Psychology Factsheets www.curriculum-press.co.uk
Number 17
Operant Conditioning This factsheet: • summarises the main features of operant conditioning; and • assesses the role of operant conditioning in the behaviour of non-human animals.
Consequences of behaviour
We have underlined keywords that you should know for the exam.
When an animal performs a behaviour, there can be many types of consequence. The consequence may be something pleasant which will make the animal repeat the behaviour or something nasty which will stop the animal from doing that behaviour again.
What is operant conditioning?
Schedules of reinforcement
• •
•
•
It is possible to give reinforcement in two ways – continuous and partial. • Continuous reinforcement gives a reward after every response the animal makes. For example, the rat will get a pellet of food after every lever press. • Partial reinforcement gives a reward after only some responses. Skinner found four schedules of partial reinforcement: 1. Fixed ratio schedule – reward after a certain number of responses. For example, a food pellet after every 8 presses on the lever. 2. Variable ratio schedule – reward after a certain number of responses on average. For example, food after 8 presses on average, so there is sometimes a reward after the 6th press and sometimes after the 10th press. 3. Fixed interval schedule – reward following the first response after a certain interval of time. For example, food for a lever press every 5 minutes. 4. Variable interval schedule – as for fixed interval, but on average. For example, food reward about every two minutes (sometimes 1.5 minutes, sometimes 2.5 minutes).
Operant conditioning explains how animals learn certain behaviours. Operant conditioning focuses on the behaviour an animal makes and the consequence of that behaviour. The consequence of the behaviour determines how likely it is that the animal will do that behaviour again. The consequence of a behaviour can be positive (something pleasant) or negative (something unpleasant). Whether the consequence is positive or negative determines whether the behaviour will be repeated in future (i.e., learnt). There are two laws: o The law of reinforcement → a positive reward/reinforcement (e.g., food/praise) increases the chance of learning a behaviour. o The law of contiguity → the behaviour and consequence (e.g., reward) must happen close enough together in time for learning to occur.
An example of operant conditioning -Imagine that your new puppy is in the garden and wants to come in, but the door is shut. On one occasion your puppy sits by the door and accidentally taps the door with his paw. You hear this and quickly respond by opening the door. Your puppy finds this pleasant as he wants to be with you. Your puppy will gradually learn that tapping the door means that you let him into the house.
Partial reinforcement is better in conditioning a behaviour than continuous reinforcement. It is better because the occasional lack of a reward is normal. In continuous reinforcement, the animal expects a reward every time, so the lack reduces the chance of the behaviour being repeated. Extinction is when a behaviour is extinguished (unlearned) if the response is not reinforced. The behaviour stops as the animal learns that the behaviour has no consequence. Partial reinforcement leads to slower extinction than continuous reinforcement. The extinction is slower because the animal is used to getting rewards infrequently and keeps going for a while in expectation of a reward.
The effect of his behaviour (tapping the door) results in something positive (you opening the door). So, your puppy has learnt the behaviour by the effect (consequence) it had on the environment.
Table 1: Types of consequence Type of consequence
Description
Example
Positive reinforcement
Increases the chance of a behaviour occurring again by providing a pleasant consequence.
Receiving a reward (e.g., food) after pressing a lever.
Negative reinforcement
Increases the chance of a behaviour occurring again by removing (or escape from) an unpleasant consequence.
Moving away from an electric shock.
Secondary reinforcement
Primary reinforcers (rewards) are linked with secondary reinforcers.
Food is a primary reinforcer whereas a mother is a secondary reinforcer because she provides the food.
Punishment
Decreases the chance of a behaviour occurring again by providing an unpleasant consequence. There are two types of punishment positive punishment and negative punishment.
Positive punishment provides a negative consequence by giving something.
Giving an electric shock.
Negative punishment provides a negative consequence by removing something.
Removing food
Exam Hint: Learn the Keywords underlined. Candidates should be able to describe, give types of consequence and schedules of reinforcement for operant conditioning,
1
Psychology Factsheet
17 Operant Conditioning
Another name for operant conditioning is ‘instrumental conditioning’. The theory of operant conditioning developed from classical conditioning (see Glossary). Thorndike developed this theory and then Skinner took it further (see text boxes below). Skinner’s rats in a ‘Skinner box’ In 1938, B.F. Skinner examined learning using rats. He placed a rat in a cage that had a lever inside. The lever connected to a food dispenser on the outside (a ‘Skinner box’). Pressing the lever dropped a pellet of food inside the cage. Skinner observed that the rat first pressed the lever by accident but soon learnt the link between the behaviour (pressing the lever) and the reward (food).
Thorndike’s cat in a puzzle box Edward Thorndike thought that learning happens by trial and error. He tested this in 1898 by placing a hungry cat in a ‘puzzle box’ with food hanging nearby just outside the box. The box could be opened by pulling a string which opened the latch. He observed that the cat performed several behaviours such as meowing and scratching. The cat accidentally opened the latch by pulling the string and escaped. The next time he put the cat in the box, it opened the box faster than the first time. After a few more trials, the cat learnt what to do and released the catch immediately.
Exam Hint: The exam question may ask for operant conditioning or classical conditioning, or both. Stick to the type of conditioning asked for in the question and do not get them confused!
Evaluating the role of operant conditioning in the behaviour of non-human animals In this section, we will look at the strengths and limitations of operant conditioning and how it relates to behaviour of non-human animals.
Strengths • • • •
Operant conditioning explains a wide range of phenomena, such as learning language and phobias. It has practical applications, such as training animals (see Tortoises and Shaping text boxes below). It examines learning in controlled experiments in the laboratory. The findings are easily replicated. Operant conditioning enables animals to behave efficiently in their environment. For example, operant conditioning enables the best foraging strategies in birds. They learn where they are most likely to find food and return to these areas to constantly check for food (rewards).
Tortoises Weiss and Wilson (2003) used positive reinforcement to train four tortoises to approach a target (a red plastic ball on a stick) and hold their head still while touching the target. They were given food as a reward for doing this. Training the tortoises to do this meant that it would be easier to take blood samples than if someone had to pick the tortoise up.
Cognitive Factors Animals may use reasoning (cognitive factors) in learning a behaviour. Operant conditioning ignores cognitive factors in learning. However, some research has found that animals do use reasoning. For instance, in Mackintosh’s (1994) study, rats pressed the lever in a Skinner box for a food reward. The experimenter then gave the rats an injection to create an aversion to the food. They found that the rats eventually stopped pressing the lever. This finding is difficult to explain without cognitive factors as the experimenters did nothing to directly stop the rats pressing the lever. Instead the rats had associated the lever with the aversion.
Shaping Operant conditioning can train animals to do complex tasks by rewarding animals for a behaviour that is similar to the desired end behaviour. The training happens step by step until the animal is doing the complex behaviour. For example, to get a tortoise to hold still at target, Weiss and Wilson began by rewarding the tortoise if it moved slightly towards the target. They then rewarded it for standing next to the target and so on until it would hold its head still on the target.
Pigs Skinner proposed that any behaviour could be conditioned in any situation (he called this ‘equipotentiality’). However, it is more difficult to train some animals on some behaviours than others. For example, Breland and Breland (1961) tried to train pigs to put a wooden token into a piggy bank. The pig would then get a reward. However, the pigs would not do this. Instead they would pick the token up and drop it. The pig was doing its natural ’rooting’ behaviour for finding food. It appears that animals only learn behaviours through operant conditioning that are like their instinctive behaviour.
Limitations • Experiments on operant conditioning use animals and so there are ethical issues. • As these experiments use just a few species of animal (e.g., cats, pigeons, rats), the findings may not generalise to other species. • Experimenters study operant conditioning in laboratories and so the findings may not be ecologically valid. • Animals are learning unnatural behaviours that they would not do in the wild (e.g., pressing a lever). • Operant conditioning does not take cognitive factors into account (see Cognitive Factors text box). • Operant conditioning does account for innate abilities that adapt a species to its environment (See Pigs text box). • Operant conditioning does not consider other forms of learning such as observational learning. For example, Sherry and Galef (1984) found that birds learnt to open a tub of cream by watching other birds do it.
Glossary Adapt: a change in a species’ behaviour to fit in with its environment. Aversion: a strong dislike. Classical conditioning: proposed by Pavlov, this focuses on how a stimulus (e.g., a bell) is associated with a bodily response (e.g., salivating). This response is not under the animal’s control. Ecologically valid: An ecologically valid experiment is one that can be generalised to real-life events. Foraging: behaviour involved in finding a suitable source of food. Innate: abilities that an animal is born with. Replicated: the findings of an experiment can be repeated (replicated) to test whether the same result is found.
2
Psychology Factsheet
17 Operant Conditioning
Worksheet: Operant Conditioning Name 1. What is operant conditioning? ...................................................................................................................................................................................................................... ...................................................................................................................................................................................................................... ...................................................................................................................................................................................................................... 2. What are the two laws in operant conditioning? ...................................................................................................................................................................................................................... ...................................................................................................................................................................................................................... ...................................................................................................................................................................................................................... 3. What types of consequence strengthen a behaviour? ...................................................................................................................................................................................................................... ...................................................................................................................................................................................................................... ...................................................................................................................................................................................................................... 4. What is continuous reinforcement? ...................................................................................................................................................................................................................... ...................................................................................................................................................................................................................... ...................................................................................................................................................................................................................... 5. Explain two types of partial reinforcement. ...................................................................................................................................................................................................................... ...................................................................................................................................................................................................................... ...................................................................................................................................................................................................................... ...................................................................................................................................................................................................................... 6. Give two advantages of operant conditioning. ...................................................................................................................................................................................................................... ...................................................................................................................................................................................................................... ......................................................................................................................................................................................................................
Example Exam Question (a) Describe the nature of operant conditioning. (12 marks) (b) Assess the role of operant conditioning in the behaviour of non-human animals. (12 marks) To answer part (a), you need to outline the main features of operant conditioning. It requires a description but not an evaluation of the theory. You should define operant conditioning and describe the laws, types of consequence and reinforcement schedules. You should also briefly mention Thorndike and Skinner. Try to use an example of operant conditioning (e.g., Skinner’s rats). To answer part (b), you need to discuss the strengths and weaknesses of operant conditioning. Any points from the strengths and limitations sections above are relevant. Include examples of animal behaviour in your evaluation, such as Breland and Breland’s study or Weiss and Wilson’s experiment. Acknowledgements: This Psychology Factsheet was researched and written by Amanda Albon. The Curriculum Press, Bank House, 105 King Street, Wellington, Shropshire, TF1 1NU. Psychology Factsheets may be copied free of charge by teaching staff or students, provided that their school is a registered subscriber.
3
