CH 10: AVERSIVE CONTROL, AVOIDANCE + PUNISHMENT

Avoidance behaviour

• Originated in classical conditioning studies from Vladimir Bechterev

  • Bechterev interested in associative learning in humans 

  • How is lifting the finger in the presence of the CS (and not a US) a departure from typical classical conditioning?

Avoidance vs classical conditioning

• Tone = CS

• Shock = US

• CR = running

  • Avoidance group could avoid the US by running

  • classical conditioning group could not avoid 

Escape

• Terminate aversive by performing response

• Learned first

Avoidance

• Prevent aversive by performing response 

• Learned second 

Apparatus and procedure: looking at avoidance and escape

• Shuttle box with metal bars that deliver small shocks 

• Cue lights (CS)

• One side of the box can be understood as safe (no shock) and not safe (shock)

  • If rats run to safe side during shock → escape 

  • if rats run to safe side when CS comes on → avoidance

Discriminated avoidance procedure

• 2 way shuttle avoidance: animal moves in both directions on successive trials

• 1 way shuttle avoidance: animal placed on some side of apparatus and always moves to the other side at end of trial 

Two-process theory of avoidance

2 mechanisms:

• plains avoidance behaviour based on subject escaping from fear, instead of the avoidance behaviour based on preventing shock 

• Prevention of shock is by-product of the avoidance behaviour 

  • Reduction of fear is reinforcing avoidance behaviour 

2 mechanisms of Two-process theory of avoidance

• Classical conditioning

  • Organism has not learned to make the avoidance response and CS-US pairing come to elicit fear

• Instrumental conditioning

  • Reinforcement of avoidance through fear reduction 

Escape from fear experiments (EFF)

• Separating the contributions of the two processes

1. First train CS-US pairing (tone+shock) with no avoidance as possible (establish fear to CS) 

2. Later exposed to CS but now can turn off the US with instrumental response → no shocks are ever given (see if they can alleviate fear) 

• Second phase lets us see if reduction in fear is enough to drive the response 

Extinction of avoidance behaviour through: response blocking and CS-alone exposure

• When attempting extinction, they disconnected shock generator so no shocks occur regardless of response

• Avoidance response still turned off the CS (warning stimulus)

  • Avoidance response would continue for a long time 

Response blocking

• Preventing a participant from making the avoidance response and exposing them to the CS (extinction of fear) 

• Also called flooding:

flooding

• Extinction + response prevention

  • Preventing avoidance response is key to successful extinction

• The longer the exposure to the CS without US, the better extinction 

• Length of exposure matters more than how many trials of exposure

• Can be used to treat phobias and anxiety 

Nondiscriminated (free-operant) avoidance

• Uninstalled shocks are avoided by making a response within a set period from the previous response/shock

• Each time a participant makes a response, there is a period of safety where no shocks occur

• Repeatedly making the response will keep delaying shocks indefinitely 

  • S-S = shock-shock interval

  • R-S = response-shock interval (safe period) 

Difficulties with two-process theory

• Fear avoidance can be difficult to extinguish 

  • Avoidance happens very quickly, little opportunity to extinguish

• CR may no longer be present (no fear)

Other theoretical accounts of avoidance behaviour

1. Positive reinforcement through conditioned safety signals

2. Reinforcement through reduction of shock frequency

3. Species-specific defense reaction

4. Predatory imminence: defensive and recuperative behavior

5. Expectancy theory 

Positive reinforcement through conditioned safety signals

• Safety signal hypothesis: safety signals that accompany avoidance responses provides positive reinforcement for avoidance behavior 

  • Signals (like skeletal cues, running, etc) can be soothing bc they indicate that you are safe from avoiding stimulus 

  • Explicit feedback stimuli facilitate learning of avoidance response 

Shock-reduction theory (one factor theory) 

• States that classical conditioning component is not necessary

• Says avoidance of shock can serve as reinforcement

  • Theory can apply to the real world → not always referring to shock but just something aversive

• Sidman avoidance task (free-operant avoidance task/nondiscriminated avoidance)

Species-specific defense reactions (SSDR)

• Aversive stimuli and situation elicit strong unconditioned innate responses that helped animals survive and reduce pain and injury (species-specific defense reactions) 

  • Ex. flight, running, freeze, fight, burying, seeking dark 

Predatory imminence

• Different defense responses occur based on level of danger faced by animal 

  • Ex. status of prey

    • Flinch → stare → back up → run

  • Ex. see stranger person

    • Stare → slow walking → turn around →  run

Expectancy theory of avoidance 

• Encounter with aversive stimuli triggers conscious process of threat that generates expectations of future threat levels → human based

  • All based on cues and responses

  • Only responding to cue bc we know what they mean and what we should do about them 

Time out

• Removal of reinforcers or the availability of reinforcers

  • Needs to actually do so

  • Short time out periods usually better

Response cost

• Problem behaviour happening → specified amount of reinforcer is removed 

• Ex. speeding ticket → lose some money (specified reinforcer) 

Overcorrection

• Correct a mistake, and also take it even further

  • Ex. child draw on walls → clean walls they marked and all the other walls 

Point-loss experiment

• Using a points system (like money) 

  • Losing intangible things, as long as they have value, punishment can be felt 

Punisher intensity

• Stronger punishers will produce greater decreases in behaviour

• We need to consider:

  • Ceiling effect (there is a perceived limit to punishment)

  • Importance of appropriate level

    • Not too intense and not too low

  • Punisher should stop behaviour when it occurs 

  • Do not start low and ramp up → causes too much harm 

Contingent aversive stimulation 

• aversive event occurs because of the behaviour

Independent aversive stimulation 

• aversive event occurs regardless of the behaviour 

  • Unrelated aversive events can reduce operant responding 

Delayed punishment

• all delays lead to huge decreases in punishment value

  • Contiguity → delays causes problems and makes it harder to link events 

  • Contingency → harder to link punishment without contingency 

Reinforcement considerations

• Most behaviours occur bc they are reinforced/reinforcing 

  • Generally need to remove reinforcement

  • Avid aversives that are actually reinforcing 

  • Reduce motivation for bad reinforcement 

  • Allow for alternate behaviour that produces reinforcement 

    • Helps still reduce behaviour by allowing them to do other things

Sensitivity to reinforcer: cocaine rats

• Subset of cocaine-addicted rats sought cocaine despite punishment  

• Some things more reinforcing than the punisher 

  • Not true for sucrose/sugar

  • Drugs like cocaine → change the way our brains connect

Effects of a discriminative stimulus for punishment 

• Responding is punished in the presence of a discriminative stimulus

• Not punished when the stimulus is absent 

• Problematic for real world applications → people learn when it is safe to do smth vs when it is not 

Signal for the availability of positive reinforcement

• Why would people seek out punishment?

  • Individuals seek punishment if positive reinforcement is available only when the instrumental response is also punished 

• Punishment becomes S^D for availability of positive reinforcement 

  • Ex. child seeking out punishment for attention  

Theories of punishment

• Avoidance = theoretical

• Punishment = empirical/practical

Conditioned emotional response theory of punishment

• Punishment suppresses behaviour 

• Stimuli/behaviours just before punishment = CSs

• These stimuli elicit freezing

• Explains many punishment effects  

Avoidance theory of punishment

• Punishment as a form of avoidance behaviour

• Stimuli that accompany the punished response become aversive

• Animals avoid these stimuli by doing smth besides the punished behaviour 

Punishment & negative law of effect

• S-R association weakened by having aversive consequence following 

  • Reinforcement strengthens behaviour, punishment weakens behaviour 

• Thorndike (opposite of law of effect) 

Premack principle of punishment

• Different behaviours have different probabilities of occurring

  • Ex. watching tv → high probability; studying → low probability

• Premack principle:

  • H = high probability response 

  • L = low probability response 

    • If L allows H, reinforces L 

      • Studying forces TV, no change in studying 

    • If H forces L, punishes H

      • TV forces studying, decrease in TV

Punishment in real life 

• Ethical issues

• Parenting

• Often used incorrectly

• Later problems

• Used to prevent serious harm 

Noncontingent punishment

• Learned helplessness

  • Perceived inescapable punishment

• Experimental neurosis

  • Unpredictable punishment 

Learned helplessness

• Seligman and Maier (1967) 

  • Dogs exposed to predictable but inescapable shock do not try to escape when later allowed to 

  • Model for depression

• Perceived inescapable punishment

• Alleviating helplessness

  • Force subject to escape 

  • Introduce post-shock signal

Experimental neurosis

• Animals exposed to unpredictable (unsignalled) punishers develop neurotic-like symptoms 

• Unpredictable punishers are stressful (hypervigilance) 

• Lives in constant heightened fear

• Model for PTSD

Benefits of punishment

• Appeasement of behaviors

• Modifies attention

• Distraction from motivator

• Usually very rapid

Problems with punishment

• Punishment does not identify alternative behaviour 

• Discriminative stimuli for punishment become stimuli for avoidance 

• Punishment elicits emotional responses 

• Modeling and imitation 

• Reinforcing value of using punishment 

• Internalizing 

• Psychological harm from noncontingence 

• Habituation

Effective punishment

• Do NOT delay punishment 

• Use consistent contingency and intensity 

• Negative punishment is preferable 

• Intense enough to stop behaviour 

• Explain punishment (if possible) 

• Provide DRO 

• Never punish out of frustration or anger

• Provide safety signals / warnings 

• Use only when necessary