Punishment - theories Flashcards
What does Thorndike think about punishment?
Punishment is not effective
Conditions for punishment
Occurs every time Immedietly High intensity every time Not signaled (alternative behavior possible)
which research shows that even a small frequency of punishment oppresses behavior?
Azrin
what are punishment mechanisms?
- Looking at up and down sides of behavior always occurs
- target response that has had highly frequent positive reinforcement = less sensitive to punishment
- Partial posititve reinforcement = punishment is easyer
- With VI or FI schemes > less behavior altogether > respond less with interval schemes
-With FR schemes > post-reinforcement pause longer
But less effect on ratio run
alternative behavior & punishment
Reinforcing a alternative behavior -> leads to better punishment
Especially if:
alternative behavior is incompatible with original response> better punishment
with the same outcome
Who said when a alternative response was presented = decrease in original behavior when paired with punishment
Herman & Azrin
avoidance theory of Dinsmoor
Continues CER > aversive classic conditioning of stimuli that occur before Response
Animal also learns to avoid those stimuli > increase of behavior that is not compatible with punished Response
look up!
What is avoidance experiment of Bechterev?
Pulling pack finger after touching hot plate:
But this is a wrong analysys > you pull your finger back > is a response which you have influence on the outcome!
Is not pavlovian > but operant conditioning!
Wrong analysys
Non signal experiment in avoidance
SIDWELL
Non discrete trials
Similar to skinner box > but negative reinforcement
Procedure: in set intervals shock > just in time jumping to other side of cage. Size of jumping benefit dependent on ‘ timing’ jumping away
Every response makes sense
2 factor theory Mowrer
Pavlovian: fear conditioning
If no R > CS = US
Conditioned fear for CS
OPERANT: Negative reinforcement through avoiding response R
If R > CS = no US
Fear decreases
Avoidance = escaping conditioned fear > not reinforced by nothing - but by reduced fear!
Reduction of fear > avoidance = only possible by learning through classical conditioning!
Bolles: Species specific defense reactions
Natural, unconditioned defense responses
Not much experimental proof > proven that situational influence is less > more probability / closeness of danger or hurt
Predatory imminence criterium
Fanselow
How close danger is > and chance of attack or hurt calculation
Based on that fight, flight or freeze
Cognitive explanation avoidance -
Avoidance behavior only inference:
CS > US
If I don’t want US > need to avoid
Seligman and Johnson
