Download - 1. Learning1
-
8/10/2019 1. Learning1
1/53
PowerPointPresentationby Jim Foley
2013 Worth Publishers
Chapter 7Learning
-
8/10/2019 1. Learning1
2/53
Types of Learning
Classicalconditioning:learning to link twostimuli in a way thathelps us anticipate
an event to whichwe have a reaction
Operantconditioning:
changing behaviorchoices in response
to consequences
Cognitive learning:acquiring newbehaviors and
information throughobservation andinformation, rather
than by directexperience
-
8/10/2019 1. Learning1
3/53
How it works: after repeatedexposure to two stimulioccurring in sequence, weassociate those stimuli with eachother.
Result: our natural response toone stimulus now can betriggered by the new, predictivestimulus.
Associative Learning:Classical Conditioning
Here, our response tothunder becomesassociated withlightning.
Stimulus 1: Seelightning
Stimulus 2: Hearthunder
After Repetition
Stimulus: See lightning
Response: Cover ears to avoid sound
-
8/10/2019 1. Learning1
4/53
Child associates his response (behavior) with consequences. Child learns to repeat behaviors (saying please) which were
followed by desirable results (cookie).Child learns to avoid behaviors (yelling gimme!) which were
followed by undesirable results (scolding or loss of dessert).
Associative Learning:Operant Conditioning
-
8/10/2019 1. Learning1
5/53
Cognitive LearningCognitive learning refers to acquiring new behaviors
and information mentally, rather than by directexperience.Cognitive learning occurs:1.by observing events and the behavior of others.
2.by using language to acquire information aboutevents experienced by others.
-
8/10/2019 1. Learning1
6/53
BehaviorismThe term behaviorism was used by John B. Watson
(1878-1958), a proponent of classical conditioning,as well as by B.F. Skinner (1904-1990), a leader inresearch about operant conditioning. Both scientists believed the mental life was much
less important than behavior as a foundation forpsychological science.
Both foresaw applications in controlling humanbehavior:
Skinner conceived ofutopian communities.
Watson went intoadvertising.
-
8/10/2019 1. Learning1
7/53
Ivan Pavlovs Discovery
While studying salivation indogs, Ivan Pavlov found thatsalivation from eating foodwas eventually triggered bywhat should have beenneutral stimuli such as:
just seeing the food.
seeing the dish.
seeing the person whobrought the food.
just hearing that persons
footsteps.
-
8/10/2019 1. Learning1
8/53
Before Conditioning
No response
Neutralstimulus
(NS)
Neutral stimulus :a stimulus which does not trigger a response
-
8/10/2019 1. Learning1
9/53
Unconditionedresponse (UR):dog salivatesUnconditioned
stimulus (US):yummy dog food
Before ConditioningUnconditioned stimulus and response :
a stimulus which triggers a response naturally,before/without any conditioning
-
8/10/2019 1. Learning1
10/53
Unconditionedresponse (UR):dog salivates
Neutralstimulus
(NS)Unconditionedstimulus (US)
During ConditioningThe bell/tone (N.S.) is repeatedly presented with
the food (U.S.).
-
8/10/2019 1. Learning1
11/53
Conditionedresponse:
dog salivates
After Conditioning
Conditioned(formerlyneutral)stimulus
The dog begins to salivate upon hearing the tone(neutral stimulus becomes conditioned stimulus).
Did you follow the changes?The UR and the CR are thesame response, triggered bydifferent events.
The difference iswhether conditioningwas necessary for theresponse to happen.
The NS and the CS are thesame stimulus.The difference iswhether the stimulustriggers the conditionedresponse.
-
8/10/2019 1. Learning1
12/53
Find the US, UR, NS, CS, CR in the following:
Your romantic partner always uses the same
shampoo. Soon, the smell of that shampoo makesyou feel happy.
The door to your house squeaks loudly when youopen it. Soon, your dog begins wagging its tail whenthe door squeaks.
The nurse says, This wont hurt a bit, just beforestabbing you with a needle. The next time you hear
This wont hurt, you cringe in fear.
You have a meal at a fast food restaurant that causesfood poisoning. The next time you see a sign for thatrestaurant, you feel nauseated.
-
8/10/2019 1. Learning1
13/53
If the dog becomes conditioned to salivate atthe sound of a bell, can the dog beconditioned to salivate when a lightflashesby associating it with the BELLinstead of with food?
Yes! The conditioned response can betransferred from the US to a CS, then fromthere to another CS.This is higher-order conditioning : turning aNS into a CS by associating it with anotherCS.A man who was conditioned to associate joywith coffee, could then learn to associate joywith a restaurant if he was served coffeethere every time he walked in to therestaurant.
Higher-Order Conditioning
-
8/10/2019 1. Learning1
14/53
14
Acquisition
What gets acquired? The association between a neutral
stimulus (NS) and an unconditionedstimulus (US).How can we tell that acquisition has
occurred? The UR now gets triggered by a CS
(drooling now gets triggered by a bell).
Timing
For the association to be acquired,the neutral stimulus (NS) needs torepeatedly appear before theunconditioned stimulus (US)about ahalf-second before, in most cases. The
bell must come right before the food.
Acquisition refers to the initialstage of learning/conditioning.
-
8/10/2019 1. Learning1
15/53
Acquisition and Extinction
The strength of a CR grows with conditioning.Extinction refers to the diminishing of a conditioned response. Ifthe US (food) stops appearing with the CS (bell), the CR decreases.
-
8/10/2019 1. Learning1
16/53
Spontaneous Recovery [Return of the CR]
After a CR (salivation) has been conditioned and then extinguished: following a rest period, presenting the tone alone might lead to aspontaneous recovery (a return of the conditioned response despite alack of further conditioning) . if the CS (tone) is again presented repeatedly without the US, the CR
becomes extinct again.
-
8/10/2019 1. Learning1
17/53
Generalization and DiscriminationPlease notice the narrow, psychological definition .
Ivan Pavlov conditioneddogs to drool whenrubbed; they then alsodrooled when scratched.
Ivan Pavlov conditioned dogsto drool at bells of a certainpitch; slightly differentpitches did not triggerdrooling.
Generalization refers to thetendency to have
conditioned responsestriggered by related stimuli.
MORE stuff makes you drool.
Discrimination refers to thelearned ability to only
respond to a specific stimuli,
preventing generalization.
LESS stuff makes you drool.
-
8/10/2019 1. Learning1
18/53
Insights aboutconditioning ingeneral It occurs in all
creatures. It is related tobiological drivesand responses.
Insights aboutscience Learning can be
studiedobjectively, byquantifyingactions andisolatingelements ofbehavior.
Insights fromspecificapplications Substance abuse
involvesconditionedtriggers, andthese triggers(certain places,
events) can beavoided orassociated withnew responses.
Ivan Pavlovs Legacy
-
8/10/2019 1. Learning1
19/53
John B. Watson and ClassicalConditioning: Playing with Fear
In 1920, 9-month-old Little Albert was not afraidof rats.
John B. Watson and Rosalie Rayner then clangeda steel bar every time a rat was presented toAlbert.
Albert acquired a fear of rats, and generalizedthis fear to other soft and furry things.
Watson pridedhimself in his abilityto shape peoplesemotions. He laterwent intoadvertising.
-
8/10/2019 1. Learning1
20/53
BeforeConditioning
NS: rat
No fear
UCS: steel bar hitwith hammer
Natural reflex:fear
Little Albert Experiment
-
8/10/2019 1. Learning1
21/53
DuringConditioning
NS: rat UCS: steel bar hitwith hammer
Natural reflex:fear
Little Albert Experiment
-
8/10/2019 1. Learning1
22/53
AfterConditioning
NS: rat
Conditioned
reflex:fear
Little Albert Experiment
-
8/10/2019 1. Learning1
23/53
How it works:
An act of chosen behavior (aresponse) is followed by areward or punitive feedbackfrom the environment.Results:
Reinforced behavior is morelikely to be tried again.Punished behavior is less likely
to be chosen in the future.
Operant Conditioning
Response:balancing a ball Consequence:receivin food Behaviorstren thened
Operant conditioning involves
adjusting to the consequences of ourbehaviors, so we can easily learn todo more of what works, and less ofwhat doesnt work. Examples
We may smile more at work afterthis repeatedly gets us bigger tips.
We learn how to ride a bike usingthe strategies that dont make uscrash.
-
8/10/2019 1. Learning1
24/53
-
8/10/2019 1. Learning1
25/53
B.F. Skinner: Behavioral ControlB. F. Skinner saw potential forexploring and using EdwardThorndikes principles much morebroadly. He wondered:
how can we more carefully
measure the effect ofconsequences on chosenbehavior?
what else can creatures be taughtto do by controllingconsequences?what happens when we changethe timing of reinforcement?
B.F. Skinnertrained pigeons toplay ping pong,and guide a videogame missile.
-
8/10/2019 1. Learning1
26/53
B.F. Skinner: The Operant ChamberB. F. Skinner, like Ivan Pavlov, pioneered more controlled
methods of studying conditioning.The operant chamber, often called the Skinner box,allowed detailed tracking of rates of behavior change inresponse to different rates of reinforcement.
Recording
device
Bar or leverthat an animal
presses,
randomly atfirst, later forreward
Food/water dispenserto provide the reward
-
8/10/2019 1. Learning1
27/53
Reinforcement
Reinforcement refers toany feedback from theenvironment that makesa behavior more likelyto recur.
Positive (adding)reinforcement:adding somethingdesirable (e.g.,warmth)Negative (takingaway) reinforcement:ending somethingunpleasant (e.g., the
cold)
For the meerkat,
this warm light isdesirable.
This meerkat has justcompleted a task outin the cold
-
8/10/2019 1. Learning1
28/53
A cycle of mutualreinforcement
28
Children who have a temper tantrumwhen they are frustrated may get
positively reinforced for this behaviorwhen parents occasionally respond bygiving in to a childs demands.
Result : stronger, more frequenttantrums
Parents who occasionally give in totantrums may get negativelyreinforced when the child responds byending the tantrum.
Result : parents giving-in behavioris strengthened (giving in soonerand more often)
-
8/10/2019 1. Learning1
29/53
Discrimination
Discrimination refers to the abilityto become more and more specificin what situations trigger aresponse.Shaping can increasediscrimination, if reinforcementonly comes for certaindiscriminative stimuli.For examples, dogs, rats, and evenspiders can be trained to search forvery specific smells, from drugs toexplosives.Pigeons, seals, and manatees havebeen trained to respond to specificshapes, colors, and categories.
Bomb-finding rat
Manatee thatselects shapes
-
8/10/2019 1. Learning1
30/53
How often should we reinforce?
Do we need to give a reward every single time? Or isthat even best?B.F. Skinner experimented with the effects of givingreinforcements in different patterns or schedulesto determine what worked best to establish andmaintain a target behavior.In continuous reinforcement (giving a reward afterthe target every single time), the subject acquires thedesired behavior quickly.In partial/intermittent reinforcement (givingrewards part of the time), the target behavior takeslonger to be acquired/established but persists longerwithout reward.
-
8/10/2019 1. Learning1
31/53
Fixed interval schedule: reward every hourVariable interval schedule:reward after a changing/randomamount of time passes
We may scheduleour reinforcements
based on aninterval of time
that has gone by.
Fixed ratio schedule: reward every five targeted behaviorsVariable ratio schedule: rewardafter a randomly chosen instanceof the target behavior
We may plan for a
certain ratio ofrewards pernumber of
instances of thedesired behavior.
Different Schedules ofPartial/Intermittent Reinforcement
Which Schedule of Reinforcement is This?
-
8/10/2019 1. Learning1
32/53
Which Schedule of Reinforcement is This?Ratio or Interval?Fixed or Variable?
1. Rat gets food every third time it presses the lever
2. Getting paid weekly no matter how much work is done
3. Getting paid for every ten boxes you make
4. Hitting a jackpot sometimes on the slot machine5. Winning sometimes on the lottery you play once a day
6. Checking cell phone all day; sometimes getting a text
7. Buy eight pizzas, get the next one free
8. Fundraiser averages one donation for every eight housesvisited
9. Kid has tantrum, parents sometimes give in
10. Repeatedly checking mail until paycheck arrives
FRFIFR
VRVI/VRVIFR
VR
VRFI
-
8/10/2019 1. Learning1
33/53
Rapid respondingnear time for
reinforcement
Fixed interval
Rapidrespondingnear time for
reinforcement
Fixed interval
Results of the different schedules of reinforcementWhich reinforcements produce more
responding (more target behavior)?
Fixed interval: slow,unsustained respondingIf Im only paid for mySaturday work, Im notgoing to work as hard onthe other days.
Variable interval : slow,consistent respondingIf I never know which daymy lucky lottery numberwill pay off, I better play itevery day.
Steadyresponding
Variable interval
Eff ti f th ti h d l f
-
8/10/2019 1. Learning1
34/53
Reinforcers
Effectiveness of the ratio schedules ofReinforcement
Fixed ratio : high rate ofrespondingBuy two drinks, get one
free? Ill buy a lot of them! Variable ratio : high,consistent responding,even if reinforcementstops (resists extinction)If the slot machine
sometimes pays, Ill pullthe lever as many times as
possible because it may pay this time!
Variable ratio
Fixed ratio
-
8/10/2019 1. Learning1
35/53
Operant Effect: PunishmentPunishments have the opposite effects of reinforcement.
These consequences make the target behavior less likelyto occur in the future.
+ PositivePunishment
You ADD somethingunpleasant/aversive(ex: spank the child)
- NegativePunishment
You TAKE AWAYsomething pleasant/
desired (ex: no TVtime, no attention)--
MINUS is thenegative here
Positive does not mean good or desirable and
negative does not mean bad or undesirable.
-
8/10/2019 1. Learning1
36/53
When is punishmenteffective?
Punishment works best in naturalsettings when we encounterpunishing consequences fromactions such as reaching into a fire;in that case, operant conditioning
helps us to avoid dangers.Punishment is effective when wetry to artificially create punishingconsequences for others choices;these work best when
consequences happen as they doin nature.Severity of punishments is not
as helpful as making thepunishments immediate and
certain .
-
8/10/2019 1. Learning1
37/53
Punished behaviors may restart whenthe punishment is over; learning is notlasting.Instead of learning behaviors , the childmay learn to discriminate amongsituations , and avoid those in whichpunishment might occur.Instead of behaviors , the child mightlearn an attitude of fear or hatred ,which can interfere with learning. Thiscan generalize to a fear/hatred of alladults or many settings.Physical punishment models aggressionand control as a method of dealingwith problems.
Applying operant conditioning to parentingProblems with Physical Punishment
-
8/10/2019 1. Learning1
38/53
Dont think about the beach
Dont think about the waves, thesand, the towels and sunscreen,the sailboats and surfboards.Dont think about the beach.
Are you obeying theinstruction? Would you obeythis instruction more if youwere punished for thinkingabout the beach?
-
8/10/2019 1. Learning1
39/53
Problem:Punishing focuses on what NOT to do, which does notguide people to a desired behavior.
Even if undesirable behaviors do stop, anotherproblem behavior may emerge that serves the samepurpose, especially if no replacement behaviors aretaught and reinforced.
Lesson:In order to teach desired
behavior, reinforce whatsright more often thanpunishing whats wrong.
-
8/10/2019 1. Learning1
40/53
More effective forms of operant conditioning The Power of Rephrasing
Positive punishment: Youreplaying video games instead ofpracticing the piano, so I am
justified in YELLING at you.
Negative punishment: Youreavoiding practicing, so Im turningoff your game. Negative reinforcement: I willstop staring at you and buggingyou as soon as I see that you arepracticing. Positive reinforcement: Afteryou practice, well play a game!
-
8/10/2019 1. Learning1
41/53
Summary: Types of Consequences
Adding stimuli Subtract stimuli Outcome
Positive +Reinforcement
(You get candy)
Negative Reinforcement
(I stop yelling)
Strengthenstarget behavior
(You do chores)Positive +
Punishment
(You get spanked)
Negative Punishment
(No cell phone)
Reduces targetbehavior
(cursing)
= uses desirablestimuli
= uses unpleasantstimuli
M O t C diti i A li ti
-
8/10/2019 1. Learning1
42/53
More Operant Conditioning ApplicationsParenting
1.Rewarding small improvements toward desired behaviors worksbetter than expecting complete success, and also works better thanpunishing problem behaviors.2.Giving in to temper tantrums stops them in the short run butincreases them in the long run.
Self-ImprovementReward yourself for steps youtake toward your goals. As youestablish good habits, thenmake your rewards moreinfrequent (intermittent).
R l f Bi l i C diti i
-
8/10/2019 1. Learning1
43/53
Role of Biology in Conditioning
Classical ConditioningJohn Garcia and others found it was easierto learn associations that make sense forsurvival.Food aversions can be acquired even if theUR (nausea) does NOT immediately followthe NS. When acquiring food aversionsduring pregnancy or illness, the bodyassociates nausea with whatever food waseaten.
Males in one study were more likely to seea pictured woman as attractive if thepicture had a red border.Quail can have a sexual response linked to afake quail more readily and strongly than toa red light.
C iti P
-
8/10/2019 1. Learning1
44/53
In classical conditioning In operant conditioning
Cognitive Processes
When the dog salivates at thebell, it may be due to cognition(learning to predict, evenexpect, the food).Conditioned responses canalter attitudes, even when we
know the change is caused byconditioning.However, knowing that ourreactions are caused byconditioning gives us theoption of mentally breaking the
association, e.g. deciding thatnausea associated with a foodaversion was actually caused byan illness.Higher-order conditioninginvolves some cognition; the
name of a food may triggersalivation.
In fixed-intervalreinforcement, animals domore targetbehaviors/responses aroundthe time that the reward ismore likely, as if expecting thereward.Expectation as a cognitive skillis even more evident in theability of humans to respondto delayed reinforcers such asa paycheck.
Higher-order conditioning canbe enabled with cognition;e.g., seeing something such asmoney as a reward because ofits indirect value.Humans can set behavioral
goals for self and others, andplan their own reinforcers.
-
8/10/2019 1. Learning1
45/53
Learning, Rewards, and Motivation
Intrinsic motivation refers tothe desire to perform abehavior well for its own sake .The reward is internalized as afeeling of satisfaction.Extrinsic motivation refers todoing a behavior to receiverewards from others .Intrinsic motivation cansometimes be reduced byexternal rewards, and can beprevented by usingcontinuous reinforcement.One principle for maintainingbehavior is to use as few
rewards as possible, and fadethe rewards over time.
What might happenif we begin toreward a behaviorsomeone wasalready doing andenjoying?
L i b Ob i
-
8/10/2019 1. Learning1
46/53
Learning by ObservationCan we learn new behaviors and skills without conditioning
and reward?Yes, and one of the ways we do so is by observationallearning: watching what happens when other people do abehavior and learning from their experience .Skills required: mirroring , being able to picture ourselves
doing the same action , and cognition , noticing consequencesand associations .
ModelingThe behavior of others serves as a model, anexample of how to respond to a situation ; we may trythis model regardless of reinforcement.
VicariousConditioning
Vicarious: experienced indirectly, through othersVicarious reinforcement and punishment meansour choices are affected as we see others getconsequences for their behaviors.
Observational Learning Processes
-
8/10/2019 1. Learning1
47/53
Albert Banduras Bobo Doll Experiment (1961) Kids saw adults punching an inflated doll while narrating
their aggressive behaviors such as kick him. These kids were then put in a toy- deprived situationand acted out the same behaviors they had seen.
h
-
8/10/2019 1. Learning1
48/53
Mirroring in the BrainWhen we watch others doing or feeling something,
neurons fire in patterns that would fire if we weredoing the action or having the feeling ourselves .These neurons are referred to as mirror neurons ,and they fire only to reflect the actions or feelings of
others .
-
8/10/2019 1. Learning1
49/53
From Mirroring to ImitationHumans are prone to spontaneous imitation of both
behaviors and emotions (emotional contagion). This includes even overimitating , that is, copying adultbehaviors that have no function and no reward.Children with autism are less likely to cognitively mirror,and less likely to follow someone elses gaze as a
neurotypical toddler (left) is doing below.
Mi i Pl Vi i R i f t
-
8/10/2019 1. Learning1
50/53
Mirroring Plus Vicarious ReinforcementMirroring enables observational learning; we cognitivelypractice a behavior just by watching it.
If you combine this with vicarious reinforcement, we areeven more likely to get imitation.Monkey A saw Monkey B getting a banana after pressingfour symbols. Monkey A then pressed the same four symbols
(even though the symbols were in different locations).
-
8/10/2019 1. Learning1
51/53
Prosocial Effects of Observational Learning
Prosocial behaviorrefers to actionswhich benefit others,contribute value togroups, and followmoral codes andsocial norms.Parents try to teachthis behavior throughlectures , but it maybe taught bestthrough modeling especially if kids cansee the benefits ofthe behavior tooneself or others.
-
8/10/2019 1. Learning1
52/53
Antisocial Effects of Observational Learning
What happens when we learnfrom models who demonstrateantisocial behavior , actions thatare harmful to individuals andsociet y?
Children who witness violence intheir homes, but are not physicallyharmed themselves, may hateviolence but still may becomeviolent more often than theaverage child.Perhaps this is a result of theBobo doll effect? Under stress,we do what has been modeled forus.
-
8/10/2019 1. Learning1
53/53
Media Models of Violence
Do we learnantisocialbehaviorsuch asviolencefrom indirectobservationsof others inthe media?
Research shows that viewing media violence leads toincreased aggression (fights) and reduced prosocial behavior(such as helping an injured person) .This violence-viewing effect might be explained by imitation ,