Learning
Instrumental Conditioning
& Observational Learning
"He who learns and
runs away will live to learn another day."
E. L. Thorndike
Important Announcement!
- The second thought paper is
due on October 18th.
Brief Review
- Learning: Change in behavior,
thoughts, or feelings of an organism that results from experience.
- Classical conditioning: The principal way
that organisms learn the meaningful relations between events.
Instrumental Conditioning
- Organisms learn how their own
behaviors can produce outcomes; they learn that their own behavior
is instrumental in producing rewards or penalties.
Thorndike's Law of Effect
- If a response in a particular
situation is followed by a satisfying or pleasant outcome, the
response will be strengthened.
- If a response in a particular situation is
followed by an unsatisfying or unpleasant outcome, the response
will be weakened.
Thorndike's Law of Effect
- Key words: "particular
situation," "satisfying or pleasant," "strengthened."
- Learning is situation dependent.
- Strengthened = Response is more likely to
occur in the future in that situation; weakened = response is
less likely to occur in the future in that situation.
- What is "pleasant" or "satisfying?"
Reinforcement
- Reinforcer: A stimulus that
increases the likelihood that the behavior associated with it
will happen again.
Two major types of reinforcement:
- Positive reinforcement: Presentation
of a stimulus after a response increases the likelihood of the
response.
- Negative reinforcement: Removal or cessation
of an unpleasant stimulus after a response increases the likelihood
of the response. NOT the same as punishment.
Examples of Positive
Reinforcement
What Makes a Stimulus
a Reinforcer?
- "Bliss point"
- Response deprivation theory of reinforcement
Schedules of Reinforcement
- Continuous: Behavior is reinforced
every time it is produced.
- Partial: Behavior is reinforced some of the
time, but not always.
- Partial reinforcement produces
greater resistance to extinction.
Types of Partial Reinforcement
- Fixed-ratio, variable-ratio,
fixed-interval, and variable-interval.
- Affects type of responding & resistance
to extinction.
Shaping: Getting the Organism
to Behave in the Desired Manner
- Reinforcement is delivered for
successive approximations to the desired response.
- Let's try it . . .
Problems with Reward
- Extrinsic vs. intrinsic reinforcers
- Extrinsic: Not inherently related
to the activity being reinforced (e.g., praise, money, prizes).
- Intrinsic: Inherently related to the activity
being reinforced (e.g., enjoyment, self-satisfaction).
- Excessive extrinsic reinforcement
can reduce intrinsic motivation.
Lepper, Greene, &
Nisbett
Why?
- Payment = work.
- Extrinsic reinforcers are perceived as "controlling."
- "Bliss point."
- Implications: Grades, allowance.
Punishment
- Stimulus that decreases the
probability of a response, either through the application of
an unpleasant stimulus ("positive") or the removal
of a pleasant one ("negative").
- Usually less effective than reinforcement.
- May lead to unintended consequences:
Unintended Consequences
of Punishment
- Punishment can increase the
likelihood of aggressive behavior.
- Punishment may cause injury.
- May cause fear of the punisher or the context
(what's the mechanism?).
Effective Punishment
- "When you take a knife
away from a child, give him a piece of wood instead."
- Combine with reinforcement of desired behavior.
- Make sure individual understands why the
behavior is being punished.
- Implement punishment immediately.
Effective Punishment
- Administer punishment that is
sufficiently intense to stop the behavior, but no more.
- Make punishment consistent and inescapable.
- Use penalties, i.e., the removal of pleasant
stimuli (negative punishment), rather than physical or emotional
pain (positive punishment).
Observational (Social)
Learning
- We learn by observing the behavior
of others.
Importance of Observational
Learning
- Gender identification and gender-role
development.
- Television violence.
Albert Bandura's Experiments
- Pre-school children watched
a film showing an adult who punched, kicked, and hit with a hammer
a Bobo doll. The film ended with the adult being rewarded, punished,
or neither.
- Children were then allowed to play with the
Bobo doll. Children who saw adult rewarded were more likely to
behave aggressively; children who saw adult punished were less
likely to behave aggressively.
APA Commission Findings
- Higher levels of viewing violence
on TV are correlated with increased acceptance of aggressive
attitudes and increased aggressive behavior.
- Strength of relation is about the same as
between smoking and lung cancer.
Summary
- Instrumental conditioning is
the way that organisms learn how their own behaviors can produce
outcomes.
- Behaviors become more or less likely to occur
depending on their outcomes.
- Reinforcement makes behaviors more likely;
punishment makes them less likely.
- Pattern and persistence of responding
depend in part on the schedule of reinforcement.
- Not all learning requires direct reinforcement;
we learn by observing the behavior of others.