Psychology: Voluntary Behaviors and Operant Conditioning

Learning Voluntary Behaviors

• Classical conditioning associates neutral stimuli with involuntary behaviors.
• Thorndike (1898) demonstrated the environment can elicit voluntary behaviors.

Operant Conditioning

• Shaping techniques are used to develop new behaviors.
• Primary reinforcers are rewards like food, treats, or access to mates.
• Secondary reinforcers are signals like a click or light, associated with primary reinforcers.
• Skinner observed superstitious behavior in pigeons, where a 15-second interval between trials became associated with a specific behavior.

Operant Conditioning in Motor Impairments

• Learned non-use occurs when a damaged ability, if unused, can cause further injury.
• Constraint-Induced Movement Therapy (CIMT) rehabilitates motor impairments.
• CIMT involves constraining the unaffected limb to force use of the injured limb.
• Shaping is used to improve performance.
• Placebo controls show minimal benefit in CIMT.

Schedules of Reinforcement

• Ratio schedules: Reinforcement based on number of responses.
  • Fixed Ratio (FR-n): Reinforcement after every 'n' responses.
  • Variable Ratio (VR-n): Reinforcement after an average of 'n' responses; produces the highest reinforcement effect, exemplified by gambling.
• Interval schedules: Reinforcement based on time elapsed.
  • Fixed Interval (FI-n): Reinforcement after every 'n' seconds.
  • Variable Interval (VI-n): Reinforcement after an average of 'n' seconds.

Acquisition

• Fixed Ratio (FR): Shows a characteristic "scalloping" pattern; initial slow response that accelerates with reinforcement and then pauses.
• Variable Ratio (VR): Results in a moderate and steady response rate.
• Variable reinforcement: Moderate to high variability as reinforcement is unpredictable.

Extinction

• Reinforcers are withheld to diminish a behavior, eventually eliminating it.
• Extinction is particularly challenging with fixed ratio schedules.

What is Learned?

• Hebb et al. suggest that operant conditioning involves continuous learning about future rewards in a continuous time, as well as the value of those predicted rewards.
• Learning consists of reward predictions and maximizing value selections.

Dopamine Neurons

• Schultz et al. (1993) found dopamine neurons peak when a reward is given.
• Dopamine influences neurons throughout the brain (neuromodulation).
• The dopamine signal indicates reward prediction error.
  • A difference between expected and actual reward triggers a dopamine response.
  • Receiving a predicted reward does not trigger a response.
  • Not receiving a predicted reward leads to a negative dopamine response.

Reinforcement Learning

• Value of a state (Vs) depends on anticipated future rewards (rt).
• Y represents the delay rate.
• It's a dominant approach for modelling human decision-making in diverse situations.
• It's used to train AI for games and real-time action control.

Orbital-Frontal Cortex

• The Orbital-Frontal Cortex (OFC) plays a role in value-based decision making
• The OFC processes decisions for simpler and enhanced value.
• Brain activity in the OFC can be measured to decode monkey decision-making.