Learning

Cards (54)

  • Animals learn by association through associative learning. The brain connects events that occur in sequence and behaviours become associated with their context.
  • Conditioning is the process of learning associations. There are two main types of conditioning: classical conditioning and operant conditioning.
  • In classical conditioning, the organism learns to associate 2 stimuli and thus anticipate events.
  • In operant conditioning, the organism learns to associate a response or behaviour to its consequence.
  • Ivan Pavlov was the first to study classical conditioning. Pavolov's experiment was aimed to study if an animal would learn a link between 2 stimuli. Before placing food in a dog's mouth, Pavolov would sound a tone. After several pairing of tone and food, the dog began salivating at the sound of the tone only(without sensing the food).
  • An unconditioned response is a response that occurs without any conditioning and does not need to be learned (eg. a dog salivating at the smell of meat).
  • An unconditioned stimulus is a stimulus that produces a response without any conditioning(eg. meat for a dog)
  • A conditioned response is a response that occurs after conditioning has occurred (eg. a dog salivating at the sound of a tone).
  • A conditioned stimuli is a stimuli that produces some response after conditioning has occurred (eg. the tone that causes the dog to salivate).
  • There are 5 major conditioning processes: acquisition, extinction & spontaneous recovery, generalization, and discrimination.
  • In acquisition, the CS must come immediately before the US(most of the times with some exception) in order to lead to conditioning. Higher-order conditioning can occur when a new neutral stimulus becomes a new conditioned stimulus. For example if a tone regularly signals food, then regularly flashing a light before sounding the tone can also signal food; the light becomes the second-level or higher-order stimulus. Higher-order conditioning tends to be weaker than first-stage conditioning.
  • Associations can influence attitudes. Conditioned likes and dislikes are even stronger when subjects notice or are aware of the associations they have learned.
  • In extinction and spontaneous recovery, the CS begins to occur repeatedly without being followed by a US for a period of time. As a result the CR is elicited less and less in a process called extinction. However, even after a break of not presenting the CS for some time and then suddenly presenting it again will cause a CR again, though weakened.
  • In generalisation, organisms learn to respond to some degree of similar stimuli to the paired CS. Emotional reactions to one stimulus can generalise to other similar stimuli. For example, after a dog has been conditioned to salivate to being rubbed, it will also drool a bit to being scratched.
  • In discrimination, an organism can learn to distinguish between a conditioned stimulus and other irrelevant stimuli. Organisms can also learn that slightly varying stimuli can cause drastically varying consequences.
  • Animals can learn the predictability of an event: the more predictable the association, the stronger the conditioned response.
  • Learned helplessness: Animals that are repeatedly faced with traumatic events and have no way of avoiding them come to feel hopeless and depressed and do not make efforts at avoiding the conflict when later presented the same situation but with a way of avoiding it.
  • Conditioning relies on both biological and cognitive processes.
  • In biological processes, an animal's capacity for conditioning is limited by its biology and influenced by its predisposition to learn associations that will enhance their survival.
  • An exception to the acquisition rule stating that the US must come immediately after the CS is that some associations can be learned even if the US follows the CS hours later(eg. getting food poisoning 4 hours after eating seafood can still cause an association to dislike for the smell of seafood afterwards).
  • The sense(touch, sight, smell, etc.) to which different species more readily develop associations with also varies.
  • There may also be a natural association between a conditioned stimulus and the unconditioned stimulus that it predicts(eg. the colour red being associated with women's sexuality).
  • In operant conditioning, organisms learn to associate their actions with a consequence. Their behaviour operates on the environment to produce a rewarding or punishing stimulus.
  • Skinner's Experiment involved placing an animal inside a box that contained a bar and a key, one which produced food and the other produced water. The experimenter on the outside would record the animal's response.
  • Shaping behaviour is the use of reinforcers to guide an animal's actions toward a wanted behaviour. Successive approximations is the process of rewarding an animal each time it gets closer to performing the final desired action until it has reached the final goal.
  • The law of effect states that rewarded behaviour is likely to incur.
  • Discriminative stimulus is a stimulus that will elicit a response after association through reinforcement and contrasted with other stimuli that are not reinforced and thus will not elicit a response.
  • Positive reinforcement strengthens a response by presenting a typically pleasurable stimulus such as food or praise.
  • Negative reinforcement strengthens a response by reducing or removing an unpleasant or undesirable stimulus(ie. it welcomes the end of something bad).
  • Primary reinforcers are unlearned, such as getting food to satiate hunger.
  • Conditioned reinforcers are also called secondary reinforcers and are achieved through association with some primary reinforcer, such as desire for money to buy food to satiate hunger.
  • Immediate and delayed reinforcers have different effects depending on how complex the animal is. Humans, unlike some other animals learn to respond to delayed reinforcers. However, sometimes, smaller but more immediate consequences override our judgement to pursue larger but delayed consequences.
  • Continuous reinforcement is the act of reinforcing the desired response each time it occurs. In this case, initial learning occurs faster but extinction also occurs faster.
  • Partial reinforcement, also known as intermittent reinforcement is the act of periodically or unpredictably reinforcing responses. In this case, initial learning is faster but produces a greater resistance to extinction.
  • Fixed-ratio schedule: reinforces behaviour after a fixed number of responses. For example, a coffee shop rewarding customers with a cup of free coffee after 30 purchases.
  • Variable-ratio schedule: provide reinforcers after an unpredictable number of responses, for example in gambling.
  • Fixed-interval schedule: reinforcing the first response after a fixed amount of time. For example, people tending to check the mailbox more often, knowing that the mail delivery date is approaching.
  • Variable-interval schedule: reinforcing the first response after a varying amount of time. For example, finally receiving mail after checking for it for some amount of time without knowing specifically when the mail will arrive.
  • Ratio schedules produce higher response rates than interval schedules. The more predictable the reinforcement schedule, the higher the response rate, thus why fixed schedules produce higher response rates than variable schedules.
  • A punisher is any consequence that decreases the frequency of a preceding behaviour.