Ch 8 Learning

Generalization

In classical conditioning, it is the observation that a new stimulus *similar* to the CS can also elicit the the CR. The more similar this new stimulus is to the original CS, the more likely it will elicit a CR. In operant conditioning, a stimulus *similar* to the one that cues a reinforced response will also cause a response. The more similar this new stimulus is to the original cue, the more likely the reinforced response will be exhibited.

Extinction

A procedure in which the reinforcement of a previously reinforced behavior is discontinued. In classical conditioning, the UC is no longer paired with the US. Extinction describes the "process" by which a previously learned response disappears as a result of non-reinforcement.

Learning

A relatively permanent change in an organism's behavior due to experience.

Six Conditions that increase Modeling

1. The model's behavior is observed being reinforced. 2. The model is liked/respected (e.g., Michael Jordan). 3. Observer knows (and likes) the model (e.g., peer influence). 4. Model's behavior is clear and salient for the observer (little leaguer imitates Buster Posey batting stance) 5. Observer is competent to perform the behavior. 6. Model's words match model's deeds

Operant chamber

A chamber also known as a *Skinner Box*, typically containing a bar or key that an animal can manipulate to obtain a food or other reinforcer. The chamber usually has devices to record the animal's rate of bar pressing or key pecking. Used in operant conditioning research.

Classical conditioning

A learning process in which an originally neutral stimulus, by repeated pairing with a stimulus that normally elicits a response, comes to elicit a similar or even identical response. Known as Pavlovian conditioning. Sometimes called Stimulus-Stimulus learning.

Acquisition and Extinction in Schedules

As a general rule, the easier a schedule of reinforcement is to acquire, the easier it is to extinguish. Ratio schedules are more quickly learned (acquisition) and more quickly extinguished. Interval schedules are learned more slowly, but are harder to extinguish. The typical order in ease of acquisition is FR, VR,FI, and VI.

Cognitive Processes vs Behaviorism

Early *behaviorists* such as Skinner felt that a scientific psychology should not resort to unobservable mental concepts. Cognitive psychologists, however, pointed out that many basic learning phenomena such as spontaneous recovery cannot be satisfactorily explained without resort to such concepts as predictability and expectancy.

The Four-Way Matrix of Rewards and Punishments + Extinction

If something is presented (good or bad) it is *positive* in Skinner's approach. If something is withdrawn (good or bad), it is *negative*. So, presenting something good (e.g., a word of praise) is *positive reinforcement*. Presenting something bad (e.g., yelling at someone) is *positive punishment*. If something bad is withdrawn (putting on your seatbelt to stop the warning buzz) it is *negative reinforcement*. If something good is withdrawn (you are "grounded" for the weekend) this is *negative punishment*. If nothing occurs following a behavior, that behavior is extinguished.

Acquisition

In classical conditioning, the initial pairing of the CS and US, when one links an initially neutral stimulus (NS) and an unconditioned stimulus so that the previously neutral stimulus(now called the CS) begins triggering the conditioned response. In operant conditioning, the process of strengthening of a reinforced response. In both cases, learning is said to be *acquired*.

Discrimination

In classical conditioning, the learned ability to distinguish between a conditioned stimulus and a similar stimulus that does not signal an unconditioned stimulus. This can also be seen in operant conditioning, where a cue such as a green light elicits a response that gets rewarded and which can be discriminated from a yellow light that never gets rewarded. One could think of discrimination as the opposite of generalization

Variable Interval Schedule

In operant conditioning, a reinforcement schedule that reinforces a response at unpredictable time intervals that averages a specific time interval.The response must be emitted sometime during each interval.

Fixed Ratio Schedule

In operant conditioning, a reinforcement schedule that reinforces a response only after a specified number of responses. The ratio could be every time a response is emitted (1:1 ratio) or *continuous* reinforcement. It could require, say, five responses before a reinforcement (5:1 ratio). A specific number of responses is required to get a reward (or punishment).

Fixed Interval Schedule

In operant conditioning, a reinforcement schedule that reinforces a response only after a specified time has elapsed. The required response must emitted sometime during that interval.

Variable Ratio Schedule

In operant conditioning, a schedule of reinforcement in which the required number of responses to get a reward (or punishment) *varies* around an *average* value. For example, on average 5 responses may be required to gain a reward, but it may vary from only 2 responses to as many as 8 responses in unpredictable ways.

Classical and Operant Conditioning and Fear Response

John Watson demonstrated that Little Albert could be made to fear a white furry bunny by sounding a startling gong in the presence of a bunny. This fear resisted extinction, because Little Albert moved away from the bunny and became calmer. This part is explained by Operant Conditioning. Little Albert reduced his anxiety by moving away, thereby *negatively reinforcing* the fear/avoidance response.

Associative learning

Learning that certain events occur together. The events may be two stimuli (as in classical conditioning) or a response and its consequences (as in operant conditioning). No mentalistic concept is necessarily implied.

Observational Learning/Modeling

Not all learning appears to involve reinforcement. Animals can learn by observation. Bandura noted that *observers* looked at *models* and imitated their behavior. Observational learning has been observed in humans, primates and other mammals.

Thorndike's Instrumental Learning and Law of Effect

Prior to Skinner's work on operant conditioning, Thorndike noted that rats in a maze will learn the route to a reward. His *Law of Effect* stated that rewarded behavior will be strengthened. Instrumental Learning and Operant Conditioning are the same concept.

Spontaneous Recovery

Recurrence of an extinguished conditioned response, usually following a rest period. The response "recovers" (at least partially) from its extinguished state. This phenomenon is difficult for *strict behaviorists* to explain without resorting to some mentalistic concept.

Schedules of Reinforcement

Schedules that determine when and how often reinforcers will continue. Four types of schedules are identified: Fixed ratio, variable ratio, fixed interval and variable interval. Ratio schedules offer reinforcement based on a discrete number of behaviors required to obtain reinforcement. Interval schedules require that the behavior be emitted during an interval of time, but the number of discrete behaviors is not a factor.

Operant Behavior vs Respondent Behavior

Skinner used the term "respondent behavior" as synonymous with classical conditioning. In operant behavior, the organism "operates" on the environment not merely "responding" to a stimulus.

Biological predispositions

Some biological predispositions to learn such as Garcia's demonstration of one-trial food aversions also present limitations on the behaviorist idea of built-up associations through repeated trials. Different species have different genetically directed learning abilities.

Chemotherapy and learned nausea

Some chemotherapy patients get drugs than have nausea as a side-effect. Patients soon associate the medical facility with nausea and exhibit a conditioned response of nausea to a facility and may skip therapy. Psychologists have tried to *counter-condition* this response by having patients practice a competing relaxation response in the facility.

CR

The Conditioned Response is emitted after the presentation of the conditioned stimulus (CS). The CR is similar or identical to the UCR or UR.

CS

The Conditioned Stimulus is an originally neutral stimulus that comes to elicit a particular response/reflex as a result of pairing with an unconditioned stimulus (UCS). Pavlov sounded a tone that was repeatedly paired with food that caused salivation. When the tone (CS) was presented without food, the dog still salivated.

UCR or UR

The Unconditioned Response is a response that naturally (without learning) results from the presentation of an unconditioned stimulus. Example: An eye blink (UR) results from a puff of air (US) into the eye.

UCS or US

The Unconditioned Stimulus is the term in classical conditioning that naturally (without learning) produces a response called the unconditioned response. Example: Food is an unconditioned stimulus that causes salivation.

Association

The idea of two stimuli connected by association (usually in time). Two things that occur together are associated.

Latent Learning and Cognitive Maps

When not-hungry rats were allowed to explore a maze without reward, the now-hungry rats (under reward conditions) subsequently learned the maze much more quickly than rats in a control condition. They were said to have "hidden" or *latent learning*. Tolman first proposed that animals can have a mental model of the spatial environment called *cognitive map*