Training Time Responses Compared to "Real Life" Responses

This week I had multiple students tell me that their dogs are great in training but not in "real life."  It's actually not something that comes up too often, usually the dogs are better in "Real Life" than I would have expected.  At the same time it is a complaint that I often hear online.... "My dog is good at class but doesn't listen at home," or "He only listens if I have a treat."

In response, I have these questions:

1. How does your dog know the difference?
2. What do you do differently in training compared to other times?
3. How are we going to change things so that your dog is more responsive?

In watching the teams work, I usually note these things:

• The dog does not have 'fluent' behaviors. Sometimes the person has to ask more than once. The dog is inattentive. The dog is slow to respond or doesn't seem to fully understand. The precision can be lacking. The dog moves slowly or takes too long to respond.
• The person is often luring with food. Sometimes the human knows this, sometimes he or she does not know the luring is happening. The dogs learn that food is guarenteed only if they see if first.  The humans see the dog respond immediately when food is visible, thus the human is reinforced by a fast response.  
• Poor responses are reinforced. When a dog is just learning, we do reinforce approximations. If we are in a very busy environment we will reinforce lower quality responses.   But we do not always reinforce these less than desirable things. And especially a dog that has been through multiple classes, we really should be more selective about what to reinforce and when we decide to move and try again.

The changes we make:

• There is no difference between training and real life:  Training always can be happening. The dog is working for all of his or her food. The food and treats are dispensed throughout the day for all sorts of good behaviors. Training is always happening. This does a few things; it strengthens the behaviors, the person is wearing the treatpouch/has treats accessible and so reinforces more often, the behaviors are reinforced in real life settings.  
• Take away the cues: Often these dogs have learned treat pouch = I am 100% sure I will be reinforced. No treat pouch =  no chance of reinforcement.   We wear the pouch and do not pull reinforcers from the pouch. We have other ways for treats to be available; containers throughout the house, pants pocket, jacket pocket, food bin on a nearby table.  
• More good training:  Perfect behaviors. One cue, one response, then get the reinforcer.  Get fluent behaviors before adding cues.  Work on new behaviors and perfecting those in place.  Only getting out the reinforcer after the click happens, never before. No more luring!
• Specific activities to teach the dog that reinforcers will happen:  Ask for a super simple behavior. Then run to get really, really great treats.  Systematically teach secondary reinforcers.  Use the "Airplane Security" game*

*Somehow I haven't ever shared this game....it will be added to the "to do" list. I used to use it all the time...but have not in recent years.  Dump all your treats and toys onto a table, just like at airport security. Lead your dog away a bit, ask for a behavior, run to get a treat/toy.  Go away again, ask for something. Return to get a reinforcer. The dog is learning to work even if the reinforcers are not immediately present. The person is learning how to teach the dog to work when reinforcers are not immediately present.

More cute dog picture...not relating to the post.  Isn't this adorable?