PSY 380 Exam 2 Class notes

Operant Conditioning/Instrumental Learning

Learning the effects and consequences of our own behavior

Experience, Learning, Behavior Change, Utility

• E: repeated exposure to a response, followed by a consequence

• L: learn the nature of a relationship between a response and an outcome (response produces or removes an outcome)

• BC: increase or decrease in behavior depending on consequence

• U: learn how to adapt in variable circumstances, and therefore more likely to survive + reproduce

Discrete learning trials

Each learning experience is a separate, start and end, learning trial

• mazes

Free operant procedures

Animal can do whatever, whenever

• boxes and food pellets

Response —> Consequences (!!!)

• This is our focus, because it outlines essentially the whole theory

• 3 term contingency: setting/response/consequences

  • here, R —> C

Positive contingency

Response makes something more likely to occur —> production

• P(outcome|response) > P(outcome|no response)

Negative contingency

Response makes something less likely to occur —> prevention

• P(outcome|response) < P(outcome|no response)

Zero contingency

Response has no impact on something

• P(outcome|response) = P(outcome|no response)

Extinction

• procedure: stop consequence

• often, initial increase in response and THEN long-term declines

  • vending machine accidents!

  • selective ignoring for poor behaviors in grade school children

Spontaneous recovery

• time — weeks go by and the organism forgets the learning

Disinhibition

• something changes (new babysitter, for example)

Definition method in operational learning

• Contingency type: positive or negative (between response and consequence)

• Behavior change: increase or decrease

Negative reinforcement

• escape: remove something physically there

• avoidance: preventing something from occurring

• behavior increase, negative contingency

Chart of ± contingencies, behavior change

• Positive reinforcement: behavior increase, positive contingency

• Negative reinforcement: behavior increase, negative contingency

• Positive punishment: behavior decrease, positive contingency

• Negative contingency: behavior decrease, negative contingency

Kevin chart: contingency type and consequence type

• Positive reinforcement: positive contingency, appetitive consequence

• Negative punishment: negative contingency, appetitive consequence

• superstition/learned laziness: no contingency, appetitive behavior

• positive punishment: positive contingency, aversive consqeunce

• negative reinfrocement: negative contingency, aversive consequence

• learned helplessness: zero contingency, aversive consequence

Superstition in the lab

B.F. Skinner gave pigeons food for one hour at 60S intervals

• most pigeons were completing random behavior that they seemed to believe ensured the consequence

  • we have a BIAS to perceive control where there is none.

  • smoke detector principle

Learned laziness

Impacts future learnign! Hard to adjust to difficult situations bc it slows down future learning

Learned helplessness

Demonstrates the importance of control

• Martin Seligman: rats, in identical experimental chambers

• Grp 1: shock —> shuttle —> shcok gone

  • future experiments: normal learning

• Grp 2: no control over the shocks, even though they were yoked to grp 1, which had control

  • cognitive impact: decreased future learning

  • affective impact: depression

  • immune system suppression

• THEN, inject them with cancer cells that confer a 50% mortality rate

  • in rats who experienced learned helplessness at a young age, more than 50% die

Continuous Reinforcement Schedule

Every response gets reinforced

• FR1 = continuous

Fixed Ratio Schedules

Constant number of responses required to be reinforced (FRn)

Variable Ratio Schedules

Varying number of response required to meet an aver # of responses

• element of surprise! which leads to dopamine —> pleasure —> addiction

Fixed Interval Schedules

Interval has to elapse before next reward is available

Variable Intervable

Uncertain amount of time before reward is available

Schedule Impacts

In general, ratio schedule produces higher amounts of responding than interval schedules because you get rewarded more.

• fixed schedules usually show post-reinforcement pauses because you KNOW you have to wait

• variable schedules produce more consistent responding because the next one might always be THE one!

Continuous record

• pattern of response over time

• straight line = steady response

• FR: stair steps

• FI: scallops

• VR: steep, straight line

• VI: flatter, straight line

Factors Influencing Operant Conditioning

1. magnitude of reward/consequence

2. Evolved predispositions

3. miracle learning

4. contingency

5. contiguity