Learning — 02/23 & 02/25

Learning

Process of acquiring new information or behaviors through experience

First: classical conditioning, aka “Pavlovian conditioning”

• Back to behaviorism…insistence on measuring only observable, quantifiable behavior; dismissal of mental activity as irrelevant and unknowable

  • Only tangible things you can see, not mental activity

Stimulus —> response

Basic Learning Concepts

We learn

• To expect and prepare for significant events (classical conditioning)

• To repeat acts that bring rewards, avoid acts that bring unwanted results (operant conditioning)

We learn new behaviors

• By observing events and people, and through language

• We learn things we have neither experienced nor observed (cognitive learning)

• Behaviorists would not focus on this because they believe you cannot study the mind

Pavlov’s Classic Experiment (Dogs) - Conditioning

1. An unconditioned stimulus (US) produces an unconditioned response (UR)

2. A neutral stimulus (NS) produces no salivation response

3. The US is repeatedly presented just after the US. The US continues to produce a UR.

4. The previously neutral stimulus alone now produces a conditioned response (CR), thereby becoming a conditioned stimulus (CS).

• Food = unconditioned stimulus

• Salivation = unconditioned response

• Tone = neutral stimulus

• Later — Tone = conditioned stimulus

• Later — Salivation = conditioned response

Classical conditioning terms

• Unconditioned stimulus (US)

  • Stimulus that naturally and automatically triggers unconditioned response (UR)

• Unconditioned response (UR)

  • Naturally occurring response to unconditioned stimulus (US)

• Neutral stimulus (NS)

  • Stimulus that elicits no response before conditioning

• Conditioned stimulus (CS)

  • Originally neutral stimulus that, after association with unconditioned stimulus (US), comes to trigger conditioned response (CR)

• Conditioned response (CR)

  • Learned response to previously neutral (but now conditioned) stimulus (CS)

Pavlov’s later work on conditioning

Association

• Initial stage, when association is presented repeatedly so that NS begins triggering CR

  • Neutral Stimulus (NS) needs to be pretty close in time to US

  • It takes many pairings (repetition over time)

Extinction

• Diminishing of conditioned response, occurs when US does not follow CS

  • Reappearance of conditioned response is possible if initial learning has very strong

Spontaneous Recovery

• Reappearance, after a pause, of extinguished CR

  • Response is not as strong but it is still there

Generalization

• Tendency, once response has been conditioned, for stimuli like CS to elicit similar responses

Discrimination

• Learned ability to distinguish between CS and other irrelevant stimuli

• There are some pairings that are easier to make for mammals and some that are harder for them to learn

Preparedness

An animal’s capacity for conditioning is limited by biological constraints — for mammals, taste is a good signal of safety to eat

(on the other hand - birds use sight)

• So, taste/sickness easy to associate

How does Classical conditioning represent learned behavior?

The “learning part” is during the conditioning (middle part)

• The pairing of NS & US

• Actions are linked to rewards/punishments

Operant Conditioning

Example:

A. Behavior: being polite

B. Consequence: Getting a treat

C. Behavior strengthened

• Rewarded behavior tends to recur (if you get rewarded, more likely to do it again)

• Operant conditioning occurs when the subject’s voluntary behavior

• (Not an automatic stimulus —> response)

• Is associated with the occurrence of a significant event

• Rewarding when behavior gets close to what the experimenter wants

Separate components of operant conditioning

• Consequences of the behavior:

  • Reinforcement strengthens the behavior (increases its frequency; makes it more likely to occur)

  • Punishment weakens the behavior (decreases its frequency; makes it less like to occur)

• The stimulus:

  • Positive (present a stimulus, positive = add/apply)

  • Negative (remove a stimulus, negative = remove/subtract)

Operant conditioning matrix

• Increases likelihood of a behavior by presenting a pleasant stimulus after the behavior occurs

Ex. After every hour of studying —> give yourself a little treat —> increase studying behavior!

• Reinforcers vary by animal but can include praise, attention, food, water, money

• Shaping: gradually guiding the behavior to the desired behavior —— changing behavior slowly

• Successive approximations: getting closer to that behavior, step by step —— almost a changed (desired) behavior

• Positive reinforcement is very strong

• Superstitions…

• Positive punishment —— Decreases the probability of a behavior by adding an unpleasant stimulus after the behavior occurs

  • Driving too fast —> get speeding ticket —> less speeding behavior

• Negative Reinforcement —— Increases likelihood of a behavior by removing an unpleasant stimulus after the behavior occurs

  • Car beeping —> put on a seat belt to stop the obnoxious beeping —> leads to more seatbelt use

  • Engage in compulsive behavior in OCD —> relieves the obsession feeling —> leads to more compulsive behavior

• Negative Punishment —— Decreases the probability of a behavior by taking away a pleasant stimulus

  • Kid hits sibling —> take away their iPad —> less hitting