unit 4 learning Flashcards
learning
long-lasting change in behavior resulting from experience
classical conditioning
Ivan Pavlov) dogs learned to pair sounds w/ food
- deduced basic principle of classical conditioning: ppl & animals can learn to associate neutral stimuli (e.g. sound) w/ stimuli that produces reflexive, involuntary responses (e.g. food) and will learn to respond similarly to new stimulus as they did to old one (e.g. salivate)
unconditioned stimulus (US / UCS) - classical conditioning
original stimulus that elicits a natural, reflexive response (e.g. food)
unconditioned response (UR / UCR) - classical conditioning
natural, involuntary response to unconditioned stimulus (e.g. salivation)
conditioned stimulus (CS) - classical conditioning
when associating unconditioned stimulus w/ neutral stimulus (e.g. sound) together -> conditioned response (CR) is no longer neutral stimulus but rather conditioned stimulus (CS) when bell elicits salvation
acquisition
when learning takes place, animals respond to CS w/o presence of US
- animals acquired a new behavior
- repeated pairings of CS & US -> stronger CR
- most effective: present CS first and introduce US while CS is still evident [delayed conditioning]
less effective methods of learning than delayed conditioning
- trace conditioning: presence of CS, followed by short break, followed by presence of UC
- simultaneous conditioning: CS & US presented at the same time
- backward conditioning: US presented first & followed by CS
extinction
process of unlearning a behavior
- when CS no longer elicits CR
- achieved by repeatedly presenting CS w/o US -> breaks association
spontaneous recoery
after CR has been extinguished & no further training, response briefly reappears upon presentation of CS
generalization
animals conditioned to a certain stimulus also respond to similar stimuli (response smaller in magnitude)
discrimination
subjects trained to discriminate (distinguish) btwn bells -> only salivate to specific bell
classical conditioning w/ humans - little Albert
John B. Watson & Rosalie Rayner: conditioned little Albert to fear white rat
- originally liked white rat but repeated pairing w/ loud noise -> cry when he saw the rat
- US: loud noise-> elicits involuntary, natural response of fear
- CS: rat
- CR: crying in response to rat alone
- generalized: crying to other white, fluffy things
aversive conditioning
conditioned to have negative response
second-order / higher-order conditioning
once CS elicits CR, use the CS as a US in order to condition a response to a new stimulus
e.g. training) presence of bell (CS) + light (new!) = salvation -> acquisition) presence of light = salvation
learned taste aversions
example of animals & humans being biologically prepared to make certain conditions more easily than others
- e.g. eating food -> sick // avoiding that food
- result in powerful avoidance responses on basis of single pairing
- adaptive (helpful for survival) bc helps avoid dangerous things in future
- CS (food) must be salient (outstanding) in order for us to learn to avoid it
John Garcia & Robert Koelling - Garcia effect
experiment on how rats learned to make certain associations more than others
- rats learned to associate noise w/shock & unusual tasting water w/ nausea BUT no connection btwn noise & nausia // water & shock
operant conditioning
learning based on association of consequences w/ one’s behaviors
law of effect - Edward Thorndike
if consequences of behavior are pleasant, stimulus-response (S-R) connection will be strengthened & likelihood of behavior will increase (instrumental 수단이 되는 learning)
Skinner box - reinforcement
B.F. Skinner
- way to deliver food to an animal & lever to press/disk to peck in order to get the food
- reinforcer: makes behavior more likely to occur (food)
- reinforcement: defined by its consequences (process of giving food)
- positive reinforcement: addition of smth pleasant
- negative reinforcement: removal of smth unpleasant (e.g. loud noise terminates when pressing lever) -> results in escape learning: allows one to terminate aversive stimulus
- vs. avoidance learning: allows one to avoid unpleasant stimulus altogether
punishment
affecting behavior by using unpleasant consequences, makes behavior less likely
- positive punishment: addition of smth unpleasant
- negative punishment (omission training): removal of smth pleasant
shaping
reinforces steps used to reach the desired behavior
- rewards approximations of desired behavior -> increases likelihood of desired behavior
chaining
animals taught to perform a number of responses successively to get a reward
primary reinforcer
rewarding
secondary reinforcer
things we’ve learned to value
- generalized reinforcer: can be traded for virtually anything (special, e.g. money) -> practical application: token economy where every time someone perform desired behavior, they’re given token -> allowed to train token for any variety of reinforcers
