When we’re teaching or shaping a new dog behavior, many of use the click/treat strategy that marks the behavior (click or verbal marker like YES!) just before we reward the behavior (with food or a toy). Dogs need to know every time they get it right … at first. Later, we introduce more intermittent reinforcements, where the dog doesn’t know which correct response will earn a reward. We’re told that a variable reinforcement pattern is the MOST motivating one out there.
It is, in fact, the one upon which gambling casinos operate. Slot machines, in particular, are all about variable reinforcement.
Not kidding.
Still Feeding After All These Years
Some SIX years into our odyssey into healing, training away, or at least mitigating Lilly’s fears, I continue to reinforce EVERY correct behavior in many, many scenarios.
Seeing or hearing other dogs or having them in close proximity is a good example, where pretty much any and every time that happens, Lilly is getting food and lots of it.
The Occasional Freak Out
Lilly is doing WAY better in many potentially scary situations than she was when things went from shy-puppy-worries to reactivity-a-go-go when she reached social maturity (around 2 1/2 years old).
These days, she better measures out her fear responses, especially if a snark is required. Smaller, less frequent snarks = normal.
The No Visiting Rule
The No Visiting Rule helped this process along … a lot. It simply means that I neither ask (nor allow) Lilly to engage with dogs we see in public. None.
Seriously. None.
Our behaviorist from Colorado State University prescribed this solution so that Lilly NEVER has to worry about meeting or having to interact with another dog. She never has to guess. She never has to wonder. She never has to believe that this time might be the one time.
Now, I cannot control loose dogs or other crappy situations, but in your average hey-we-both-have-dogs situations, we politely refuse to “say hi.”
I highly recommend using whatever version of NO VISITING works for your dog’s particular fears … at least until you have a solid foundation of behavior modification work done.
Always + Never = Success
So, if Lilly ALWAYS gets rewarded for offering the right behavior in the face of “scary” things, and she NEVER is forced to interact with dogs she’d really rather not know, then you end up in this place of coping pretty darn successfully in the world, generally speaking.
Variable Fear Scenarios
My question today is this: If variable reward schedules reap big rewards, then is it also true that variable fear scenarios — one BIG scare every once in a while — does an equal amount of damage?
I ask because a friend’s fearful dog had a rotten experience this last weekend. (Again, it isn’t my story to tell.)
I’m thinking mostly about known, specific triggers.
I’ve read some dog-fear info that talks about how quickly they can take seed and how long-lasting they can become, but I don’t think I’ve read anything about how a variable scare-inforcement schedule might work (or in this case, work against you).
Have you?
***
P.S. While looking online for a description of variable reinforcement schedules, I came across a blog by some psuedo-sauve expert who talked about using variable reinforcement on women. Ok. Seriously? Ick!
What a great question! I’ve been trying to think if there has been any one variable fear event that has occurred with my Daisy that reinforced a fear in her brain (she is a fearful dog too).
The only thing I could think of is the one time her friend (yes, friend) bit her pretty bad when they struggled over a stick. I supposed it could have been an event for Daisy, but I didn’t overreact. I didn’t baby her and instead kept her walking. It never led to any fear. She still loves other dogs and plays with them.
But maybe that isn’t a big enough variable fear event? You’ve got me thinking. I wonder…
I don’t think fear increases with variable reinforcement like learning does only because the dog is using two different parts of her brain. A dog is thinking while she’s learning, but reacting to fear. The fear reaction isn’t a cognitive one.
With that said, I think any experience that kicks a dog into survival mode has the potential to become hard wired quickly. Then, the more that brain pathway is used, the more efficient it becomes. It doesn’t seem like running into a trigger at a variable schedule would cause extra damage. It seems like frequency and intensity (like KB’s PTSD example and K’s reaction to her bad crash) matter most.
I’ve been wondering about dogs and memory lately. Today Kona *ran* to the rut that housed the rat she found two days ago. She obviously remembered the rat had been there. That was one experience, that lasted only a few seconds and happened two days ago. Kona still remembered. Makes me more cautious about putting her into a scary situation.
I’ve read psychology experiments that show that one single very scary event causes the release of such a huge amount of stress hormones that it becomes forever imprinted in the brain. The extreme case is PTSD – scientists are finally starting to understand why people with PTSD cannot ever repress the memories of their bad events. I actually had a minor case of PTSD after my dog was murdered a long time ago and then the man turned the gun on me and started conversationally discussing how he was going to murder me. I swear – it was as if that event implanted a brand new brain in my head. The world never looked the same again although counseling helped me learn to function in it again.
From an evolutionary view, it makes sense. If a bear meets a person with a gun once in its life but escapes, that bear will be most likely to survive for a long time if the event makes a HUGE impression on the bear and he forever makes it his mission to avoid people.
In a fearful dog, I wonder if similar mechanisms are at work but are in overdrive. For example, K still remembers the exact spot where I had a very bad (and scary from her viewpoint) mountain bike crash FIVE years ago. Every time we get to that spot (we pass it daily), she shrinks behind me to watch whether I’m going to become a human projectile. Then, once I get through it, she scurries forward to catch up. So, I’ve successfully negotiated that spot at least several hundred times but the one scary time drives K’s behavior. I bet that Lilly has a bunch of similar “memories” helping drive her behavior.
That’s what I was thinking too, KB. So … if one incident can imprint, can occasional encounters restart the process? I fear that’s the case. At least K will move past the spot once you clear it. I suspect Lilly would refuse to budge.
Holy synchronicity! I would never have, er, bet that we would both blog about reinforcement schedules ( of all things) on the same day.
This is a fantastically written article. What I like about it? 1/ it’s easily digested. None of these huge paragraphs am infamous for. 2/ the author blatantly knows what she’s talking about.
I’d heard about gambling being so addictive because of it being on a random ratio. Funny you reminded me.
I am also the proud owner of a scaredy dog, but he is still on a pretty much constant reinforcement schedule of: here’s a dog, here’s a treat for not freaking out.
I do let him do the meet and greet (but never force him) and I have to say and any set backs are quickly overcome. He was severely phobic and we’ve gone a long long long way.
Am not a big fan of meet and greets on the leash for obvious safety reasons, though (most dogs are far more frustrated and potentially aroused on the leash).
I can’t help but feel a little embarrassed at my ratio of reinforcement, though. It’s heartening to see that others keep rewarding too, way past the initial learning of the behaviour.