One of the most frequent questions clients ask at an orientation, when being introduced to clicker training is, “Yeah, but will I always need to use the clicker and treats?” The answer is, yes and no.
The clicker is a precise teaching tool, useful for teaching new behaviors and refining behaviors which need to be cleaned up for a particular aspect of fluency. Once a behavior meets your criteria, you can absolutely transition to a verbal marker to maintain the behavior. Because behavior is in a constant state of flux, you may occasionally want to pull your clicker back out – if there is too much latency in a response or if you need a straighter, neater, more precise version of the known behavior. For pet dogs, the clicker can often be abandoned as soon as a cue has been added to the behavior or shortly thereafter.
Once you get rid of the clicker, it’s time to begin reducing your reliance on food. The first step is getting the food off your body. The presence of your treat bag or food in your hands may very well become a cue for the behavior, giving you a “show me the money” dog. It is a good idea to hide small containers with non-perishable treats throughout your house, so that you may ask for a behavior without food on your body and quickly run with your dog to the nearest “treat spot.” Once your dog is readily responding to your cues without needing to see food, there are a variety of techniques you can use to further decrease your reliance on the primary reinforcer of food. These techniques include:
- introducing reinforcement variety – reinforcement variety is the key to reliable, consistent responses. If you want your dog to think “outside the treat bag,” you must as well. Generally, this is as simple as paying attention to your dog to find out what she likes. Truly, reinforcement variety could fill a number of blog entries by itself as reinforcement values are not static and are highly individual in nature. For this entry though, you may want to consider the following additional types of reinforcers: life rewards (going for a walk, getting out of the crate, greeting a new person or dog, going for a ride, sniffing a fire hydrant, chasing a squirrel, getting out of the bath tub, coming in from the rain, going swimming, entering the training classroom or field, etc.), tactile rewards (petting, scratching, etc.), play rewards (with any toys your dog likes, games your dog likes, access to social interaction with appropriately matched doggy playmates, etc.), and conditioned secondary reinforcers (a smile, eye contact, or “good girl” from the handler).
- raising your criteria – if you consistently reward slow or sloppy behaviors, your dog will consistently offer slow and sloppy behaviors. Once your dog understands a new behavior, it is time to start raising criteria. Raising criteria can be most effective when combined with a differential reinforcement strategy which means better behaviors earn better reinforcement. If I cue a “sit,” for instance, a response within one second of the cue may earn a hot dog, a response within three seconds a “good girl” and a scratch, and a sit later than three seconds, no reinforcement at all. Behavior can always be improved – better, faster, neater, performed around higher level distractions, etc. To use this technique effectively, you need to have a good idea of your dog’s performance baselines – on average, how long does it take for your dog to respond to your cue? Below-average responses should not earn reinforcement.
- more behaviors per reinforcement – instead of reinforcing a single repetition of a behavior, you may begin asking fr a number of responses to earn a single reinforcement. Instead of clicking and treating Mokie for each step of loose leash walking, a single sit or a single down, I may require fifteen steps in heel position, a sit, five more steps in heel, a down for thirty seconds, and six hand targets for a single reinforcement. If this was my training plan for a session and Mokie happened to offer me the snappiest, straightest, most perfect down of all, I may choose to forgo the six planned hand targets and reinforce that one beautiful, perfect behavior in the sequence.