#14082 - Operant And Classical Conditioning - Psychology

Operant and classical conditioning

Constraints on learning – learned taste aversion

Maladaptive learned behaviour – examples from phobias and addiction

• The behaviourist approach makes certain assumptions: that external rather than mental factors are responsible for behaviour and that simple associations of either the classical or operant kind are the building blocks of all learning. However, it now appears that different mechanisms of learning are involved in different species i.e. lending support to the biological perspective.

• Learning: a change in behaviour that results from practice. The study of learning considers the conditions in which associations are established (as opposed to roles in retrieval and storage as is studied in memory). Habituation is the loss of an effect caused by a stimulus e.g. if a loud noise is repeated over and over then the startle response diminishes.

Classical conditioning

• Classical conditioning is a learning process by which a previously neutral stimulus becomes associated with another stimulus through repeated pairing with that stimulus. In the early 20^th century, Ivan Pavlov carried out his famous set of experiments measuring the salivation rate in dogs. Dogs will normally salivate (unconditioned response) in response to seeing/smelling food (unconditioned stimulus). Pavlov had the dogs set up so that salivation could be measured and meat powder could be dispensed when needed. Each time, prior to meat powder being delivered, a light was turned on. The light was then turned off when the meat powder arrived. Eventually the light (conditioned stimulus) alone caused salivation (conditioned response).

• Classical conditioning plays a role in emotional reactions such as fear: e.g. if a cat repeatedly scratched a child when the child was young, the child has a negative association between the cat (conditioned stimulus) and being scratched (unconditioned stimulus) and is likely to be fearful or even have a phobia of cats when older. Gradual and increased exposure to cats may reverse the association and cure such a fear or phobia.

• For many years, the behaviourist view was dominant in classical conditioning but then some researchers argued that the critical factor behind conditioning is what the animal knows (cognitive view) i.e. the animal acquires new knowledge about the relationship between two stimuli which is why it acts as it does therefore there are some cognitive factors which must be considered when looking at classical conditioning.

• Contiguity vs predictability: since Pavlov’s time, researchers have tried to determine the critical factor required for classical conditioning to occur. Pavlov believed that the critical factor was ‘temporal contiguity’ of the conditioned stimulus and the unconditioned stimulus i.e. the two must occur close together in time for an association to develop. However, an alternative view is that the conditioned stimulus must be a reliable predictor of the unconditioned stimulus i.e. there must be a higher probability that the UCS will occur when the CS has been presented that when it has not. Research has shown that it is the predictive power of the CS (conditioned stimulus) which is the most important factor in classical conditioning e.g. in experiments where there was a control group of dogs which always received a CS and a test group which only sometimes received a CS, the control group became conditioned much more rapidly than the test group.

• Predictability is also important for emotional reactions. If a CS reliably predicts pain for example, then the absence of that CS predicts that pain is not coming and the individual can relax. The CS is therefore a danger signal and its absence is a safety signal. If there is no reliable predictor then fear is constant whereas with a reliable predictor, fear is only seen when the danger signal is present. E.g. if a child is told by the dentist that something won’t hurt then it actually does, the child is likely to become fearful of going to the dentist (as opposed to if the dentist says something may hurt a bit in which case the child is fearful until the procedure is over but has a good predictor of the pain).

Operant conditioning

• In operant conditioning certain responses are learned because they operate on or affect the environment. Animal must elicit the response in order for the reinforce (uncondition stimulus) to be delivered.

• Thorndike studied the process of operant conditioning. He placed cats in a cage which had a door which could be opened by pressing a simple latch with a piece of fish just outside the cage. At first the cat would try and reach the fish by putting its paws through the cage then when that didn’t work, the cat moved around the cage, engaging in a variety of behaviours. Eventually the cat inadvertently presses the latch which opens the door and allows the food to be reached. When placed back in the cage, the cat goes through roughly the same series of events but over a series of trials, much of the irrelevant behaviour prior to opening the latch is reduced until the cat eventually opens the latch as soon as it is put in the cage. Thorndike explained this: the cat is engaging in trial-and-error behaviour and when a reward follows one of those behaviours, the learning of the action is strengthened (the ‘law of effect’).

• Skinner carried out some simpler experiments into operant conditioning. A rat is placed in a cage with just a small lever in it. The rat is left to explore and the number of times that the rat pushes the lever is taken as the baseline level. Then, the cage is switched on so that pressing the lever delivers a small pellet of food to the cage. It is seen that the frequency of lever pushing increases dramatically. Similarly, if the cage is switched off (so pressing lever doesn’t deliver food) then the rate of lever pushing diminishes dramatically.

Reinforcement (of desired behaviour) and punishment (of undesired behaviour)

• Positive reinforcement: A pleasant stimulus that follows a desired behaviour – increases the likelihood of that behaviour.

• Negative reinforcement: removal of an unpleasant stimulus after a desired behaviour occurs – increases the likelihood of that behaviour.

• Positive punishment: presentation of an unpleasant stimulus after an undesired behaviour occurs – decreases the likelihood of the undesired behaviour.

• Negative punishment: removal of a pleasant stimulus after an undesired behaviour occurs – decreases the likelihood of the undesired behaviour.

Constraints on learning

• Do the principles of learning apply to the arbitrary pairing of stimuli and responses as occurs in classical conditioning? i.e. saying that any stimulus can be conditioned to cause a response? Evidence suggests that we are biologically ‘prepared’ to learn and to associate certain stimuli with certain responses according to which species we are. This has been investigated most extensively using taste aversion.

• In taste-aversion experiments, rats were allowed to eat food and each time they ate, there was a light plus a click. They were then exposed to the same food which was mildly poisoned and made them sick. Which stimulus (taste, light or sound) did they come to associate with being sick? According to classical conditioning, the stimuli should have been equally effective but they found that taste was the most strongly associated with feeling sick. In concordance, rats select their food in the wild on the basis of taste. In contrast, birds associate feeling sick more with visual factors and accordingly, they naturally select their food in the wild on the basis of appearance rather than taste.

• Kamin’s blocking effect

  • Conditioning of an association between two stimuli, a conditioned stimuls and an unconditioned stimulus is impaired if during the conditioning process the conditioned stimulus is presented together with a second conditioned stimulus that has already been associated with the unconditioned stimulus

  • E.g when dog is exposed to light (1^st conditioned stimulus), together with food (unconditioned stimulus), after repeated pairings of conditioned stimulus and unconditioned stimulus the dog salivates when light comes on.

  • When there are more conditioning trials, this time with light and tone together with unconditioned stimulus, animal doesn’t salivate to tone. This is because association between the tone and the unconditioned stimulus is blocked because an association with another conditioned stimulus already exists

  • Rescorla Wagner model

    • If one CS already fully predicts that US will come, nothing will be learned about the second CS that accompanies the first

Maladaptive learned behaviour

• Anticipatory nausea and vomitting – occurs before chemotherapy drugs are administered and occurs due to classical conditioning

  • Conditioned stimulus: ie the sight of a nurse, is paired with an unconditioned stimulus (chemotheraphy) which reliably produces an unconditioned response (nausea)

  • Following conditioning period (repated chemotherapy treatments) the conditioned stimulus provokes the conditioned response

  • In the situation of patient receiving chemotherapy, the patient finds themselves at treatment location surrounded by unfamiliar sights/sounds/smells. These stimuli also become associated with the chemotherapy treatment and the subsequent nausea and vomitting

  • The best method to avoid development of ANV is to adequately prevent both vomiting and nausea from the first exposure to chemotherapy.

  • If anticipatory side effects develop, behavioral treatment techniques, such as systematic...