Classical Conditioning: Extinction Procedures

• Extinction occurs when a previously conditioned stimulus (CS) is presented without its associated unconditioned stimulus (US).
• Extinction does not result in 'unlearning'.

Extinction Procedures

• The extinction procedure involves removing the outcome when it has previously occurred, reducing conditioned responding to near-baseline levels.
• Extinction increases behavioral variability.
• Extinction can produce emotional effects like frustration and aggression.

Applied Significance

• Extinction research aims to understand how stimuli can lose their triggering effects in cravings, addictions, phobias, PTSD, and depression, all of which have their own stimulus triggers.
• Drug addicts crave drugs in certain environments.
• Gamblers have urges to continue placing bets.
• PTSD sufferers are anxious in the presence of specific stimuli.
• Phobics experience fear of specific stimuli or situations.
• Eating disorders are accompanied by a loss of control over eating decisions.
• Depressives encounter stimuli that evoke sadness.
• A target of therapy is identifying the environmental events that lead to unwanted behavior and then how to extinguish these behaviors.

Procedure & Process

• Extinction does not result in unlearning.
• One key issue in extinction is determining which psychological and neural processes are responsible for the decreased responding.
• It is key to ascertain whether response loss is transient or permanent.

Signature Phenomena of Extinction

• Spontaneous Recovery
• Renewal
• Reinstatement
• These phenomena highlight that extinction is transient, rather than a complete elimination of prior learning.
• Extinction involves new inhibitory learning.

Spontaneous Recovery

• Rescorla's 2004 study used magazine approach (goal tracking) conditioning with rats to study spontaneous recovery.
• CSs (S1 and S2) are paired with a US during the Acquisition phase.
• During the Extinction phase, both CSs are extinguished.
• S2 is tested immediately after the extinction phase, while S1 is tested after an 8-day delay.
• Responding to S1 spontaneously recovers over this time interval.
• The results imply extinction does not totally eliminate original learning.
• Pavlov observed this phenomenon first and interpreted it to mean that while "excitation" was learned in the acquisition phase, "inhibition" was key after extinction.

The Renewal Effect

• The basic ABA renewal design involves first teaching rats a CS-US association in Context A, then extinguishing said association in a physically distinct Context B.
• The animal would then be tested in either Context A OR Context B.
• Animals typically display more conditioned responding when tested in Context A than in Context B.
• Bouton et al. 2011 ABA instrumental study first trained rats to press a lever for food in Context A before being extinguished in Context B, at which point the rats were tested in both Contexts A and B.
• Rats show "renewed" responding once tested in the conditioning context (A).
• Renewal reflects that extinction memories are difficult to retrieve when testing occurs outside the extinction context.
• Testing should occur in the training context.
• ABC renewal.
• Renewal has been interpreted as a form of occasion setting, the extinction context triggers the extinction memory.
• Spontaneous recovery can be considered a special case of ABC renewal since time acts as a contextual stimulus.

The Reinstatement Effect

• Two groups are given acquisition training followed by extinction.
• Then, one group is exposed to the US by itself.
• The effects of this exposure on responding to the CS are then observed.
• This test will show more CRs in the group given additional USs.
• The CR to the CS has been "reinstated."
• A - context; CS - tone; + shock
• CS/+ - unpaired CS and shock presentations
• P – paired CS and shock
• U – unpaired CS and shock
• R – reinstatement
• N – no reinstatement

Factors That Affect Extinction

• Number and Spacing of extinction trials: More trials and more widely spaced trials lead to greater extinction.
• Repetition of extinction/test cycles: More repetitions lead to greater extinction.
• Conducting extinction in multiple contexts: More contexts lead to more extinction learning.
• Reminder cues that are present during extinction can be effective even when the CS is tested in a different context.
• Rescorla's 2006 study showed a stimulus extinguished with another previously trained stimulus results in deepened extinction.

Paradoxical Reward Effects

• Magnitude of reward effects on extinction: Greater reward magnitudes used in training lead to faster extinction in some learning tasks.
• Partial reinforcement extinction effect (PREE): Partially reinforced stimuli extinguish more slowly than continuously reinforced stimuli.

Theories of the PREE

• Frustration Theory suggests partially reinforced trials induce frustration, which serves as a cue for reinforced responding.
• Sequential Theory (Capaldi) suggests if the previous trial was nonreinforced, its memory serves as a cue for reinforcement on the next trial.