- 2 primary characteristics of Skinner’s work
1. Atheoretical
2. Inductive - Built on Thorndike’s work
Expanded Thorndike’s law of effect to an entire system of reinforcement
Thorndike experiment: Hungry cat learned to pull a string in order to leave a box and eat
food from a bowl placed just outside the box
Law of Effect: Behavior is controlled by its consequences - Behavior is emitted from the organism
A consequence occurs
The organism adapts its behavior accordingly
Focus on S-R-C (stimulus-response-consequence)
Not S-R (stimulus-response) - Rewards impact an entire class of behavior
Operant is a class of behavior
Not a single response
Answering the phone - Fictions
People are responsible for their own behavior; people are autonomous
Free will is a superstition - Intend doesn’t counts
Reinforcement & Punishment is not in the intent but in the effect
Approach
- Radical behaviorism
S-R theory can account for all overt behaviors
Took ideas of Watson to logical extreme - Social Darwinism
assumes we are nothing more than a bundle of behaviors shaped by environment - Concentrated on variable and environmental forces, not person
- Sought general principles of behavior
- Relied on animal research (mostly rats and pigeons)
Elegantly simplistic theory
- Functional analysis
- 1 subject at a time (laws of behavior must apply to every subject)
- Internal structures are “fiction”
can’t be directly observed
can’t operationally define
can’t systematically test them
unnecessary to posit internal forces - Personality and personality theories are superfluous
internal states (if they exist) are the by-product of behavior - Operational definitions
Clear definitions not open to interpretation
Didn’t infer internal states (hunger, etc)
# of hrs not eaten
Did not hypothesize drive, insight or any internal process
Skinner’s experimental approach
- Manipulated when a reward was received
- Built a body of knowledge on replication
- Used single subject designs (N=1)
- Rejected statistical analyses
Operant Conditioning
- Also called instrumental conditioning:
Responses operate on the environment and are instrumental in receiving reward - 3 Components
- 1. Antecedent condition
Circumstances that indicate when to respond
The antecedent can be in the form of a discriminative stimulus
– green light = cross.
– red light = don’t cross. - 2. Behaviour
- 3. Consequence
The outcome, result of behavior
Reinforcement = positive outcome
Punishment = negative outcome
- 1. Antecedent condition
- 2 bi-polar dimensions of consequences
Give-take
Posit
Negate - Good-bad (like-dislike)
Reward
Punish - 4 consequence conditions
Positive reinforcement
Positive punishment
Negative reinforcement
Negative punishment
Reinforcement
environmental $ that occurs after response & increases likelihood response will reoccur
increases likelihood of operant reappearing
- 3 types
- Primary reinforcer
satisfies biological need, works naturally, regardless of prior experience - Secondary reinforcer
becomes reinforcing because of association with a primary reinforcer - Generalized conditioned reinforcers
type of secondary
praise and affection
- Primary reinforcer
- 2 ways to apply
give +
take – - Positive Reinforcement
stimulus after response makes response more likely to occur in the future - Negative Reinforcement
response terminates aversive stimulus, strengthens response
also called escape-learning
removing impending doom
avoidance learning: response prevents aversive event from occurring
child cleans his room to avoid parental nagging
5 schedules of reinforcement
- Continuous reinforcement
Shaping
Reinforcer is obtained for every response - Fixed interval (FI) (scalloped)
after the elapse of N minutes - Fixed ratio (FR): every Nth response
- Variable interval (VI) (resistant to extinction)
on average, after N minutes - Variable ratio (VR) (very resistant to extinction)
average is every Nth response
Intermittent schedules: Reinforcer is not obtained for every response
- Rewards should be given deferentially
Parents should reward behaviors they want and ignore (extinguish) behaviors they don’t want.
Behavior can be shaped by rewarding successive approximations
Practice without reinforcement doesn’t improve performance
Punishment
- Punishment (positive and negative) decrease the likelihood an operant reappearing
- 2 ways to apply: give and take
Punishment decreases the likelihood that a response will occur
Examples of punishing situations - Presentation of an aversive stimulus (Positive punishment)
Parent spanks a child for taking candy…
Owner swats a dog who has chewed her slippers… - Removal of a reward (Negative punishment)
Teenager who stays out past curfew is not allowed to drive the family car for 2 weeks…
Husband who forgets anniversary sleeps on couch for a week. - Difficulties in Punishment
Learner may not understand which operant behavior is being punished
Learner fear, rather than learn association between action & punishment (avoids teacher)
Punishment may not undo existing rewards for a behavior
Using punishment when the teacher is angry
Punitive aggression may lead to future aggression
Blocks behavior, not eliminate it
Application
- Teaching pigeons to play table tennis
- Language development
Chomsky - Programmed instruction
Teaching machine (or books with small quizzes)
Small bits of info presented in ordered sequence
Each frame or bit of info must be learned before allowed to proceed to the next section
Assumes proceeding to the next section is thought rewarding
Therapy
3 steps
- identify the behaviors that are maladaptive,
- remove them
- substitute more adaptive and appropriate behaviors
- No need to review the individual’s past or encourage reliving it
not dependent on self-understanding or insight
Operant conditioning chamber
- hated the popular title of “Skinner box”)
“Baby Tender” crib
air conditioned, glass box
used for his own daughter for two and a half years
commercially available, not a popular success
Theoretically successful but practically unaccepted applications - WWII missile guidance system
Pigeons as “navigators”
Army rejected it out of hand. - Token economy (retarded, industrial, prison)
- Social Utopia
Walden II (behaviorally engineered society designed by a benevolent psychologist)
Beyond Freedom and Dignity (most major problems caused by human behavior
Criticisms
- Can’t handle intentionality