Classical and Operant Conditioning
January 16, 2001
Classical Conditioning
• Reminder of Basic Effect
• What makes for effective conditioning?
• How does classical conditioning work?
Classical Conditioning
• Reflex-basic unit of behavior
• Ring a bell and give a sour ball—soon you will salivate to the sound of the bell = conditioned reflex.
Basic effect
If Unconditioned Stimulus ? Unconditioned Response (meat powder) (salivation)
then pair
Conditioned Stimulus with the Unconditioned Stimulus
(bell) (meat powder)
then eventually
Conditioned Stimulus ? Conditioned Response (bell) (salivation)
Who didn’t know this already? Who didn’t know this?
If Unconditioned Stimulus ? Unconditioned Response (smell of food) (approach)
then pair
Conditioned Stimulus with the Unconditioned Stimulus (sound of can opener) (smell of food)
then eventually
Conditioned Stimulus ? Conditioned Response
Measurement
• You can’t answer questions effectively without an experimental method
• It’s not enough to say “cat comes”
• How many times must it hear the can opener?
• What if sometimes I open a can of soup?
• How quickly will the cat come?
• Can any stimulus be associated with any response?
• WHY does the cat come?
Many questions you could ask
• What makes an effective CS and US?
• How might classical conditioning work?
What makes effective CS & US
• Belongingness
• Taste->vomiting, sight->shock
• Effect observed in humans, too. Fear conditioning to snakes/spiders vs.
flowers/mushrooms (dv = GSR)
What makes effective CS & US
• Novelty
• Bell alone, then bell?food – Bell associated w/ background – Bell associated w/ no food
• Food alone, then bell?food
How does CC work?
• CC can be thought of as adding predictability to the animal’s environment.
• Learning that one stimulus is conditional on the other.
• If one stimulus is not conditional on the other, you won’t get learning.
How does CC work?
• Importance of one stimulus being conditional on another.
– If you present CS and US randomly, you don’t get learning.
– Animals should ignore stimuli that don’t have predictive value.
Predictive value--blocking
Group 1: Tone Shock ToneLight Shock Light Shock Group 2: x Light Shock Light Shock
Training 1 Training 2 Test
Learning = bad
Learning = good
The Point of Blocking
The animal only learns what light means if light carries new predictive information
Rigor allows prediction
Note how different this enterprise is than the casual observation of your cat.
Final phenomenon-- secondary conditioning
What do you think would happen if you taught a dog Light Food,
and then taught it Bell Light,
Answer
The dog would learn it, and would eventually salivate to
bell.
How does secondary
conditioning apply here?
US = food, UR = approach, hovering CS = arm motions, CR = approach, hovering Secondary CS= looking up, CR = approach, hovering
What’s happening, and what should the birds do?
What’s happening: removal of secondary CS What should the birds do?: extinction of CR
Moments later, birds are leaving
Application to humans?
Application to humans?
• Food anticipations--salivation
• Food aversions
• Drug tolerance & addiction
Drug Addiction and Overdose
CC plays a role in deaths caused by drug overdoses
Person who usually takes a drug in a particular setting develops a CR to that place.
Drug ? Big Response (e.g. hypothermia) and body tries to return to homeostasis
Drug Addiction and Overdose
Drug ? Body attempts to counteract (raise body temp.)
US UR
Setting (e.g. bathroom)?Drug
CS US
Bathroom ? Body attempts to counteract drug
CS CR
What happens if the drug is taken in a different room?
Drug Addiction and Overdose
CR does not occur (user’s body does not try to counteract drug) and the user can not tolerate the higher dose.
Drug Addiction
Craving for drug is an attempt to get back to homeostasis:
Craving is caused by Conditioned Stimuli e.g.: handling money
seeing a friend take drug talking about drugs being in specific setting
Operant Conditioning Operant Conditioning
• Conditioned reflexes couldn’t account for all behavior
• Active response ? future change in response depending on consequences.
Operant Conditioning
In classical conditioning, the presence of one stimulus (e.g. meat powder) is conditional on the presence of another stimulus (e.g., a bell)
What else can an animal learn, besides the relationship of two stimuli?
Operant Conditioning
It is also possible for the animal to generate a response and for that response to have consequences:
e.g., act cute, you get pet
What makes OC effective?
• Temporal contingency
• Schedule of reinforcement
• Belongingness
Temporal Contingency
• The delay between the animal’s act that you are reinforcing, and the reinforcer.
– Immediate is more effective than delayed for animals.
– Humans can learn effectively after delayed reinforcement.
Operant Conditioning
Relies on reinforcement:
The process by which consequences lead to an increase in the likelihood that the response will occur again.
Reinforcement
• Positive Reinforcement: desired event is presented after a response.
– example: food when animal presses bar
• Negative Reinforcement: removal of an unpleasant event
– example: removal of shock when animal presses bar.
Schedules of Reinforcement
• Fixed ratio – number
• Variable ratio – number
• Fixed interval – time
• Variable interval – time
Fixed ratio
Reinforcement is given after a fixed ratio of responses.
Time Number of
Responses
Example: factory piecework Steady response Easy to extinguish
Variable ratio
Reinforcement is given after a variable ratio of responses.
Example: slot machine
Rapid response Hard to extinguish
Time Number of
Responses
Fixed interval
Reinforcement is given for a response emitted after a fixed interval of time.
Example: studying for exams
Little response until just before reinforcement:
then rapid response Fairly easy to extinguish Time
Number of Responses
Variable interval
Reinforcement is given for a response after a variable amount of time.
Example: checking mailbox (sort of)
Steady response Hard to extinguish Number of
Responses
Operant conditioning--what makes it effective?
• Temporal contingency
• Schedule of reinforcement
• ** Belongingness
Belongingness
• Thorndike: Cat and puzzle box.
– Pressing lever led to door opening – Not yawning or scratching
• Motivational state can also influence;
a hungry animal does more for food- seeking behaviors. . .
Applications
• Animal training
• Superstition
• Teaching Machines
• Token Economies
Animal Training
• Revolutionized animal training – Shaping
• Importance of temporal contingency
• Exclusive use of positive reinforcement
• Complexity of behaviors when these rules are followed.
Superstition
• Skinner left pigeons alone, reinforced every 15 seconds. Reported that they developed “superstitious” behavior, each bird having a different behavior. Pigeons appeared to believe that they were “making the food appear”
• Temporal contingency--birds were doing something when the food appeared. . .
Superstition
Superstitious behavior: depends on accidental association between action and consequence
Teaching
• Apply operant conditioning principles to learning
– Make sure student doesn’t make mistakes; guide behavior via successive approximations – Review frequently
• Little enthusiasm. Teachers don’t like it and students are bored.
Behavior Modification
• Token economies – Secondary reinforcement – “dehumanizing”?
Operant and Classical
• CC: Neutral stimulus comes to have meaning
• OC: Neutral response comes to have meaning
Are they really different?