Friday, 22 June 2012

Learning And Training

In the past I have written a short post on the learning theory and Parelli. This post is further going to discuss learning and training in the horse.

When training a horse it is important to take into account the way they learn. This will lead to more effective training strategies.

This is the waning of a response due to repeated stimulus presentations, not followed by any form of reinforcement. For example, the horse may get less scared of an object after seeing it for a number of times. This is also affected by the temperament of the horse.

Christensen et al (2006) looked at responses of horses in frightening situations and said they are important for both equine and human safety. Their objectives were to investigate which of 3 different training methods (habituation, desensitisation and counter-conditioning) was most effective in teaching horses to react calmly in a potentially frightening situation. In the habituation condition, horses were exposed to a carrier bag in training sessions until they were habituated. In the desensitisation group they were gradually introduced to the stimulus and habituated to each step before they moved on. In the counter-conditioning group they were trained to associate the stimulus with a positive reward.

Horses trained with the desensitisation method showed fewer flight responses in total and needed fewer training sessions to learn to react calmly to the carrier bag. In addition, all horses on the desensitisation method eventually habituated to the test stimulus whereas some horses on the other methods did not. Desensitisation appeared to be the most effective training method for horses in frightening situations.

Classical conditioning
This is when there is an unconditioned response (reflex)(UCR) to a specific stimulus such as food. An association is made between the unconditioned stimulus (UCS) and a conditioned stimulus (CS) to produce a conditioned response. This is also called Pavlovian conditioning. This can be better explained when looking at Pavolv's dogs the stimuli were as follows; food (UCS), salivation (UCR), bell sounded (CS). They were conditioned to salivate when the bell sounded as they associated this with the food. 

Operant conditioning
This is an extension of classical conditioning and is The foundation of equine training. It is not necessary that they understand what to do at first, they use trial and error with reinforcement which increases the likelihood of them performing the behaviour. It becomes a conditioned response to a conditioned stimulus. 

There are 3 stages to learning;
1. Pavlovian acquisition- link between external stimuli. What they have done and a reward. 
2. Action outcome.  They monitor the actions and the rewards.
3. Habitual responding- habit is formed and there is stimulus response learning. More like a reflex. Not monitoring it in the same level. You can get errors because you are not fully registering it.

Main problem with positive reinforcement is that there is only so much you can get the animal to do. You can get a sense of how much they enjoy something or how much they want to avoid it by how much work they will do for you (if the costs outways the benefits).  The horse has evolved a social structure from both positive and negative reinforcement but largely negative. So from an evolutionary perspective they have evolved for negative reinforcement. Therefore this may be more effective. When discussing negative reinforcement an example is applying the leg aids to a horse then releasing them when they perform the movement, the reward is the releasing of pressure. 

The intelligence of animals is difficult to assess. Intelligence may be shown when the animal learns to ignore irrelevant stimuli, just as it also learns to react to significant stimuli.Maze tests have been used to measure the learning ability. They are good because they are used without human presence. Foals have been found to learn quicker than their dam.

Training generally means drawing out desirable behaviours and suppressing undesirable innate behaviours to institute novel behaviours. The aim of training is to install signals that result in predictable behaviour patterns.

Clicker training is an example of a use of secondary reinforcers. The animal expects a reward after the clicker. This bridges the gap between the animal performing the behaviour and receiving the primary reward. 

There are welfare implications in failing to identify adequately the mental abilities of all animals in the care of humans. Overestimating an animals mental ability must be seen as a major welfare issue when it manifests as abuse, wastage, stress and conflict behaviours.

Love Laura

1 comment:

Carly said...

Really interesting write up!

I've never really been a fan of Parelli. I've always thought the ideas were cool, but that he was more in this for the money now than the horses. I do use a few of his concepts however- just without purchasing the tools he brands/recommends.

I'm a huge fan of Monty Roberts, though! It'd be really interesting if you did a write up on him sometime. :) i like to see other peoples thoughts on the different ways of natural horsemanship

Related Posts Plugin for WordPress, Blogger...