Back to Math Class – Understanding Positives and Negatives in Behavioural Training
This entry was posted on Thursday, October 05. 2006 and is filed under Clicker Training,General Training.
In a previous article on Clicker Training we touched lightly on the terms positive and negative when referring to reinforcers and punishers. There is much general confusion about terminology of behavioural training and the number one issue is probably the use of the words negative and positive to describe consequences. In today’s article we’ll look at why it is necessary for us to really understand these concepts and why it does matter if we are doing the right thing but don’t understand the ‘scientific’ terminology.
| ...the terms negative and positive are only correctly used, when they are applied as if we were back in school solving a math problem. |
On one level you could say that the ‘doing’ is much more important than the terminology. If you only ever trained your own horses and were doing everything correctly this is fine.
However, for most of us we do more than that. We talk to others about our horses and training programs and we learn from others by watching demos and clinics or going to seminars and reading articles or books on how to get the best from (and for) our horses.
This is where the significance of understanding the terminology comes in. If we are going to talk to one another then we need a common language. This is particularly important when the things we are going to talk about have rules that go along with them in order for them to work as expected.
We commonly use the word negative to mean bad and positive to mean good. This in itself means that people are subjective about how they use the words. What is good and fine to one person may be totally unacceptable to another, so you can immediately see how this rapidly becomes a problem for communicating ideas.
So let’s clear up this confusion now. When we are discussing behavioural modification (training) the terms negative and positive are only correctly used, when they are applied as if we were back in school solving a math problem.
- Negative means remove/subtract/take away/minus.
+ Positive means add/plus.
Recently because there is so much confusion about the use of negative and positive, many people in the training world have started to replace the word negative with removal i.e. removal reinforcement or removal punishment. Others have devised a symbol system using the math symbols plus and minus e.g. R+ (or +R) for positive reinforcement and R- for negative reinforcement and so on.
Okay, so now we know that negative and positive refer to removal or addition of ‘something’, but it still may not be clear why we should know this or how it helps us to train better and problem solve if we do know. This is where our rules come in. Behaviour changes over time because of the consequences that occur to the animal as a result of that behaviour. The rules for each type of consequence, the things that will cause animals to change their behaviour over time, are quite different from one another.
We must be able to identify what is causing behaviour in order to change and then apply the right ‘rules’ in order to either correct it (if we don’t want that behaviour) or encourage it (if it’s something we do want). Perhaps most importantly we need to understand the rules for each type of consequence because it’s when we break them that we tend to get the wrong behaviour entirely or unwanted side-effects. By having the ability to strip anything behaviour to its most basic elements and say ‘what is happening here’, instead of labeling the horse naughty or stupid or disrespectful, we have a much better chance of spotting problems and sorting them out.
Let us then, take a look at the four major consequences that we use when training animals (or indeed humans).
1. Negative Reinforcement R- Negative = Remove something, Reinforcement = Increase Behaviour
- This is also known as removal reinforcement or pressure-release. This is the most commonly used method of training.Both traditional and ‘natural’ horsemanship use Negative Reinforcement.
The rules for a negative reinforcement are:“Something the horse doesn’t like is removed when it does the thing you want” or you can turn it around to “Something the horse doesn’t like is applied until the horse does what you want (when you must immediately remove it).”
- The most important part of negative reinforcement is the RELEASE. The pressure IS NOT what teaches the animal what you want and this makes this the most misunderstood training method.
- Example: A leg aid is a negative reinforcement. You apply something the horse doesn’t like,in this case the pressure of your leg\spur, then release it only when the horse moves forward.
- Common Problems – Incorrect release timing (too soon), No release or release is too late. E.g. Removing the leg prior to the horse moving forward or not removing the leg when the horse has moved forward.
2. Positive Punishment P+ Positive = Add something, Punishment = Decrease Behaviour
- Positive punishment is also often used with horses and other animals.
The rules for positive punishment are: “Something the horse doesn’t like is added when it does something you don’t want” or “You add something that (you hope) will decrease behaviour”
- Example: A smack with the whip after refusing a fence is a positive punishment.
- Common Problems: Timing and easily becomes abuse – refer to article on ‘problems with punishment’.
3. Positive Reinforcement R+ Positive = Add something, Reinforcement = Increase Behaviour
- Clicker Training uses positive reinforcement.
The rules for positive reinforcement are: “Something the horse likes is added when it does something you want” or “You add something that the animal wants at that particular moment in order to increase a behaviour”
- Example: The horse enters a float calmly and is rewarded with a haynet on entry.
- Problems: Incorrect timing & rewards are provided instead of reinforcers i.e. we assume that rewards are reinforcers but this is not necessarily so.
4. Negative Punishment P- Negative = Remove something, Punishment = Decrease Behaviour
- Negative punishment is also often used with children but frequently forgotten with animals.
The rules for negative punishment are: "Something the horse doesn’t like is added when it does something you don’t want” or “You add something that (you hope) will decrease behaviour”
- Example: A timeout (removal from friends or interaction).
- Same problems as positive punishment, especially if the trainer believes that negative punishment is automatically ‘not as severe’ as positive punishment.
The four consequences listed are not the final word on how behaviour changes. We must recognise that not all consequences are equal. Punishers are unreliable at changing behaviour in the way we think they should and “No Reinforcer” is also a consequence. If we ignore or more accurately ‘don’t reinforce’ a behaviour, that will also decrease behaviours (more reliably than punishment).
In conclusion, being able to speak a common language to one another will help us in our understanding of how different methods of training operate and whether to expect various techniques to work for us long term, without unwanted side-effects or not. Understanding the correct usage of the terms; positive and negative, reinforcement and punishment, gives us a universal foundation upon which we can discuss behaviour and training with one another, confident in the knowledge that we are all talking about the same things.
copyright Horse Play Ltd 2006
