“Please, just do anything else. Literally, anything.” – some poor trainer and their last ounce of patience.
Since we’ve gone over Differential Reinforcement of Alternate Behavior, and Differential Reinforcement of Incompatible Behaviors, today we’re going to be talking about DRO or Differential Reinforcement of Other Behavior.
In a nutshell, DRO is a behavior modification plan where you reinforce any behavior other than the target behavior. While DRA and DRI want to introduce a different behavior, DRO cares about the subject not doing the target behavior. Usually, this procedure is paired with a fixed interval reinforcement schedule, as in, you set a specific amount of time that your subject gets to perform anything other than the target behavior. Once that time limit has expired, you can reinforce your subject.
Let’s dig into some examples.
The most typical example of DRO can be seen in work with children. Back in college, I had a professor who would teach her behavior classes with examples from her work with teaching autistic children. One of the children she worked with was named Tina, who would self-stimulate by rapidly tapping her fingers together in front of her eyes. This made teaching her difficult since she would stop doing everything else to self-stimulate. Luckily, Tina also loved playing with her doll she named Allie. My professor decided that a DRO procedure would be the best approach to reduce the self-stimulating behavior of Tina because you could reward Tina with playtime.
With a DRO procedure in mind, Tina’s caretakers would start a timer for 5 minutes before they began teaching her. They told Tina that if the timer reaches zero, and she didn’t interrupt, then she would get playtime with Allie.
Each time they would reinforce Tina with playtime, they would add a little bit more time to the timer. So instead of avoiding self-stimulation for 5 minutes, they would have her avoid self-stimulation for 5 minutes and 10 seconds, and then 5 minutes and 20 seconds, and so on. However, the increments of 10 started to be too much for Tina to handle. So instead, they lowered the time and then switched to increments of 5 seconds which she was able to handle more easily.
As a dog trainer, I love using DRO for modifying jumping behavior – especially, if a dog has a bad habit of jumping on people to greet them.
While it’s not as interesting at the previous example, I accomplish this is by petting the hell out of the dog while they are on the ground. However, the second they jump up, I stop, turn around, and put my hands in my pockets.
This accomplishes two goals. First, the dog no longer gets pets or attention from me. This is a negative punishment because I’m removing the reinforcer of attention. Second, this also prevents any self-reinforcing behaviors. If you’ve worked with dogs before, you know that sometimes dogs will try and chew on parts of your clothes, lick your hands, lick your face, and plenty of other behaviors that make jumping up self-reinforcing. By turning around, you’re giving the dog almost nothing it can work with to make the effort of jumping on you worth it.
Once the dog is back on the ground, I wait a moment for them to calm down and then go back to petting them. While this doesn’t use a timer the way DRO is traditionally used, I’m making sure I only reinforce the dog for staying on the ground. Eventually, they get the idea that “paws on ground means pets” while “paws on person means no pets.”
This is exactly how DRO wants to work – it reinforces your subject for doing anything other than the targetted behavior. While more common uses of DRO will provide reinforcement at the end of the time interval, with dogs, I continuously reinforce the subject while they are doing anything other than the target behavior (in this case, jumping on people). Be aware, doing this will fit into a continuous reinforcement schedule and should be treated as such.
While this usually works for most dogs, I did have a few cases back in my shelter days where I needed to do a bit more to train dogs out of bad habits. A particular black Akita-mix named Balto and his obsession with jumping up and licking people’s faces come to mind. Situations like this would usually call for a more formal DRO procedure.
First, I set up a 10-second timer on my phone that would go off with a clicking sound. Every 10 seconds Balto didn’t jump on me would result in a click and a small treat. Eventually, I would increase this timer every 5-ish times Balto would meet the time limit.
Second, whenever Balto did jump on me, I would stop, turn around, put my hands in my pocket, and then reset the timer. Once Balto was back on the ground, I would wait a second, then restart the timer and continue interacting with him.
Thankfully for me, Balto was a quick study. He picked up that I was restarting the clock whenever he jumped on me and did his best to stop jumping. He would still jump on me sometimes – old habits die hard, after all – but the moment I started turning around he would immediately try to get back to the ground as soon as possible.
Be aware, if you are going to do this with your dog, make sure you always ALWAYS provide a treat when you click. This is an unspoken rule between you and your animal when you are training them – a click always gives a treat. Which, if you’re planning on using this kind of procedure in the near future, leads me to…
If you’re planning on using DRO, be aware that you should establish your baseline first. I recommend recording how long your subject goes between times where the targetted behavior appears. After that, you want to start them at a little higher than their baseline, and then increase it in increments that are comfortable and doable for them.
Another important consideration to keep in mind is what to do when your subject fails. First, do not apply any punishers or reinforcers. Reinforcers must be earned. Second, reset. Encourage your subject to try again and reinforce when they meet the time threshold. Always set up your subject for success. If your subject doesn’t reach the time threshold, then lower the threshold to an area that your subject will be comfortable with and then try again.
Differential Reinforcement of Other Behaviors is a behavioral procedure where you reinforce the subject for doing anything other than the behavior you’re trying to get rid of.
Both Tina and Balto were reinforced when they avoided the targeted behaviors in the time limit. Tina would have to avoid self-stimulation behaviors, but if she was able to avoid self-stimulation 5 minutes, then she would be reinforced with playtime.
Balto had to avoid jumping on people for 10 second time intervals. Every 10 seconds he didn’t jump on someone, I would reward him with a small treat.
For both of these cases, every time the subject successfully reached their time limit, they were reinforced for not doing the targeted behavior, and then the time limit was increased. This would get our subjects used to going longer and longer periods without the problem behavior occurring and reinforcement.
Thank you for reading!