can bring that "thumping" into the ring with me.

No comments: