Today I thought I’d address a question that we occasionally get asked by clients…
“At what point can I stop using (the clicker/treats) with my dog for the behavior I am working on improving?”
This is actually a really interesting question to answer since it brings to light several aspects of dog learning and helps everyone understand better why dogs are motivated to do certain behaviors. In the example below (jumping up on the person when exiting the crate) here’s how I answered…
How long it takes to get to a point of consistency (where you will be phasing out the clicker/treats) depends on three factors…
- How long the dog has been doing the behavior (i.e how much learning do we need to undo)
- What percentage of the time the dog continues to practice the behavior (success versus non-success rate)
- And finally, how often you practice the training activity and reward the appropriate behavior (e.g. open crate, click (or mark) for 4-on-the-floor, rewards on the floor, return to crate, repeat)
I like to think of the point of decision for a dog to jump or not jump as a giant balance-scale. Think of each successful repetition (rewards for not jumping) for a behavior as one pebble on the scale of choosing to not jump in the future. The other side of the scale is filled with pebbles for each time that a dog has gotten to jump up and thereby rewarded herself for the inappropriate behavior. If the scale balances out, there is a 50/50 chance that she will chose to jump up or not to jump up in the future.
After some training, once the scale starts to lean more heavily on the side of “not jumping is better” (e.g. your dog has been rewarded 300 times for not-jumping and has jumped up 100 times in their life and gotten the reward of attention, now they are three times more likely to not-jump than to jump) first the clicker is phased out (while still giving treats for good choices) and then afterwards the treats are also gradually phased out, being replaced with verbal praise and petting for staying on the floor.
Because of the difficulty quantifying the factors involved (How many times has the dog jumped up in the past? How often will you practice? How effective is your management to prevent future practicing of the behavior?) it is difficult to pinpoint for an individual dog how long this is is going to take.
If your management of jumping (using baby-gates, tethers and other techniques) is effective close to 100% of the time and you are spending about 20-30 minutes a day in 5-minute sessions rewarding staying on the floor, you should make good progress. Alternately, the progress will be much slower if your management is only 60% effective and/or if practice happens for a shorter amount of time during the day.
Let us know what you think about your dog’s behavior and whether you also have seen what I described to be true.