7a: Associative Learning

ASSOCIATIVE LEARNING

TYPES OF LEARNING

  • Classical Conditioning

    • Definition: Learning that stimulus X predicts outcome Y.

    • involuntary (reflexive)

    • e.g. flinching when you hear thunder

  • Operant Conditioning

    • Definition: Learning that doing X leads to outcome Y.

    • voluntary (intentional)

    • e.g. learning that workiing harder earns you a bonus

  • Erroneous Learning

    • Definition: Illusory correlations—‘learning’ that X and Y co-occur when they are actually random.

CLASSICAL CONDITIONING

  • Definition: Learning association between a stimulus and an outcome, leading to new behavior.

  • Notable Practitioner: Ivan Pavlov and his dogs.

Phases of Classical Conditioning
phases of classical conditioning

  1. Before Conditioning

    • Food: Unconditioned Stimulus (US).

    • Salivation: Unconditioned Response (UR).

  2. During Conditioning

    • Tuning Fork: Neutral Stimulus (NS).

    • Salivation: Response upon presentation of Food (US).

  3. After Conditioning

    • Tuning Fork: Now a Conditioned Stimulus (CS).

    • Salivation: Conditioned Response (CR).

Key Terms in Classical Conditioning

  • Unconditioned Stimulus (US): A stimulus that naturally elicits a response prior to learning.

    • automatically encripted stimulus (the activation button)

  • Unconditioned Response (UR): The response elicited by the US.

    • automatically encripted response

  • Conditioned Stimulus (CS): A stimulus that did not initially elicit a response but does so after learning.

    • manually encripted stimulus (后加的 activation button)

  • Conditioned Response (CR): The response elicited by the CS after conditioning.

    • manually encripted response


Processes in Classical Conditioning

  • Acquisition: The process of learning the association between CS and US, resulting in the CR.

    • acquiring

    • the process

  • Latent Inhibition: Prior presentation of a CS alone can impair further conditioning.

    • a phenomenon in classical conditioning where an organism learns more slowly to associate a stimulus with a consequence after it has been repeatedly exposed to that stimulus on its own.

  • Extinction: The conditioned response is gradually eliminated by presenting the CS alone without the US.

  • Renewal: After extinction in one context, presenting the CS in a new context can reinstate the CR.

Stimulus Control in Classical Conditioning

  • Stimulus Generalisation: The CR is elicited by stimuli resembling the CS.

    • e.g. CS = bell, but also responds to chime and buzz

  • Stimulus Discrimination: The CR is not elicited by stimuli that do not resemble the CS.

    • e.g. CS = bell, and does not respond to chime and buzz

“Little Albert” Experiment (Watson & Rayner, 1920)

  • Overview: Focused on pairing a white rat (CS) with an aversive sound (US).

    • US: Aversive sound (e.g., loud noise).

    • UR: Fear response to the sound.

    • CS: White rat.

    • CR: Fear response to the white rat.

  • Results:

    • Albert learned to associate the white rat with fear, demonstrating a CR where none existed before.

    • Generalization: Albert began to fear other fluffy white objects, such as a Santa Claus mask.

    • Discrimination: He did not exhibit fear towards all objects with hair, only certain ones.

Thinking Exercise in Preparation for Lecture

  • How can phobias arise through classical conditioning?

  • How does modern behavioral therapy utilize extinction?

  • What strategies can effectively “undo” or extinguish Little Albert’s learned fear? What steps could prevent renewal?

OPERANT CONDITIONING

  • Definition: The association between behavior and its consequences, leading to increased or decreased behavior via rewards and punishments.

  • Key Figure: B.F. Skinner.

Key Concepts in Operant Conditioning

  • Reinforcement: Increases the frequency of a behavior.

    • Positive Reinforcement: Addition of a positive stimulus to increase behavior (e.g., praise for good behavior). (+)+(+)=+

    • Negative Reinforcement: Removal of a negative stimulus to increase behavior (e.g., taking a painkiller to alleviate headache). (+)-(-)=+

  • Punishment: Decreases the frequency of a behavior.

    • Positive Punishment: Addition of a negative stimulus to decrease behavior (e.g., parking fine for illegal parking). (-)+(-)=-

    • Negative Punishment: Removal of a positive stimulus to decrease behavior (e.g., removing a toy for bad behavior). (-)-(+)=-

Misunderstanding Reinforcement and Punishment

  • Clarification of negative reinforcement (NOT punishment).

  • Consequences aimed at decreasing behavior are a form of punishment.

  • Consequences aimed at increasing behavior are reinforcement.

Reinforcement Schedules

  • Continuous Reinforcement: Reinforcement occurs every time the behavior is performed—results in quicker learning but less robustness (strong and healthy).

    • behaviour is rewarded every single time it happens

    • learns quickly

    • behaviour is easily lost if reward stops

  • Partial Reinforcement: Reinforcement occurs only occasionally—results in slower learning but more robustness.

    • rewarded occasionally

    • slower learning

    • behaviour more stable and lasts longer

Types of Partial Reinforcement Schedules

  • Fixed-Ratio Schedule: Reinforcement after a fixed number of responses (e.g., rewarding oneself after studying two lectures).

    • reward after 2 lectures watched

  • Fixed-Interval Schedule: Reinforcement after a specific amount of time (e.g., rewarding oneself with a break after one hour).

    • reward after 1hour lecture watched

  • Variable-Ratio Schedule: Reinforcement after an unpredictable number of responses (e.g., gambling).

    • reward after a varied amount of actions

  • Variable-Interval Schedule: Reinforcement after unpredictable time intervals.

    • reward after a varied amount of time

Graphs of Cumulative Responses

  • Illustrate different reinforcement schedules' effects (fixed vs. variable ratios and intervals).

red line=behaviour exhibited
Reinforcers: Strength and Origin

  • Factors that determine effective reinforcers depend on the organism.

  • Premack Principle: Higher probability behaviors can reinforce lower probability behaviors.

    • If behavior A is less likely and behavior B is more likely,  

      then access to B can reinforce A.

    • you can have ice cream after you finish your homework

  • Types of Reinforcers:

    • Primary Reinforcers: Essential for survival (e.g., food, water). the actual goods

    • Secondary Reinforcers: Conditioned to be associated with primary reinforcers (e.g., money).

    • reinforcers that become rewarding through experience by being associated with primary reinforcer

    • e.g. money, praise (linked to social approval or affection), grades (linked to future opportunities)

    • the currency

Thinking Exercise in Preparation for Lecture

  • What schedule of reinforcement would best increase studying behavior and why?

ERRONEOUS LEARNING

  • Definition: The propensity of the human brain to pick up on associations can lead to misconceptions about the relationship between events (illusory correlations).

  • Illusory Correlations: Misinterpretation of random variation as systematic.

Examples of Erroneous Learning

  • Coin flip example:

    • HHHHHH vs. TTTTTT vs. THTHTH: People might incorrectly assume a pattern or cause.

  • Superstitious behaviors and projective tests (e.g., Rorschach blots).

Explanation of Misperception

  • Normal fluctuations in random events can mislead people to conclude unvalidated causal relationships, especially when interpreting personal experiences (e.g., familial structures).

  • Cognitive mechanisms predispose (incline) individuals to see patterns even within randomness, leading to sustained erroneous beliefs.

Risks of Illusory Correlations

  • Cognitive biases and their implications include the perpetuation (the continuation or preservation of a situation/idea) of superstitions and mistaken reasoning.

  • Strategies to mitigate (make something less severe, serious) this include education on randomness and scientific reasoning.

Thinking Exercise in Preparation for Lecture

  • What are the risks of forming illusory correlations?

    • risk because reinforce false associations that distor judgment and fuel stereotypes

    • undermines evidence based reasoning

  • Which strategies can protect against or help overcome illusory correlations?

    • promote statistical literacy

    • encourage critical reflection