Operant and Instrumental Conditioning: Principles, Measurement, and Contingencies
Foundations of Instrumental and Operant Conditioning
- Elicited vs. Emitted Behavior:
- Classical conditioning involves elicited responses that are involuntary, automatic, or unconscious.
- Operant and instrumental conditioning involve emitted behaviors that are voluntary and based on choice.
- Historical Terminology:
- Edward Thorne Diet (Edward Thorndike) introduced instrumental conditioning, asserting that behavior is instrumental in producing consequences.
- Thorndike's Law of Effect (1911): Behaviors followed by satisfying consequences are more likely to be repeated, whereas behaviors followed by punishing consequences are less likely to recur.
- B. F. Skinner, writing 25 years later, introduced the term operant conditioning to reflect an organism operating freely on its environment.
- Contingency Frameworks:
- Analyzed as ABC: Antecedent (situation/stimulus), Behavior, Consequence.
- Also represented as SRO: Stimulus, Response, Outcome.
- Organisms learn both the behavior-consequence (B−C) relationship and the stimulus-behavior (A−B) relationship; the antecedent (A) sets the context that determines whether a consequence is reinforcing or punishing.
Measurement Methods: Thorndike vs. Skinner
- Thorndike's Puzzle Box Experiments (Cats):
- Relied on indirect measures of learning over successive trials.
- Latency: The time/slowness taken to pull a lever or escape the box decreases as learning occurs.
- Speed: Running speed along a straight runway toward food increases.
- Error Rate: In a T-maze apparatus, incorrect choices decrease across trials.
- Skinner Box (Pigeons):
- Provides a direct measure of response rate (e.g., counting pecks on illuminated key-pecks).
- Demonstrates preference under varying schedules (e.g., choosing between an immediate 1 pellet of grain via a red light versus a 10-second delay for 5 pellets of grain via a green light).
Behavioral Modification Techniques: Shaping and Chaining
- Shaping:
- Defined as reinforcing successive, closer approximations toward a single target goal behavior.
- Used to establish novel responses that do not occur spontaneously (e.g., teaching a dog to sit by rewarding approach, crouching, and complete sitting sequentially).
- Chaining:
- Used to link a sequence of distinct behaviors into a continuous string.
- Illustrated by training a pet to use a stool to access a couch, or teaching a young child a multi-step routine such as getting dressed.
Reinforcement and Punishment Mechanics
- Core Definitions:
- Reinforcement: Any consequence that strengthens behavior, directly resulting in an increased response rate.
- Punishment: Any consequence that weakens behavior, directly resulting in a decreased response rate.
- Evaluation Criteria:
- A stimulus is defined purely by its functional effect on behavior, not by intent.
Classes of Reinforcers
- Primary Reinforcers (Unconditioned):
- Innately satisfying biological needs without requiring prior learning (e.g., food, water, sleep, sex, and substances that alter brain chemistry directly).
- Secondary Reinforcers (Conditioned):
- Stimuli that acquire reinforcing value through learning and experience (e.g., money, grades, praise, tokens, shopping, winning prizes).
- Behavioral Reinforcers:
- Engagement in one voluntary behavior can serve as the reinforcer for another behavior (e.g., completing homework to earn television time).
Positive vs. Negative Reinforcement and Situational Context
- Positive Reinforcement:
- Strengthening a behavior by presenting (adding) a desirable stimulus following the response.
- Negative Reinforcement:
- Strengthening a behavior by removing, reducing, or avoiding an aversive stimulus following the response (e.g., taking aspirin to relieve a headache, buttoning a coat to escape cold, putting up an umbrella to avoid rain, applying a Band-Aid to stop bleeding, waiving a final exam for a student holding a 95 average).
- Situational Relativism:
- No stimulus is universally reinforcing or punishing across all contexts; the antecedent determines its value (e.g., coconuts on a desert island versus in daily life).
- A single action can yield multiple, simultaneous consequences (e.g., eating pizza can provide positive reinforcement via taste, negative reinforcement via hunger reduction, and punishment via dietary violation or stomach discomfort).
Questions & Discussion
- Connection to Prior Concepts:
- Prompt: Chaining resembles higher-order conditioning and preconditioning by linking multiple elements together.
- Clarification: While sharing associative principles, shaping guides approximations of a single behavior, whereas chaining links discrete, sequential behaviors.
- Persistence of Maladaptive Behaviors and Addiction:
- Prompt: Inquiry regarding why substance addicts persist when punitive consequences appear to outweigh reinforcement.
- Explanation: Immediate, guaranteed reward consistently overpowers delayed, probabilistic punishment. Furthermore, sustained dopamine activity in biological reward pathways actively impairs punishment-based behavioral learning.
- Student-Generated Examples of Negative Reinforcement:
- Correcting posture to eliminate back pain.
- Fidgeting to mitigate anxiety or release excess energy.
- Drinking water to quench thirst.
- Fastening a seatbelt to terminate a warning chime.
- Packing up bags before class ends to avoid being late for the next commitment.