We’ve been talking about different forms of negative reinforcement in class. So, I thought this might be a good week to talk about positive reinforcement, negative reinforcement and different types of behavioral responses associated with negative reinforcement. First the basic behavioral definitions, just to make sure we’re all on the same page. Reinforcement is defined as anything that follows a behavior and increases the rate of that behavior. If the rate of the behavior doesn’t increase over time, reinforcement is not taking place. (I’ve talked in the past about finding good reinforcers.) Also, something must follow a behavior to be reinforcement. (So giving instructions before a behavior is not reinforcement because it comes before the behavior.)
Two Types of Reinforcement
1. Positive Reinforcement
During positive reinforcement, something follows the behavior and increases the rate of the behavior.
- The dog sits, you give her a treat, and her rate of sitting increases in the future.
- The child screams in the grocery store, the parent buys the kid some candy to get the kid to shut up, and the kid is more likely to scream on future trips to the store.
- Your horse nips at your pocket, you swat him across the nose with your hand, and his rate of nipping increases.
- The horse walks onto the tarp, you give her a long belly scratch, and she is more willing to walk over the tarp in the future.
2. Negative Reinforcement
During negative reinforcement, something is taken away following the behavior, and the rate of the behavior increases.
- You push down on your dog’s bottom, he sits, you remove your hand, and his rate of sitting increases.
- The rider tugs on the lead rope, the horse moves forward and the rider releases the pressure and stops pulling. The horse begins to learn that if she moves off at the slightest tug, she can escape most of the pressure.
- Kid’s parents nag him to take out the trash, as soon as he takes out the trash, they stop nagging. In the future, the kid takes out the trash before the parents start nagging in order to avoid the nagging.
- Going inside if it starts raining to avoid getting wet.
The important thing to remember here is that negative does not mean “bad” and positive does not mean good. In everyday language, we attach emotional meaning to these words, which behavior analysts do not. Positive reinforcement simply means something is added following the behavior and negative reinforcement only means something is subtracted following the behavior.
Picking Good Reinforcers
A great way to pick good reinforcers is to remember the letters DISC. A good reinforcer is something the animal is Deprived of, will occur Immediately after the behavior, is an appropriate Size, and is always delivered Contengent on the behavior. The rules of DISC apply well for choosing both positive and negative reinforcers. (I won’t go in to great detail about this here, if you want more information on DISC, feel free to read this post.)
Behavioral Responses Associated with the Use of Negative Reinforcement
When negative reinforcement is used, two different behavioral patterns develop, avoidance behavior and escape behavior. During escape behavior, the animal exhibits the behavior in order to get rid of, or terminate, the negative reinforcer. During avoidance behavior, the animal exhibits the behavior in order to prevent the negative reinforcer from even occurring at all.
Example: Getting a child to do their homework
- Escape Response (negative reinforcement): Every afternoon when she gets home from work, Jenny’s mom nags and yells at her to do her homework. As soon as Jenny starts her homework, her mom stops nagging. She learns she can terminate her mom’s nagging by doing her homework.
- Avoidance Response (negative reinforcement): Jenny finishes her homework every afternoon before her mother gets home from work. She can completely prevent or avoid the aversive stimulus of her mother’s yelling by starting the behavior before her mom has a chance to start delivering the negative reinforcer of nagging.
- Positive Reinforcement: Jenny’s mom compliments and praises her when she does her homework and then allows her to go outside and play with the kids who live down the street.
Another example: teaching a horse to lead at the trot
- Escape Response (negative reinforcement): The trainer starts jogging and waving his long whip or training stick behind the horse. The horse begins trotting and the trainer stops waving the whip. Anytime the horse slows down to the walk, the trainer flicks the whip until he trots again. The horse begins to respond faster in order to terminate the pressure of the whip.
- Avoidance Response (negative reinforcement): The trainer begins jogging and the horse immediately trots along beside him. The horse has learned that he can completely avoid the pressure of the whip by trotting and keeping up with the trainer.
- Positive Reinforcement: The trainer rewards the horse with a treat at the walk whenever the horse’s eye is even with his shoulder. Gradually, he walks faster and only gives the horse a treat everytime the horse stays in the correct spot for a given length of time. Once the horse understands the concept of “stay with me,” he tries jogging. Sometimes the horse trots along, sometimes it doesn’t. He rewards the horse with a treat whenever the horse trots along and the horse’s rate of trotting increases. (This actually invoves a bit of shaping as well, click here to read more about shaping.)
Why does all this matter?
Training (and life in general!) is a fairly complex matter, and many situations and trainers end up using both of these to varying degrees. It’s impossible to completely avoid either one of them. It’s good to be aware of these definitions so that we can evaluate how much of each we use in our training. Positive reinforcement means using something the animal wants and is actively working towards. During negative reinforcement, the animal is actively working to get rid of something aversive or annoying. This doesn’t have to be something big, even a slight annoyance is enough motivation for a lot of animals. Think of a horse who will swat his tail to get rid of a single fly. (Can you think of several ways you use each of these strategies when interacting with people or animals around you?)
Negative reinforcement must always be used with some caution for the following reasons.
- You must apply pressure, something annoying or something aversive in order to take it away, that’s the very nature of negative reinforcement. This can elicit confusion or an emotional response (such as frustation) from the animal, especially if it takes the animal awhile to figure out what is expected of it.
- Some animals start to try to escape from not just the negative reinforcer, but also the environment or the person delivering the negative reinforcer.
- If the trainer is not extremely careful, the animal can become desensitized to the negative reinforcer over time. (For instance, the schooling pony who needs spurs and a riding crop to convince him to move forward because he has become so desensitized to the rider’s squeeze.) To avoid this pitfall, the trainer must have a good feel for both timing and the amount of negative reinforcement to use.
- Negative reinforcement is closely tied to positive punishment and also can lead to the development of posioned cues. (These topics are extremely complex and I won’t go into them here. For more information, this article is a good introduction.)
- Negative reinforcement creates a different mindset than positive reinforcement
Using Positive Reinforcement to Create a Positive Mindset
Use of negative reinforcement can create an animal that is obedient and compliant. It knows it’s job and is willing and able to perform cues when ask. However, any extensive use of negative reinforcement without positive reinforcement leads to an animal that is only working to avoid. The animal has little incentive to offer a bigger response or new behaviors. Also, the animal is only thinking how he can avoid the next negative reinforcer, instead of developing thinking and problem solving skills to figure out how he can earn the next positive reinforcer. Many training goals work best when positive reinforcement is used. However, in some situations, combining positive reinforcement with just a bit of negative reinforcement can be extremely successful as well. Whatever strategy we use, I think it’s important to consider the following questions:
- What is motivating the animal to respond?
- Am I setting the animal up for success?
- Am I creating an environment where the animal actively wants to participate and learn?
- If I gave the animal a choice, would he stay and participate, or would he leave?
As I’ve written about previously, ultimately, we have to look at motivation from the animal’s perspective, not from our own. If we are setting the animal up for success, he is actively learning, and he stays and plays when given the choice, then our choice of reinforcers is working well. If we can’t answer yes to all those questions, it might be time to go back to the drawing board and see if we need to decrease our session time, increase our rate of positive reinforcement, or otherwise modify the situation so the animal can be more successful.