Learning and Conditioning

Learning

Learning is the capacity to change behavior based on experience. Learned behavior is a constant change in behavior acquired through experience. Learned behaviors are less rigid and stereotyped than fixed action patterns. Learned behavior is determined by the environment.

Types of Learning

Learning is determined by the environment. There are two main types:

  • By association: Associating responses to stimuli. Examples include Classical Conditioning (Watson, Pavlov) and Operant Conditioning (Skinner).
  • For comments or vicar: Copying or reproducing behavior observed in other individuals. This is the basis of Behaviorism (Watson).

Behaviorism

Behaviorism posits that learning is acquired by interaction with the environment. The only way to learn is through stimulus-response associations. It emphasizes the passive nature of the organism, with activity coming from the environment. Behaviorism holds that you can only study learning through observable experiences and that the laws of learning apply to all species.

Pavlovian Paradigm

Ivan Pavlov discovered that animals can learn to respond to certain stimuli through unconditioned reflexes. He experimented with dogs that were taught to salivate at the sound of a bell. After a while, the sound alone caused salivation, even though it wasn’t food. The dog had learned to salivate to a neutral stimulus, the bell. This form of learning is called classical conditioning.

Classical Conditioning

Definition: Classical conditioning studies the relationship between a stimulus and a response. It involves associating a stimulus that causes a behavior with another stimulus. When the second stimulus is presented, it will elicit the initial behavior or a very similar one.

Key Elements:

  • Stimulus: An event that takes place inside or outside the body and helps to trigger a response.
  • Neutral Stimulus (NS): Does not elicit a specific response regularly.
  • Unconditioned Stimulus (UCS): An event that naturally produces a measurable, unlearned response.
  • Conditioned Stimulus (CS): An initially neutral stimulus that comes to elicit a conditioned response (CR) after being paired with a UCS.
  • Response: An action that takes place at a given time, as a result of stimulation.
  • Unconditioned Response (UCR): A biologically programmed response to a UCS that is produced automatically.
  • Conditioned Response (CR): Arises as a response to a previously neutral stimulus (CS) after it has been paired with a UCS. It is similar to the UCR but is learned.

Higher-order conditioning (SOC) occurs when the CS becomes a UCS, so that when another NS is associated with it, it acquires the power to evoke the appropriate response.

Laws of Classical Conditioning:

  • Generalization: Occurs when an individual identifies different stimuli as the same and produces the same response to them.
  • Discrimination: The subject distinguishes between similar but not identical stimuli and gives different responses.
  • Extinction: If a CS is presented repeatedly without the UCS after conditioning, the CR becomes weaker and weaker until it disappears.
  • Spontaneous Recovery: After a rest period in which the CS does not appear, the presentation of the CS may lead to a new CR.
  • Systematic Desensitization: If a stimulus that causes anxiety is paired with something incompatible with anxiety, the anxiety is reduced until it disappears.

Educational Applications and Implications of Classical Conditioning:

  • Generating patterns of behavior.
  • Enhancing the generalization of learning.
  • The effect of reward or punishment is more successful the more immediate its implementation.
  • The need to use reinforcers in education.
  • Importance in language acquisition.
  • In the formation and change of attitudes.
  • Addressing school pathologies (anxiety, phobias, etc.).

New Perspectives: Biological Predisposition

The conditioning capacity of an organism is determined by biology. The biological disposition of each species makes its members susceptible to learning specific associations. For example, humans are more likely to learn to fear spiders and snakes than flies. This is consistent with Darwin’s principle of natural selection, which favors traits that aid survival.

Other Applications of Classical Conditioning:

  • Conditioning of emotional responses: The case of Little Albert (Watson, 1913).
  • Conditioning of consumer behavior.
  • Substance abuse: Crack, alcohol (Siegel, 1982).
  • Conditioning of interaction with people: Formation of attitudes toward people.
  • Conditioning of spaces: Nausea in waiting rooms.
  • Semantic conditioning: Formation of attitudes toward names of nations.
  • Conditioning by communication: Attitudes towards brands.

Operant Conditioning

Thorndike’s Experiments:

  • Law of Exercise: Repetition (practice) reinforces learning.
  • Law of Effect: Responses that are followed by favorable consequences are more likely to recur.

Definition:

Operant conditioning is a process in which the frequency of a behavior depends on the consequences of that behavior. Conduct that has pleasant consequences is strengthened and tends to repeat itself, while conduct that has unpleasant consequences weakens and tends to disappear. Therefore, the probability of a behavior occurring is a function of its consequences.

B.F. Skinner’s Experiments (1904-1990):

Skinner Box:

  • CS (lever inside the cage)
  • CR (Actuating the lever)
  • UCS (food pellets)
  • UCR (Eating)

Schedules of Reinforcement:

Schedules of reinforcement are guidelines or rules that determine how and when the occurrence of a response will be reinforced.

Types:

  • Continuous Reinforcement: The behavior is reinforced every time it is emitted.
  • Partial Reinforcement: Reinforcement is administered after some, but not all, instances of the target behavior.

Properties of Reinforcers:

  • Positive and Negative Reinforcement:
    • Positive Reinforcement: Increases the likelihood of a response by providing a pleasant stimulus after it occurs.
    • Negative Reinforcement: Increases the likelihood of a response by reducing or eliminating an unpleasant stimulus.
  • Primary and Secondary Reinforcers:
    • Primary Reinforcers: Stimuli that are innately reinforcing (e.g., food, water).
    • Secondary Reinforcers: Stimuli that acquire reinforcing power by being paired with primary reinforcers (e.g., money, grades, praise).
  • Immediate and Delayed Reinforcement:
    • The ability to delay reinforcement is important for complex behaviors.
    • Immediate reinforcements are more effective in shaping new behaviors.

Partial Reinforcement Schedules:

Types:

  • Based on the Interval:
    • Fixed Interval (FI): Reinforcement is delivered after a fixed amount of time has elapsed since the last reinforcement.
    • Variable Interval (VI): Reinforcement is delivered after a variable amount of time has elapsed since the last reinforcement.
  • Based on the Ratio:
    • Fixed Ratio (FR): Reinforcement is delivered after a fixed number of responses have been emitted.
    • Variable Ratio (VR): Reinforcement is delivered after a variable number of responses have been emitted.

Techniques for Behavior Modification:

  • Positive Reinforcement: Reinforcing the desired behavior or increasing the likelihood of it occurring by providing reinforcing stimuli.
    • Techniques:
    • Modeling
    • Token economy
    • Shaping
  • Aversive Control: Removing a pleasant stimulus or presenting an unpleasant stimulus to decrease the likelihood of a behavior.
    • Terms of Use:
    • To provide an alternative
    • To comply
    • To continue
    • To insist

Punishment:

Punishment is a negative event that decreases the likelihood of a response. It has the opposite effect of reinforcement: it tends to suppress a response.

Requirements for Punishment to Be Effective:

  • Immediacy
  • Intensity
  • Consistency

Potential Negative Side Effects:

  • Increased aggression
  • Fear responses
  • Suppression of behavior without teaching desired behavior

Conditions for Learning:

  • Temporal Contiguity: We learn that a behavior causes an event if the event occurs immediately after the behavior.
  • Relevance of the Response: Learning is more effective if the response is relevant to producing the reinforcer.
  • Nature of the Reinforcer: The magnitude and frequency of the reinforcer influence the rate of learning.
  • Response-Reinforcer Contingency: The existence of a contingency between the response and the reinforcer is crucial for learning. It provides an indication that the response is causally responsible for the consequences. Positive or negative contingency gives the subject the ability to control the appearance or elimination of important events. The absence of contingency between the response and aversive events is critical to the phenomenon of learned helplessness, which is most likely explained by the learning of uncontrollability.

Applications of Operant Conditioning:

  • Training/Education:
    • Programmed instruction
    • Modification of classroom behavior
    • Study habits
  • Organizations:
    • Motivation
    • Timing of rewards
  • Home:
    • Consumer behavior
    • Education
    • Self-behavior change (smoking, gambling, alcohol, etc.)

Observational Learning (Bandura)

argues that humans acquire skills and behaviors and instrumental operating mode and that between the observation and imitation involving cognitive factors that help the individual to decide whether mimics the observed or not. Cognitive factors relating specifically to the capacity for reflection and symbolization, as well as the prevention of process-based consequences of comparison, generalization and self-evaluation. The behavior depends on the environment and personal factors (motivation, retention, and motor production). Experiment Bobo or Bobo doll experiment Doll. (1961) STEPS TO LEARNING: Pay attention and perceive the relevant characteristics of the behavior of another individual. Memorize the observed behavior. Repeat the observed action. Be motivated to learn. Observational learning: Applications of Modeling. Models antisocial. Children of violent parents are more likely to be violent. Children of parents who beat their wives are more likely to stick to its future companions. Prosocial models. Effectiveness of the model when it is perceived that consistency between what is said and done. You learn the consequences of the model (reinforcement and punishment)