Psy 116 Punishment and Avoidance Learning

Punishment

-Cues are usually classically conditioned-Negative reinforcement usually plays a role--> relief from something bad will increase behavior- increase escape behavior-Expectancy matters--> Rescorla-Wagner predictions?-Frustration or Pain induced aggression-Escape from aversive events is a motivator

Form of learning where a neutral stimulus gains importance by its association with a non-neutral stimulus

A neutral stimulus becomes a CS after being paired with an already established CS--> chaining!

Want the stimulus without the response--> start with weak representation of the stimulusEx. Start with pictures of spiders--> Interact with the pictures--> train representations of this stimuli to not be associated with fear

-Ex.

Watson's Little AlbertConditioned Emotional Response-*CER*- produced by a neutral stimulus that has been paired with an emotion-producing stimulus--> Some stimuli automatically produce fear reactions (Ex. Loud unexpected noises)--> Example CER- Pair tone with shock--> startle response, animal jumps into the air then freezes, increase in HR, BP & breathing, secretion of catecholamines and stress hormones, fear response ~After CS-US pairing, tone elicits the same response as the shock ~No shock, now tone elicits startle response .. etc.

-Wanted to show that every fear was a result of classical conditioning-Our phobias are completely trained.. maybe..

possibly?-If completely learned, can treat them with conditioning, & if you uncondition them---> extinguish it

*Goal is to reduce behavior*Positive punishment is adding something aversive in order to reduce behavior-Ex. squirting your bat with a squirt bottle every time they scratch couchNegative punishment is taking away something good in order to reduce behavior-aka *omission training*-what you lose is called the *response cost*-time out from reinforcement (Ex. taking away your teenager's car when they break curfew)

Although it does reduce behavior, the reduction is *short lived and somewhat context dependent*-i.e.

your punishment for a behavior at school may be different than at home and punishment in one context (at school) does not necessarily transfer to the other context (at home)

*Reinforcement* has a more powerful and longer lasting effects on behavior compared to punishment *Reinforcement generalizes better!PUNISHMENT IS NOT AS GOOD AS REINFORCEMENT

-Severity-Consistency-Immediacy

Usually *more severe* means *more effective* punishment

a MUST for punishment to be effective-Have to have consistency, or shouldn't bother

a MUST for punishment to be effective-the longer the delay, the less likely the punisher will affect behaviorthe longer the delay, the less effective

All have to be changed or punishment will not be as effective-ex. change severity, but not consistency and immediacy--> not going to be as effective-if you do not decrease behavior, you're not doing it right

*Pain-induced aggression*Punishment can lead to aggression-not aggression to avoid punishment, but aggression as a result of anger at being punishedPain induced aggression is affected by previous experiences-if you reinforce non-aggressive responses to punishment, you can reduce the aggression in response to punishment -Punishers usually cause pain--> can lead to aggression-usually out of anger, not necessarily an aggressive act*Modeling*-Ex. Bandura's Bobo dollSocial Leaning Theory says we learn social behavior through imitation and modeling-Evidence--> Children who were punished in an expt later used the same method of punishment when trying to control other kids later on --> Correlation data shows that parents who abused their kids tend to have also been abused themselves (suggests an imitation of parental behavior)-Part of how you learn is by imitating and watching the behaviors that you see-Varied whether the adult was reinforced or punished--> affected behavior (Bandura's Bobo doll experiment)

Escape response is a behavior motivated by an aversive event and reinforced by the ending or removal o the aversive thing-Getting away form an aversive (punishing) stimuli-Negative reinforcement-Very powerful in affecting behavior--> Factors that affect negative reinforcement ~Intensity of the aversive event--> more intense leads to a quicker learning/responding ~Amount of negative reinforcement (i.e. how much does the escape response reduce the aversive event) ~Delay in the reinforcement-trying to escape something bad-response gets you away from something negative--> very fast learningDoes sit make behavior less, or go away?-delay--> how long is it between response and the thing going away?

-Mice have a natural tendency to enter dark environments--> passive avoidance exploits that tendency-not exactly doing a behavior response over and over again, just not going in there-Unidirectional: mouse goes from light to dark chamber-organism learns not to emit a certain response in order to avoid a punishing or aversive stimuli**No CS!***No predictive cue!****no active behavior to get away, just avoid passively

Mouse learns to avoid shock based upon the presentation of a light cue-Unidirectional: mouse is always shocked in the same chamber/location-Learn a predictive cue-Whenever the light comes on, a shock happens-Actively doing a behavior to get away-Light on both sides, so no natural tendency

Mouse learns to avoid shock based upon the presentation of a light cue, which is dependent upon location of the mouse in the apparatus-Bi-directional (hence shuttle): mouse learns to monitor for cues in both chambers that predict shock-Pair light with shock, but light can happen in same location on both sides of the chamber and shock can happen on both sides-have to pay attention to the light, not really a single place associated with the shock

-Active Avoidance-Responding in order to avoid a negative event-Generally produce quick, stable learning-Passive avoidance is just not going into the place/environment that is associated with unpleasant events--> Factors that affect avoidance learning ~Severity of the aversive event ~Delay between the CS and US-Avoidance Learning is really fast!

Uncontrollable bad events--> Perceived lack of control--> Learned helplessnessInescapable shock of avoidance learning:Control group- avoidance learning-shuttle box--> speaker going to make a tone and is going to tell the dog the shock is happening-dog should jump to the other side when the shock goes off--> avoidance learning-a well-trained dog is not going to get shockedShocked group-inescapable shock condition-took away barrier--> only 33% of the dogs learned to jump at the tone-the dogs shocked and couldn't leave--> learned to be helpless--> not going to get any behaviorex. depression--> helplessness--> lack of behavior-not many of the dogs learned avoidance response

*Very hard to extinguish*-Systemic desensitization-Flooding or response prevention is when you force the person to experience the aversive stimulus (the CS) in the absence of the US to eliminate the avoidance response (avoiding the CS)--> Effective (3 sessions and seems long lasting), but people don't like it-Flooding is really effective (ex.

throwing away hoarders stuff)