Learning Theory Flashcards
Sensory Memory
Hold info from senses very briefly, generally less than 2 seconds. Does not change with rehearsal
Feature integration theory
Perception of an object as an entity rather than a cluster of unrelated features depends on focused visual attention.
Bandura’s Social Learning Theory & Functional Value
Learn through modeling; a bx has fx value when you can anticipate desirable consequences
Other terms for Classical Conditioning
Respondent (involuntary), Pavlovian, Stimulus-Response
Major Classical Conditioning Theorists
Pavlov & John Watson
Classical cond: US –> UR
Unconditioned Stim–>Unconditioned Resp;
Lights–> Pupil Dilation (innate,universal resp)
Class cond: CS–> CR
Cond Stim –> Cond Resp (result of experience & learning, not universal)
Ice Cream Truck music–> excited
How is conditioned response learned?
Repeated pairing of a neutral stim w/unconditioned stimulus
Pavlov: Tone (NS) repeatedly paired with Meat (US), over time produces salivation (CR) when presented alone
*CR same as UR, but may be of less magnitude than the UR
How can you tell if a stim is conditioned versus unconditioned?
Ask yourself: Is the resp to the stim universal for everyone?
Ex: fishnet stockings– not everyone experiences arousal, so conditioned stim
2 key factors in classical conditioning
Contiguity: closeness in time/temporal sequence of presentation of US & NS
Contingency: US must appear to depend or be contigent upon CS (so NS signals coming of the US)
Methods of Class Cond: Delay Conditioning/Standard Pairing
CS (orig the NS) precedes the US by a very short interval
Tone (CS) sounded .5 sec before meat powder (US)(contiguity), US appears dependent of CS (contingency)
Methods of Class Cond: Trace Conditioning
CS (orig the NS) precedes the US by a period of time and stops before the US (not as short a time interval as Standard Pairing)
Dog food bucket opens (CS), then Hank gets food (US)
Methods of Class Cond: Temporal Conditioning
US presented repeatedly at same time, such that time itself becomes the CS
Hank fed everyday when we go downstairs, such that morning wakeup becomes CS
Methods of Class Cond: Simultaneous Conditioning
NS & US completely overlap
Tone (NS) sounded at same time meat powder (US)presented
*NO LEARNING takes place, because there is no contingency established
Methods of Class Cond: Backward Conditioning
US presented before the NS
Meat powder (US) presented before tone (NS)
*NO LEARNING takes place, the NS will never elicit UR because contingency is not estab and contiguity is not sufficient
Stimulus Generalization
Classical cond term; Occurs when a subject automatically demonstrates a CR to stimuli that are similar to the CS
Ex: Watson’s Little Albert experiment: Albert conditioned to fear white rat (CS) then automatically feared white rabbit & santa claus mask
Higher Order Conditioning
Classical Cond Term; Deliberate pairing of CS with another NS until the new NS becomes another CS and elicits CR
Ex: Pavlov… Tone (CS1) paired repeatedly with flash of light (new NS) until light alone elicits CR so it becomes CS2. Known as 2nd Order Conditioning
*Impossible to condition beyond 3rd order
Classical Extinction
CS repeatedly presented without the US
Ex; Little Albert study…white rat (CS) repeatedly presented without loud noise (US), eventually fear response eliminated
Spontaneous Recovery
During extinction trials, CR briefly reappears to CS
Ex: Dog undergoes extinction trials one day and stops salivating to tone. Next day returns to lab and initially salivates (CR) when tone (CS) presented. CR will vanish again in extinction trials continue
Stimulus Discrimination
Subject learns to discriminate between 2 similar neutral stimuli because only one of them has been paired with US
Ex: Discrimination between 500 Hz tone and 100 Hz tone
Experimental Neurosis
Related to Stimulus Discrimination; Occurs when two stimuli are too similar for the subject to distinguish, such that subject become agitated. If attempt to return to original discrimination that was mastered, subject will no longer be able to discriminate (as if they had a break down and cannot function any longer)
Pseudoconditioning
NS is NOT DELIBERATELY paired with US or CS comes to elicit CR over time
Light to room turned on before US or CS presented such that dog salivates (CR) to light
Habituation
After repeated exposure to US, it no longer elicits UR
Ex: person lives by a railway and over time noise of passing trains (US) no longer produces startle response (UR)
*Habituation is always to an US, NOT the CS and is not possible with all US (for example extreme electric shock)
Definition of Operant Conditioning Theory
Explains voluntary behavior; Posits that we learn as a result of reward & punishment of behaviors
Other names for Operant Cond Theory
Skinnerian conditioning and Instrumental Conditioning (bx is instrumental or goal directed toward obtaining rewards)
2 Major Operant Cond Theorists
EL Thorndike & BF Skinner
Thorndike’s Law of Effect
Bxs initially emitted in random trial & error fashion. If followed by pleasurable consequences, bx gets stronger & more frequent (reward). If followed by unpleasant consequences, bx gets weaker & less freq (punishment)
Reinforcement
Always increases target bx; brings subject to more desirable state
Punishment
Always decreases target bx ; brings subject to less desirable state
Positive; Negative
Something added; Something taken away
Positive Reinforcement
Based on REWARD; something added following target bx
Ex: Elise goes potty in public place, receives jelly bean–>inc. likelihood of going potty
Negative Reinforcement
Based on RELIEF; something annoying is removed following target bx
Ex: Elise covers ear when potty flushes, noise is not so loud/annoying–>inc likelihood of covering ears in restroom
Positive Punishment
Based on PAIN; something aversive added after target bx
Ex: Elise pushes buttons on remote, gets scolded–>less likely to push buttons without permission
Negative Punishment
Based on LOSS; something valuable removed following target bx
Ex: Elise steps off stool while I’m brushing her hair, I turn off video–>less likely to step off stool
3 questions to determine type of operant cond
- What is the target bx?
- What occurs after the target bx? Is something added or taken away?
- Is subject brought to more desirable state/more likely to emit target bx or to less desirable state/less likely to emit target bx?
Schedules of Reinforcement: Acquisition Phase
period during which new learning occurs
Extinction
period during which reinforcement is withheld
Operant Strength
term for strength of a behavior, measured by rate of responding; generally depends on which schedule of reinforcement is used
2 broad categories for Schedules of Reinforcement
Continuous & Intermittent
Continuous Reinforcement
Reinforcing every occurrence of the bx; best for acquiring new bxs
Ex: When first potty training, Elise received m&m after every trip to potty
Over time, continuous reinforcement leads to ____, so it is best to change to an intermittent sched of reinforcement, which refers to a process called____
Satiation, where reinforcer loses its value due to overuse; Thinning
2 broad types of intermittent reinforcement
Ratio & Interval
Ratio Schedule
Reinf based on how often (frequency) of target bx
Interval Schedule
Reinf based on a certain interval of time elapsing (provided subject has emitted target bx)
2 types of schedules of reinforcement (applies to both ratio & interval)
Fixed (ratio or interval does not change/consistent/predictable) vs Variable (ratios or intervals do change/inconsistent/unpredictable)
What schedule of reinforcement is a bi weekly paycheck?
Fixed Interval
What schedule of reinforcement is a slot machine?
Variable ratio
Fixed Interval schedule of Reinf
Reinf occurs when target bx emitted after the fixed interval of time has elapsed; response rate tends to be low during most of the interval and inc sig at the end of the interval
Variable Interval Sched of Reinforcement
reinf occurs after an unpredictable interval of time has elapsed, such that subject can’t anticipate when reinf will occur. So, sub performs bx at moderate rate without pause
Fixed Ratio Sched of Reinforcement
Reinf after a certain, unchanging # of responses have been emitted; response rate tends to be moderate to high, subject may pause after reinf is provided esp if many responses are required to obtain reinf
Ex: Assembly Line, earn money for every 50 garments completed
Variable Ratio Sched of Reinforcement
Reinf after unpredictable # responses emitted; response rate tends to be high with little pauses
Ex: slot machines
Which sched of reinforcement has strongest response rate?
Variable Ratio
Impact of fixed schedules on bx
increase in bx before reinf, drops after reinf
Impact of variable schedules on bx
smooth & steady before & after reinf
During acquisition, List operant strength of the different reinf schedules from strongest to weakest
*Hint: Linking reinf to bx is stronger than linking to time
Unpredictability keeps sub trying harder than predictability
- Variable Ratio
- Fixed Ratio
- Variable Interval
- Fixed Interval
During extinction, list reinf schedules from most resistant to extinction and least
*Hint: Follows same pattern as operant strength during acquisition
- Variable Ratio
- Fixed Ratio
- Variable Interval
- Fixed Interval
Operant Extinction
Ceasing to reinf bx that was previous reinf
Response Burst
During operant extinction, withholding reinf results initially in an increase in the previously reinf bx
Ex; Whining…when parents initially stops responding to whining bx, the child tends to whine more before the bx begins to diminish
Superstitious Bx
Results from non contingent or accidental reinf, where reinf is arbitrary and inconsistent
Ex: practice of “knocking on wood” to avoid TA filling
Discrimination Learning/Stimulus control
Target bxs reinf in certain circumstances and not others, such that subject learns to discriminate between situations
Ex: Elise expects jelly bean for eating meals at home, but not at school; Use curse words at home, but not at work
In Discrimination Learning, the stim that signals reinf is known as the __________
Discriminative Stim (S_d)
In Discrimination Learning, the stim that signals reinf will NOT take place is known as the _______
*Hint: Delta sounds like delete, as in reinf is deleted
S delta
Stimulus Generalization
Subject emits target bx in presence of stimuli similar, but not exactly the same as discriminative stim
Ex: slowing down when you see any emergency vehicle after being pulled over by police
Response Generalization
Subject performs a bx similar to a previously reinforced bx, hoping for reinforcement
Ex: Elise receives praise for cleaning up toys downstairs, then later cleans upstairs
Prompting
Cueing the subject regarding what bx to perform
Ex: While teaching Elise manners, initially told her “say thank you”
___ is a gradual reduction in prompting
Fading
Ex: “Say Thank You” becomes “What do you say?” in teaching manners
What is shaping?
Teaching a subject to emit a bx through reinforcing bxs that are closer and closer to the desired bx, known as using SUCCESSIVE APPROXIMATIONS
Ex: Teaching ABCs, may start with sounds, then individual letters, then letters in order
Chaining
Stringing together behaviors to accomplish a goal. Each behavior in the chain is mildly reinforced and serves as a cue to perform the next bx in the chain. Major reinforcement comes at the end of the chain.
Ex: Steps of using potty, ultimate reinforcement after completion of all steps
Ex: Going to movies
Premack Principle
Hint: Grandma’s Rule
A high frequency bx (implies strong reinforcing value) used to reinforce low freq bx (weaker reinforcing value)
Ex: Reach end of EPPP section, watch a show
Ex: Grandma’s rule- eat your veggies, go outside & play
Behavioral Contrast
When 2 bxs were initially equally reinforced, then stop reinforcing one of them. The bx that is still reinf tends to increase freq, bx no longer reinf decreases freq
Ex: Hank initially reinf for both sitting & shaking hands. Stopped reinf shaking hands, so he no longer tries to shake
Alternate term for Social Learning Theory
Observational learning
Social Learning Theory
Proposes that learning, particularly complex learning, occurs through observation & modeling of bx
Primary social Learning Theorist & Major Study
Bandura- Bobo Doll Study: children exposed to aggressive models imitated that bx, particularly when the model was high status, nurturant, and same sex as the observer
Main difference between Bandura SLT & classical/operant learning theories
Perform bxs not because we were reinforced in the past, but because we anticipate reward in the future. Social learning incorporates the COGNITIVE activity of the learner more so than classical & operant theories
Reciprocal Determinism
Social Learning Theory term; Bx regulated by an interactive triad of:
1. Person
2. Bx
3. Environment
Ex: Elise anticipates interaction if she poops in pull up at night rather than potty, so she waits until bedtime to poop. Attention received leads her to put even less effort into going on the toilet, which diminishes her motivation to apply potty training skills
4 steps of observational learning
- Attention (to model)
- Retention (remembering what is seen/heard)
- Production (imitating what was seen/heard)
- Motivation (reinforcement for accurate performance)
Ex: May explain why Elise has not cussed repeatedly even after modeling me using curse words, because I ignored/did not reinforce
Mediated generalization
another term for stimulus generalization
