Instrumental Conditioning: Reinforcement

edward thorndike
Click the card to flip 👆
1 / 48
Terms in this set (48)
edward thorndike
- an American psychologist who established experimental studies of learning

- animal intelligence

- cat puzzle box
george romanes
- studied animal intelligence by getting anecdotal evidence from pet owners

- cat and parrot feud story

- things wrong with his approach = anthropomorphizing
anthropomorphizing
attributing human characteristics to animals
morgan's cannon
- we should not over-speculate why animals behave a certain way

- we cannot study animals by comparing and relating them to humans

- law of parsimony
law of parsimony
things are usually connected or behave in the simplest or most economical way, especially with reference to alternative evolutionary pathways
trial and error, law of effect
Thorndike's Two-Stage model of learning

Phase 1 = ______________

Phase 2 = ______________ (behavior changes due to consequences)

- satisfying consequence = strong connection

- annoying consequence = weak connection
wolfgang kohler
- known for idea of "insight" learning

- very cognitively oriented - connections are made
bf skinner
- known for radical behaviorism

- said behavior can be predicted by reinforcement & punishment

- skinner box = operant conditioning chamber

- rats & pigeons
maze, t-maze, radial maze, runway
What are the 4 common methods used by psychologists to study instrumental/operant conditionings?
positive reinforcement, negative reinforcement, positive punishment, negative punishment
What are the 4 basic consequences that control behaviors?
positive reinforcementadding a stimulus after a response to increase the rate of responding in the future - adding something desirable to increase the behavior - giving child candy after doing homeworknegative reinforcementsubtracting a stimulus after a response to increase the rate of responding in the future - take away something undesirable to increase the behavior - headache goes away after taking aspirin, more likely to use aspirin in the futurepositive punishmentadding a stimulus after a response to reduce the rate of responding in the future - adding something undesirable to decrease the behavior - spanking a child so he stops screamingnegative punishmentsubtracting a stimulus after a response to reduce the rate of responding in the future - take away something desirable to decrease a behavior - take away TV after child is picking on brotherescape conditioningalready experiencing negative stimuli - if you can initiate a correct response, you can escape from the negative stimuliavoidance conditioningreceived a warning - if you can initiate a correct response, you do not have to experience the negative stimuliprimary reinforcerreinforcers you don't have to learn their values - stimuli required for survival (food and water) - stimuli that are desirablesecondary reinforcerreinforcers that you have to learn their values - money, academic diploma - based on classical conditioningsecondary, suboptimalWhat was shown by Zentall and Laude to study gambling in pigeons? - gambling is based on _______________ reinforcers that signal a big pay-out - both pigeons & humans chose the ______________ choicesocial reinforcersatisfaction we derive from behaviors of other members of our own species - praise, attention, physical contact, facial expression - smiles from adults can reinforce infants 3 months oldpre-commitment paradigmadding extra delay between the decision and reward - increases the likelihood of self-control - make a decision in advance would help reduce the influence of the incentive value of the rewardcontinuous reinforcementevery response is followed by reinforcementpartial reinforcementsome responses followed by reinforcementfixed ratiofixed # of responses before rewardvariable ratiomust emit a varied amount of responses before rewardfixed intervalmust wait a fixed amount of time before rewardvariable intervalmust wait a varied amount of time before rewarddifferential reinforcement of high rate- must emit responses rapidly - speed up responsesdifferential reinforcement of low rate- must emit responses slowly - slow down responsesdifferential reinforcement of other responsestarget an unwanted behavior and reinforce the positive behaviorpost-reinforcement pausea pause in responding that occurs after the delivery of the reinforcerfixed interval, fixed ratiowhen does post-reinforcement pause become more likely to occur? (2 different times)fixed interval scallopgradual increase in the rate of responding with responding occurring at a high rate right before reinforcement is availablerapid acquisitionWhat is one advantage of using continuous reinforcement (CRF)?rapid extinction, costly, lot of timeWhat are 3 disadvantages to using continuous reinforcement (CRF)?VR, VIWhat is the best way to use reinforcement schedule? Start with CRF --> reduce to VR or VI ________ will produce higher rate of responding ______ is better for long-runif ratio is too high, responding will stopWhat is the pitfall of using VR? (variable ratio)positive contrast effectincreased responding after the amount is increasednegative contrast effectreduced responding after the amount is reducedlabor supply curvewhen reinforcement is cut back, we work harder to try to get what we got before - at some point it is not worth it, so we cut back our workstimulus generalizationthe tendency for a conditioned stimulus to evoke similar responsesstimulus discriminationability to distinguish between two stimuligeneralization gradientstimuli similar to the original are not as strong as the original stimulusperceptual learninglearning better perception skills - ex: distinguishing two musical notes from each other * best way to learn = easy --> difficultacquired distinctivenessability to distinguish between stimuli after long exposure and practicefading in proceduremaking the difference extreme, then gradually making the difference more subtleshapingreinforce behaviors that would approximate the final behavior in a way that the reinforced behavior becomes closer and closer to the final behaviorprinciple of successive approximationa series of rewards that provide positive reinforcement for behavior changes that are successive steps toward the final desired behavior