Using Reinforcement To Change Behaviour
If you want to change your pet's behaviour, why use harsh techniques when gentle ones will work better?
Training with positive reinforcement
What do your electric can opener and your postie have in common? They are both expert pet trainers.
For instance, 'Rumbles' the Burmese quickly learnt that food was likely to materialise in her bowl when she heard the characteristic grinding of the electric can opener. The noise and food were regularly linked together so she quickly learnt that if she ran to the kitchen when she heard the noise, her behaviour was rewarded with food.
The postie, too, is a good dog trainer. Just ask 'Gizmo', the Bearded Collie. Gizmo knows that if he barks at the postie, the postie always 'runs away'. In reality, the postman is just continuing his normal route. By 'running away', the postie positively rewards Gizmo's barking. Gizmo thinks the bark-and-chase routine is a real hoot. To him, the whole scenario is joyful - so why not continue? His owners are not so sure.
For Rumbles and Gizmo, the can opener and the postie are agents of positive reinforcement. These 'agents' train each animal, progressively, to perform a particular behaviour.
You can use reinforcement in many practical ways to change your pet's behaviour for the better, too.
When you reinforce a pet's behaviour, the behaviour becomes stronger and is more likely to occur again. Reinforcement is a very strong training and behaviour conditioning tool. There are two forms - positive and negative reinforcement.
Positive reinforcement is where the performance of a behaviour results in a pleasant outcome. Of course, Gizmo's owners would much prefer that Gizmo had positively reinforced a better behaviour than his annoying, frantic and incessant barking at the postman!
Negative reinforcement is different. It is where your pet strengthens a behaviour by moving away from, or avoiding, an unpleasant stimulus.
For instance, if you are speeding down a freeway doing 120 kilometres per hour but the speed limit is 110 kph what happens when you see a police camera car parked on the side of the freeway?
You immediately lift your foot from the accelerator and feel relieved as you slip past at 109 kph, thus avoiding an embarrassing fine.
The police camera car negatively reinforces your 'driving under the speed limit' behaviour.
'Roxy' presents an interesting case of negative reinforcement. She was a bright and active German Shepherd cross but when a thunder storm arrived, she was petrified and resembled a hairy blancmange more than a dog. It took 'Roxy' no time at all to realise that running into the house and hiding under the table was the thing to do. There, she gained comfort from her owner's company and she reduced her exposure to the noise, wind and lightning of the pounding storm thundering outside. The result of her behaviour was far more pleasurable than braving the thunder and lightning.
'Roxy' had experienced negative reinforcement. Negative reinforcement occurs when an animal performs a behaviour to avoid or reduce something unpleasant. The comforting outcome of 'Roxy's' actions meant that she reinforced her 'moving-away-from-the-storm' behaviour.
What about punishment?
Punishment is the opposite to reinforcement. When you punish a behaviour, it becomes weaker and is less likely to occur again. The rolled newspaper is a primitive form of positive punishment.
Looking at your driving behaviour again, if you are travelling at 120 kph again and a police car approaches from behind with sirens screaming, you know what happens next.
You pull over and the burly policeman slowly and deliberately writes you a ticket, while his gaudy police car is behind with it's lights flashing. You can feel the mocking smirks of the other drivers as they pass by. You correctly receive a tongue-lashing from the police officer, squirm as you pay the fine and worry that you have lost a few points from your license.
The police officer is thus positively punishing your 'driving over the speed limit' behaviour.
Another example is 'Fur Ball', a perplexing Persian. She insisted on doing her whoopsies in her owner's vegetable patch. The soft, tilled soil made an easy and convenient toilet, but no one in family really wanted to eat the vegetables afterwards. No wonder!
Something unexpected happened next time 'Fur Ball' casually sauntered to the vege patch. She was saturated by her owner's new toy - a Scarecrow! This is a sprinkler activated by a motion-sensor similar to those sensors on security flood lights. It gives a three-second blast of water and then resets itself for the next assault. 'Fur Ball' soon decided that the vege patch had lost its charm. Her 'approaching the vege patch' behaviour had been punished (in a gentle manner) and was thereby weakened. This is positive punishment - where an unpleasant experience is an immediate consequence of an animal's behaviour.
Punishment, however, should not be used for aggressive behaviour problems.
Negative punishment is where you remove something pleasant or desirable to weaken an unwanted behaviour. For children the 'You will not watch TV for the rest of the week for fighting with your brother!!'
Can you use negative punishment with your pets? Time out is a form of negative punishment that works with pets in many situations.
Using reinforcement to change behaviour
Let's look at a real example. Jessie the Terrier was a teeth-gnashing terror. She was aggressive when anyone tried to grasp her collar. Her aggression was immediate and intense. However, she tolerated handling around her face and head well and, if they avoided her collar, her owners could usually handle her easily. Could Jessie's owners use positive and negative reinforcement to control this problem?
Positive reinforcement for Jessie's behaviour involved stroking her on the nose and muzzle. If no aggression occurred, then her owners gave her a food reward. She tolerated this stroking well and it was a good starting point.
After stroking her on the muzzle a few times, Jessie started to look forward to the food reward. Now we had her on the go! The owners progressed by stroking her on the head and later by quickly hooking her collar with a finger. They released the collar after a second or two. Through this progression, if Jessie showed no snapping, they rewarded her.
They progressed by attempting to hook-the-collar for longer and longer periods. Jessie enjoyed the food rewards and the whole process was quite smooth.
Negative punishment could also be used to weaken Jessie's aggression. When her owners handled her collar and she responded aggressively Jessie was immediately put outside the house or confined in a room away from the family company and activity. Jessie did not like being removed from the family. It was removing something pleasurable and was therefore negative punishment.
If you are trying a routine like this, don't 'time out' your dog in a gruff, surly manner. Just coolly and calmly walk your dog to the time out room and close the door.
For 'Jessie', this was a useful and gentle technique. After this, they tried the positive reinforcement routine again. As needed, Jessie's owners switched from positive reinforcement to negative punishment. Jessie's improvement was rapid and her owners were very pleased.
Sensitive New Age Pet Owners (SNAPOs for short) know that positive punishment - where a nasty stimulus is used to change a behaviour - is often very harsh and is the least desirable way of causing behaviour change. They also now that positive reinforcement and negative punishment are usually soft and progressive techniques and, generally, are quite effective.
So, SNAPOs - burn the rolled newspaper in protest! The time for humane animal training has arrived.
By Dr Cam Day BVSc - Last updated