Book: Teaching with Reinforcement 

Understanding Reinforcement

Author: Kay Laurence
Many of us enjoy teaching in a positive style. It demands us to be more creative than the alternatives, and certainly it can force us to take the long route rather than the easier short cuts. The benefit of this style of teaching is a really happy learner, who develops a keen interest in acquiring new skills and a really happy teacher that gets a buzz out of seeing our dogs blossom. It develops a trust that goes beyond teacher-learner relationship.

I teach many varied classes. From experienced clicker trainers interested in specialised activities to people making their first contact with positive teaching. I am also fortunate to teach in many places outside the UK and view hundreds of different styles of clicker training, not just my home-grown learners.

Many of us succeed in teaching complex skills with talented and sometimes less talented dogs, and often help a partnership find a happier, balanced life. But there are occasions that although the prescribed protocols have been followed, the results are below their expectations. What appears to be teaching with reinforcement may be hiding a process that diminishes the value of the reinforcer, thereby diminishing the behaviour.

Examples:
A dog is given lots of treats and “good dog” for a sit on greeting, but still jumps up.
Despite lots of clicks and treats when performing a behaviour or trick, the dog seems to droop and lose interest.
On the surface everything seems to be on track but when exploring a little deeper there are undercurrents affecting the behaviour: reinforcement patterns.

I have grown more respectful of these powerful currents over the recent years. We have examined many ways of teaching and communicating to dogs and learners, and have developed an extensive library of skills and tools to teach high quality behaviours with ease. We can become addicted to this learning acquisition period which brings us a great sense of achievement, and clicker training with a high rate treat delivery can ease us into a feeling of having completed the job. Now we need to explore further and really examine all the subtle elements that distort behaviours.

I am often found reminding people that it is the nature of behaviour to change. But what is changing the behaviour? The reinforcers. A young collie pup can develop an eye for herding his relatives. Every single time he gets to use this skill the behaviour gets stronger. It is an instinctive and innate activity that exponentially magnifies each time he uses it. At present this is on track and permissible. But what if he diverted his eye to herding traffic instead of his relatives? Every time the pup was exposed to traffic and allowed to practice the behaviour it would become stronger and much harder to reduce or eliminate. This meets the definition of a reinforcement in progress: a behaviour increasing or maintaining.

The reinforcement history of the behaviour can dictate how we plan to change the behaviour. Every week it may take a different strategy to manage the behaviour since the reinforcement is changing.

Treats and rewards

One of our failings is the habit of saying one thing but having a different intent. We may reward or treat our dog with the delivery of food but are somewhat disappointed in the result. We sigh as we give the treat. It “taints” the reward. We tell them they are a good dog, but look away as we secretly wish they were better. We get wrapped up in our own intent, and can be lazy in actually analysing what is happening and what is needed.

We wear a feeling of satisfaction that we have rewarded the good sit and never allowed the jumping up to gain our attention, but the jumping up still occurs. It doesn’t matter what we think we have rewarded – that is our own selfish assessment – what has been reinforced is the jumping up.

If the behaviour repeats, gets stronger or happens more frequently then it is being reinforced. It absolutely doesn’t matter what you think you are rewarding, the behaviour is being reinforced. Your intention is of no value, intentions cannot act as reinforcers.

reward n.1. something given in return for a deed or service rendered. 2. a sum of money offered, esp. for help in finding a criminal or missing property. 3. something received in return for good or evil; deserts. ~vb. 4. to give something to (someone) esp. in gratitude for a service rendered.
reinforce vb. 1. to give added strength or support to. 2. to give added emphasis to; increase; “his rudeness reinforced my determination”. 3. to give added support to (a military force) by providing more men or equipment
Reward implies an interaction, an exchange of something done for someone else, the person rewarding gains something from the actions of the person. A tip may be given as a reward for service rendered.

A reinforcer does not imply a pay-back for the deliverer of the reinforcer. It has a more independent association, where the action is being reinforced, and is not giving back to the reinforcer. A tip may be intended to act as a reinforcer but if the value is less than the recipient expected then it is not reinforcing. If I provided excellent service and was given a tip, I may expect a fiver, but if I received only £1, then my service, (actions) were not reinforced, but I am sure the customer delivering the £1 tip thought they were rewarding me.

Once we begin to see the reinforcer “currents” that surround us all the time we can begin to become more effective and more efficient in our teaching and interaction.

Exercise : Reinforcing not rewarding

Consider the difference between reinforcing a behaviour and rewarding a behaviour. Ask yourself what would be reinforcing for this dog at this moment and what would be rewarding for me? Are they the same thing?

Your dog may be jumping excitedly because you are home after a 3 hours absence.

What would be rewarding for you:

  • the dog to wait quietly whilst you put the shopping away
  • the dog to sit and wag their tail

What would be rewarding for the dog:

  • to leap in your face and lick you all over
    (be careful if this is also rewarding for you, because somedays it may not be and you will have established the habit) 

A compromise where both you and the dog can be happy would be a sincere 60 seconds of greeting, managed to prevent a leap to your face. Then you can put the shopping away. The behaviour of greeting under control was reinforced by meeting the dog’s needs at that time and you did not get a black eye! For both of you to be reinforced, what happens needs to be of value to both of you, not just one at the expense of the other. What is reinforcing is measured by the recipient, not who delivers it.

It will get clearer as we progress.

0 Comments

Submit a Comment

Your e-mail address will not be published. Required fields are marked *

Get in touch

8 + 5 =