PSY 3011 - Exam 3
Herrnstein's Results (Natural Concept Learning)
Pigeons were able to classify trees and non-trees in 80 example slides in training, and this learning extended to novel examples, where responding was as rapid as with the training slides.
Video Games and Active Participation (Violence)
Playing violent video games may have a greater effect on aggressive behavior in children than watching violence on TV because video games involve active participation.
Spence Accounted for Both
Proposed that in intradimensional training, an excitatory generalization gradient develops around the S+ and an inhibitory gradient develops around the S-.
Rosch: 3 Important Characteristics of Natural Categories
(1) people tend to agree about which examples are central and which are peripheral. (2) when people are asked to list the members of various categories, they list central instances more frequently. (3) in reaction-time tests, people take longer to decide that peripheral examples are members of the category.
Insomnia -- Reduce Incompatible Behaviors
A bed may be associated with many activities that are incompatible with sleeping including reading, watching television, eating, and thinking about the day's events or one's problems. Can reduce incompatible behaviors by only sleeping in bed, and getting up each time some other activity is occurring.
More Manageable Size (Chunking)
A common idea in human cognitive psychology is that memorizing is easier if a long list of information is divided into portions of more manageable size called chunks.
Compare with Ideal Performer (KP)
A comparison between the individual's performance and those of an ideal performer are a particularly effective form of feedback.
Conditional Discrimination (STM / Working Memory)
A discrimination task in which the subject must choose between one of two stimuli, and the correct response depends on which of the other two stimuli was presented previously.
Elaborate Types of Feedback (KP)
A general conclusion we can draw from this research is that although quantitative KR can be quite helpful in learning a motor skill, more elaborate types of feedback can produce even greater improvements in performance.
Proportion of Human Learning:
A large proportion of human learning occurs, not through classical conditioning or as a result of reinforcement or punishment, but through observation.
Major Purpose of Comparative Cognition
A major purpose of his field is to compare the cognitive processes of different species, including humans.
Radial-Arm Maze
A maze for animals in which eight or more arms radiate from a central starting area, and each of the arms may contain food at the end.
Retrospective Coding
A memory strategy in which subjects choose later responses by remembering what stimuli have occurred previously.
Prospective Coding
A memory strategy in which subjects remember what response needs to be made next, rather than remembering what events have occurred previously.
Slow Positioning Tasks (reinforcement of KR)
A motor learning task in which the subject must move an object toward a target, usually out of sight, with no time limit.
Generalized Motor Program
A motor program that can be adapted to a variety of different situations is called a generalized motor program.
Weber's law
A principle of perception that states that the just noticeable difference (the smallest difference between two stimuli that can be detected) is proportional to the sizes of stimuli).
Peak Procedure ("internal clock")
A procedure for studying animal timing abilities in which the time of its peak response rate shows how accurately the animal can time the intervals.
Directed Forgetting in Animals (Rehearsal)
A procedure for studying memory in which the learner (either human or animal) is taught that on some trials it is important to remember a stimulus and on other trials it is safe to forget the stimulus.
Matching to Sample
A procedure in which reinforcement is delivered if the subject chooses the comparison stimulus that matches a sample stimulus.
Delayed Matching to Sample (DMTS)
A procedure used to measure short-term memory, or working memory. First, a sample stimulus is presented, followed by a delay with no stimuli; then two comparison stimuli are presented, and a choice of the comparison that matched the sample is reinforced.
Nondifferential Training
A simple type of discriminative training in which the positive stimulus is presented on every trial, and there are no trials with negative stimuli.
Narrowing
A technique of stimulus control that involves gradually reducing the range of situations in which an unwanted behavior is allowed to occur.
Behavioral Theory of Timing ("internal clock")
A theory of animal timing proposed by Killeen and Fetterman that states that animals use their own behaviors to measure durations. The theory states that the rate of reinforcement controls the rate of the internal clock, which in turn controls the rate of the animal's behaviors, and the animal uses the behaviors to measure the passage of time.
Intradimensional Training
A type of discrimination training in which responses in the presence of one stimulus are reinforced, but responses in the presence of a different stimulus from the same physical continuum are not reinforced.
Presence-Absence Training
A type of discrimination training in which the presence or absence of a specific stimulus indicates whether responding will be reinforced.
Maintenance Rehearsal
A type of rehearsal that retains information in short-term memory but does not necessarily strengthen information in long-term memory
Associative Rehearsal
A type of rehearsal that strengthens the information in long-term memory.
Simple Form of Social Influence
A very simple form of social influence is social facilitation, in which the behavior of one animal prompts similar behaviors from another animal, but the behavior is one that is already in the repertoire of the imitator.
3 Term Contingency in Imitation
According to Miller and Dollard, imitative learning fits nicely into the Skinnerian three-term contingency of discriminative stimulus, response, and reinforcement. There is no need to claim that observational learning is a separate class of learning that is different from operant conditioning.
PTE "Distracts" (Rehearsal)
Acquisition of a conditioned response proceeds more slowly if some post trial episode (PTE) that "distracts" the animals occurs shortly after each conditioning trial. Surprising PTEs interfere more with learning that expected PTEs.
Fig 12-3 and Adams' Interpretation
Adams theory predicts that during the first stage of motor learning, if participants receive KR only intermittently, the perceptual trace will be strengthened on trials when KR is delivered, but will tend to decay on trials without KR. As shown in Fig 12-3, the participant's accuracy improved immediately after a trial with KR; then it gradually deteriorated during the trial without KR.
Understand Analogy/Relation (animal reasoning)
An analogy is a statement of the form "A is to B is to C is to D". To understand an analogy, one must be able to understand a relation (similarity) between two relations.
Overreliance on Immediate KR (reinforcement of KR)
An overreliance on immediate KR may lead to improved performance while learning a new motor skill, but it seems to hurt performance in a later test when the participant must perform without external feedback.
Object Permanence (animal reasoning)
An understanding that objects continue to exist even when they are not visible. Researchers have studied how and when young children develop object permanence and whether different species of animals can also develop this concept.
Teaching Language to Animals - 5 Conclusions
Animals have demonstrated many of the characteristics of human language→ ••Use of abstract symbols ••Productivity ••Grammar ••Displacement ••Use in communication
Lashley's Arguments (Motor Programs)
Argued that human reaction times are too slow to support the idea that sensory feedback from one response can serve as the stimulus for the next response in a rapid sequence. Second argument was that skilled movements and sequences of movements are still possible for individuals who have lost sensory feedback. Third argument for motor programs concerns the types of errors frequently found in rapid-movement sequences. He noted that many typing mistakes are errors of anticipation or transposition. A fourth type of evidence, not known to Lashley, is that the amount of time needed to begin a sequence of movements depends on the number of separate movements that are part of the sequence.
Distributed vs Massed Practice
As a general rule, laboratory studies have found that performance is better if rest periods are interspersed among fairly brief practice periods, than if practice occurs in one continuous block. In short, distributed practice is better than massed practice.
Social Learning Theory:
Bandera and Walters meant a combination of (1) The traditional principles of classical and operant conditioning, plus (2) the principles of observational learning or imitation.
Learning/Performance Distinction of Bandura's Theory of Imitation
Bandura concluded that reinforcement is not necessary for the learning of new behaviors through observation, but that the expectation of reinforcement is essential for the performance of these new behaviors.
Bandura's Proposal (moral standards and behavior)
Bandura proposed that a child's judgements about what behaviors are good and what behaviors are bad are largely learned by observation.
Equivalence Sets and Language Learning
Belief that ability to learn equivalence sets is similar to the ability to learn language.
Bandura vs Generalized Imitation Theory (best)
Both theories can account for the results, but they do so in slightly different ways. As in other debates between the cognitive and behavioral approaches, the debate over explanations of imitative behavior is partly about terminology and partly about how much we should speculate about processes that we cannot observe directly.
Measure Length Information Stays in Memory (STM / Working Memory)
By using delays of different durations in the DMTS procedure, we can measure how log information about the sample is retained in working memory.
Stimulus Enhancement
Category of imitation in which the behavior of a model directs the attention of the learner to a particular stimulus or place in the environment. As a result
Closed vs Open Loop Movements
Closed-loop movement→ In feedback theory, a movement during which the performer continually receives feedback about whether the movement is proceeding correctly and can adjust his or her behavior in response to this feedback. Open-loop movement→ In feedback theory, a movement that occurs so rapidly that there is no time to make any corrections once the movement begins.
Animal Ability (Time, Numbers, Serial Pattern)
Determining whether animals can also recognize and respond to such abstract dimensions such as time and number is not an easy task, but substantial progress has been made in recent years.
Self-Discipline/Achievement Motivation
Direct reinforcement and observational learning work together in shaping what we might call self-discipline and a high achievement motivation.
Discrimination Reversals
Drawback to using errorless discrimination - After errorless training, the children may have difficulty learning this, in which the roles of S+ and S- are reversed.
Rest Periods
During continuous practice, a type of fatigue builds up that interferes with performance, and this fatigue dissipates during periods, so overall performance is better if these periods are allowed. In learning motor skills there is often an improvement in performance immediately after these periods.
Fig 9-7 (Stimulus Equivalence Training)
Each letter represents a geometrical shape in an experiment on stimulus equivalence training, On each trial, a child is presented with one geometrical shape as a sample, and two choices, one "correct" and one "wrong". In this example, the children are trained with four relations: A→ B, B→ C, D→ E, and E→ F. If they are then tested with new combinations their choices will likely show that they have formed two equivalence sets, as illustrated by the two circles at the bottom.
Transfer from Previous Training - Positive vs Negative
Early theorists believed that it should be possible to observe positive transfer (in which practice in one task aids in the acquisition of a similar task) in some situations, and negative transfer (in which practice in one task interferes with the acquisition of a similar task) in other situations.
Evidence for Vicarious Learning of Phobias
Evidence for the vicarious acquisition of phobias is based on correlational evidence, case studies, and retrospective reports, and this is not the strongest type of evidence.
Fig 9-4
Examples of the types of stimuli and tests used by Lazavera and colleagues to compare the absolute and relational theories of stimulus control. After discrimination training with Circle 1 versus Circle 2, and training with Circle 5 versus Circle 6, pigeons were tested with a new pair of stimuli, Circle 3 versus Circle 4.
Number Discrimination and Weber's Law (Counting)
Experiments have shown that both rats and pigeons can discriminate between two stimuli if their durations differ by roughly 25%. → A principle of perception that states that the just noticeable difference (the smallest difference between two stimuli that can be detected) is proportional to the sizes of stimuli).
General Rule for Movement (Schmidt)
Findings consistent with the prediction of the schema theory that people can learn a general rule for movement and not just individual movement patterns.
2 Key Points of Difference from Generalized Motor Program (Dynamic Pattern Theory)
First, those who favor dynamic pattern theory argue that the evidence for relative time invariance in movement sequences could be the result of the physical properties of the body (such as the length and weight of one's limbs and flexibility of one's joints), not the product of the generalized motor program. Second, dynamic pattern theory emphasizes the important role of feedback from the body and the environment during a movement sequence.
General Shape of Forgetting Curves in Animals
Forgetting is rapid at first, with a substantial loss during the first 24-hours, but subsequent forgetting proceeds at a much slower rate.
Dalal and Meck Experiment (Chunking)
Found evidence for chunking by rats in a 12-arm radial maze. Concluded that by chunking the arms of the maze on the basis of food-type, these rats were able to decrease the burdens on their working memories and thereby perform more accurately.
4 Factors of Bandura's Theory of Imitation
Four factors that determine whether imitative behavior will occur ••Attentional Processes ••Retentional Processes ••Motor Reproductive Processes ••Incentive and Motivational Processes
Combo of Methods in Assertiveness Training
Frequently, assertiveness training consists of a combination of modeling, role playing, and behavioral rehearsal, in which the therapist describes a hypothetical situation, models an appropriate response, asks the client to imitate this response, and evaluates the client's performance.
Hulse and Dorsky Experiment (serial pattern)
Hulse and Dorsky showed that rats were able to learn a steadily increasing or decreasing sequence in fewer trials than a sequence that decreased, increased, and then decreased again-14, 1, 3, 7, 0 pellets. Their explanation is that the learning was slower with this pattern because a more complex rule is needed to describe it.
Animals Choice of Rehearsal
If an animal can choose whether to engage in rehearsal, it should eventually learn to follow the directions and rehearse the sample when it sees the remember cue, but not when it sees the forget cure. This was found in pigeons.
Context-Shift Effect (LTM, Retrieval, and Forgetting)
If you learn some new information in one context (such as a particular room), your recall of the information will be better if you are tested in the same context than in a new context (a different room).
Schmidt's Schema Theory
In an effort to go beyond Adam's theory and deal with these more flexible motor skills, he developed his schema theory of motor-skill learning.
Different Storage Capacities (LTM, Retrieval, and Forgetting)
In contrast to the very limited size of short-term memory, the storage capacity of long-term memory is very large. No one has yet found a way to measure and quantify this capacity for either animals or people.
Guidance Hypothesis (reinforcement of KR)
In motor-skill learning, a theory that if knowledge of results (KR) is given on every trial during learning, the learner becomes overly dependent on this KR and will not perform well in a later test without KR.
Reinforcement and Knowledge of Results (KR)
In motor-skill learning, feedback given to the learner about how close his or her movement came to the goal.
Pacemaker ("internal clock")
In theories of animal timing, a hypothetical internal process that pulses at a steady rate and allows animals to measure durations.
Current Task Performance (STM / Working Memory)
Information in working memory is used to guide whatever tasks the individual is currently performing.
Premack's Conclusion
Language skills are essential for analogical reasoning.
Bandura's Influence (Behavioral Therapy)
Largely because of Bandura's influential work, modeling has become an important tool for behavior therapists.
Transitive Inference in Many Species (animal reasoning)
Learning a rule about the relation between three stimuli, such as the following: "If A<B, and B<C, then A<C. Transitive inference can be found in many species including rats, mice, and pigeons.
The Sophisticated Skill of Learning by Observation:
Learning by observation is the most sophisticated type of learning we have considered in this book.
Long vs Short Term
Long-term memory→ Can retain information for months or years ••Birthday, names of friends, the fact that 4+5=9, the meaning of the word rectangle and more. Short-term memory→ Can only hold information for a matter of seconds ••Phone number you have just looked up for the first time.
Methods Similar to Those Used to Study Timing (counting)
Many of the techniques used to study animal's counting abilities are similar to those used to study timing, and the results are similar as well. Procedures that require animals to count their own responses have shown that they can do so in an approximate way, just as the peak procedure showed that animals can roughly time the absolute duration of stimuli.
Match to Sample VS. DMTS (STM / Working Memory)
Matching to Sample→A procedure in which reinforcement is delivered if the subject chooses the comparison stimulus that matches a sample stimulus. Delayed Matching to Sample→ A procedure used to measure short-term memory, or working memory. First, a sample stimulus is presented, followed by a delay with no stimuli; then two comparison stimuli are presented, and a choice of the comparison that matched the sample is reinforced.
A Country's Children Stories (Achievement Motivation)
McClelland found a significant correlation between the average level of achievement motivation depicted in a country's children's stories and its rate of economic growth during the next 25 years.
Graduated Modeling
Method of progressing from simple to more demanding behaviors is called graduated modeling, and it is a frequent component in many modeling programs.
Special Case of Operant Conditioning (Imitation)
Miller and Dollard claimed that observational learning is not an additional type of learning (besides classical and operant conditioning); rather, it is simply a special case of operant conditioning.
Acquisition of New Behaviors: --Behavioral Skills Training
Modeling is used as part of a larger program that may include verbal instruction, prompting, guided practice, and feedback.
Elimination of Fears and Unwanted Behaviors -- Value of Modeling
Modeling sometimes offers several advantages over systematic desensitization: ••Modeling can be used with very young patients, who may not be able to follow instructions ••Modeling can be a more rapid procedure and require less of a therapist's time, especially when films or videos are used. ••Modeling may produce better generalization to real-world situations.
Essential Ingredient for Learning Motor Skills
Motor skills are an essential ingredient for all types of learned behaviors
General Rule Learned First, Then Exceptions (serial pattern)
Muller and Fountain concluded that in learning these response sequences, animals first learn the general rule (the repeating sequence), and only later do they learn the exception to the rule.
Tongue Protrusion
Newborn infants have a tendency to repeat certain gestures made by adults, especially tongue protrusion.
Direct Practice of Observational Learning
Observational learning can be beneficial in learning motor skills, especially when combined with this type of practice.
Knowledge of Performance (KP)
Often it is possible to give a participant many types of feedback besides information about how close the movement came to some goal. The delivery of such information about the sequence of components of a complex movement is called the knowledge of performance (KP).
Sarah's Initiation of Conversation (animal language)
One disappointing feature of Sarah's performance, however, was that she seldom initiated a conversation. Her use of symbol language was almost exclusively confined to answering questions posed by the experimenters.
Behavioral Contrast
Phenomenon in which responding to one stimulus changes as a result of a change in the reinforcement conditions for another stimulus.
Contextual Interference
Refers to any features of the learning situation ("the context") that make performing the task more difficult (i.e. they interfere with the learner's performance during the acquisition of a new skill).
Studying Animal Working Memory
Research with DMTS and the radial-arm maze has substantially increased our understanding of animal working memory, which turns out to have many of the same properties of human working memory. ••Information in working memory may last only a few seconds, or as long as several hours. ••The amount of information that can be stored is very small so it is susceptible to proactive or retrograde interference.
Parents vs Peers as Models
Researchers have found that drug and alcohol use among adolescents is more highly correlated with the habits of their peers than with the habits of their parents, which suggests that observational learning and social reinforcement play an important role.
2 Types of Interference (STM / Working Memory)
Retroactive Interference→When the presentation of new material interferes with the memory of something that was learned earlier Proactive Interference→ When previously learned material impairs the learning of new material.
Retrospective vs Prospective Coding (STM / Working Memory)
Retrospective Coding→ A memory strategy in which subjects choose later responses by remembering what stimuli have occurred previously. Prospective Coding→ A memory strategy in which subjects remember what response needs to be made next, rather than remembering what events have occurred previously.
Conservation Task and Observational Learning
Rosenthal and Zimmerman showed that a child's mastery of the conservation task can be enhanced by observational learning.
Essential Characteristics of Language (teaching animals)
Terrace concluded that the "language" learned by chimps lacked many of the essential characteristics of human language.
Discrete vs Continuous
Some motor skills are called discrete because they are completed shortly after they have begun. Others are called continuous because they extend for an indefinitely long period of time (balancing).
Music Lyrics and Violence
Some research has found evidence that listening to music with violent lyrics can increase aggressive thoughts and emotions in adolescents. Music with violent and antisocial lyrics can have undesirable effects on the attitudes, emotions, and behaviors of children and adolescents.
Respond as if Stimuli are Interchangeable (Stimulus Equivalence Training)
Stimulus equivalence refers to a situation in which subjects learn to respond to all stimuli in a category as if they are interchangeable, even though they have been taught only a few relations between stimuli, not all possible relations.
Advanced Academic Skills (Stimulus Equivalence Training)
Stimulus equivalence training can be used for more advanced academic skills. In one study, equivalence-based training was given to college students in introductory psychology in an attempt to teach them a difficult topic. On a post-test, students who were given the trained obtained an average score of 92% compared to 57% in a control group that did not receive the training.
Advantage of Schemas
That it allows the individual to respond to new situations with a reasonable chance of success.
Metacognition
The ability to reflect on one's memories and thought processes and make judgements about them.
Motor Components of Observational Learning
The benefits of observing someone else demonstrate a new motor skill may be primarily related to the perceptual components of the task (learning to perceive and attend to the appropriate cues), and direct practice may be necessary to develop the motor components of the task.
Phobias - Animals vs People
The evidence for observational learning may actually be stronger for animals than people.
Concept of Generalization (Imitation)
The principle of reinforcement can account for some instances of novel behavior if we include the concept of generalization.
Compatible with Adams and Schmidt (Response Chain Approach)
The response chain analysis of movement sequences is compatible with theories such as those of Adams and Schmidt, which emphasize the role of feedback in the control of movement.
Attacks Response Chain Approach
The strongest line of attack against the response chain approach to movement sequencing is based on evidence for the existence of motor programs. Those who favor the concept of motor programs suggest that the response chain approach is incorrect because some movement sequences do not depend on continual sensory feedback for their proper execution.
Premack's Principle
The theory that more probable behaviors will act as reinforcers for less probable behaviors, and that less probable behaviors will act as punishers for more probable behaviors.
Human vs Animal Categorization
The three classes of theories developed for human concept learning have also been applied to animal concept learning, and as with human concept learning, there is no agreement about which type of theory is best. Like people, animals differentiate between central and peripheral instances of a category. Like humans, animals do not categorize complex visual stimuli simply on the basis of which individual features are present, but on how the features are arranged as a whole. People and animals can learn to classify stimuli according to a variety of different criteria, depending on what the task demands. Animals can learn concepts that vary in their level of generality. Animals have been questioned in their ability to learn concepts that involve abstract relationships.
Animals Capable of Relational Responding
There is strong support for the idea that animals are indeed capable of this idea...
Television Violence and Aggressive Behavior (Substantial Evidence)
There is substantial evidence that TV viewing can affect the attitudes and behavior of both children and adults.
True Imitation
This occurs when an animal imitates a behavior it has never performed before, and when it is an unusual behavior pattern for that species, which probably would not have been learned if the animal did not observe another animal performing the behavior.
Thorndike's Conclusion (reinforcement of KR)
Thorndike's conclusion was that the Law of Effect is just as important in human motor learning as it was for his animals in the puzzle box. In both cases, reinforcement "stamps in" or strengthens the correct response, so this response is more likely to be repeated in the future.
Adams's Two-Stage Theory - Two Typical Stages
Two stages in the learning of a typical motor skill. The first stage is called the verbal motor stage because in this stage improvement depends on delivery of feedback, usually in a verbal form. The instructor must supply the learner with KR because the learner does not have an accurate perceptual trace and therefore cannot discriminate good trial from a bad one. The second stage is the motor stage. At this point the individual can rely on an internal perceptual trace to judge the accuracy of the movement in the absence of external -- says that in addition to maintaining current performance level in the absence of KR, the learner can actually improve performance by redefining the precision of the motor trace.
Variable vs Specific vs Random vs Blocked Practice (Best Practice)
Variable practice is often better than specific practice, and random practice is often better than blocked practice, but there are exceptions to these rules.
General Ability to Imitate
Very little evidence for a general ability to imitate in young children. A general ability to imitate new behaviors does not appear until later in childhood.
Video Self-Modeling (Types of Behaviors it Works With)
Video self-modeling has been used to teach better social and communication skills to children with autism spectrum disorders to decrease stuttering in adults, to teach simple cooking skills to people with traumatic brain injuries, and to improve the walking and mobility of elderly patients.
Fig 10-4 (Rehearsal)
Wagner, Rudy, and Whitlow varied the time between the trial and the surprising PTE from 3 to 300 seconds for different groups of subjects. Figure 10-4 shows the median percentages of CRs to the new CS over the first 10 conditioning trials. As can be seen, the PTEs had their greatest disruptive effects when they closely followed each conditioning trial, and thereby kept rehearsal to a minimum.
Television Violence and Aggressive Behavior (Longitudinal Results)
Watching violent TV can lead to later aggressiveness, not the reverse. Furthermore, the effects of TV violence are not limited to young children. One longitudinal study found a substantial correlation between the amount of TV exposure at age 22, and assault and fighting at age 30.
Ironic Errors in Movement
Wegner has proposed a theory which states that people have a tendency to make false movements that they are trying hard to avoid, especially if their attention is distracted by some competing task.
Fire When We Act and See Others Act (Mirror Neurons)
What makes mirror neurons unique is that they fire both when an animal makes a certain movement and when the animal observes someone else make that movement.
Proactive Interference
When previously learned material impairs the learning of new material.
Retroactive Interference
When the presentation of new material interferes with the memory of something that was learned earlier
Aggressive Models as Controlling Agents
When they discipline their children, parents are serving as models as well as controlling agents
Pavlov - Automatic By-Product
Why should reinforcement of a behavior in the presence of one stimulus cause this behavior to occur in similar stimuli that have never been used in training? · Pavlov said generalization is this...
Information Processing Approach
With the rise of cognitive psychology in the second half of the twentieth century, many theorists adopted the information-processing approach in analyzing motor skills.
Motor Trace (Adam's)
relates to the workings of the action system of control systems theory. The basic idea is that in addition to learning what it feels like to produce the correct movement, a person must also learn to coordinate their muscles so the movement is indeed produced.
General Rule
rule for learning motor skills is that more difficult practice sessions often produce better long-term performance.
Perceptual Trace or Reference Input (Adam's)
when a person begins to learn a new motor skill, this is weak or nonexistent and is an important part of learning
2 Theories to Explain Simultaneous Discrimination Procedure Results
· Absolute theory of stimulus control: The animal has learned about the two stimuli separately: It has learned that choosing the medium grey card produces food and choosing the dark grey card produces no food. · Relational theory of stimulus control: The animal has learned something about the relationship between the two stimuli: It has learned that the lighter grey is associated with food.
Fig 9-2
· An analysis of peak shift based on Spence's (1937) theory. o (A) Intradimensional training is assumed to produce an excitatory gradient around S+ (550nm) and an inhibitory gradient around S- (550 nm). o (B) The net associative strength of each wavelength equals the difference between its excitatory strength and inhibitory strength. Because of the inhibitory gradient around S-, the peak of this gradient is shifted from S+ in a direction away from S-.
Fig 9-3
· An application of Spence's (1937) theory to the intermediate size problem. o (A) In initial training, an excitatory gradient develops around S+ and inhibitory gradients develop around the two S-'s. o (B) Because of the two symmetrical inhibitory gradients, there is no peak shift in the gradient of net associative strength. There is only a sharpening in the generalization gradient.
Reinforcer Habituation/Satiation Process (Behavioral Contrast)
· Basic idea: The more frequently a reinforcer is presented over a short period of time, the less effective it becomes, because of habituation, satiation, or both.
Rosch: Central vs Peripheral
· Central Instances: In research on concept formation, an example from a natural category that people judge as a "good" or "typical" example. · Peripheral Instances: In research of concept formation, an example from a natural category that people tend to judge as a "bad" or "atypical" example.
Phylogenetic Scale Effects (learning sets)
· Clear tendency for animals higher on the phylogenetic scale to develop stronger learning sets. Noteworthy characteristic of higher species is the ability to acquire more abstract information from a learning situation.
Probe Trials
· Ex: Train a pigeon to peck a yellow key by reinforcing pecks with food on a VI schedule, and the bird will respond at a fairly steady rate. In order to determine how much generalization there is between yellow keys and other color keys, we use these trials. Other colors are briefly presented to measure the pigeon's responding, but no reinforcer is given. o Advantage - Procedure can continue indefinitely without threat of extinction until sufficient data are collected. o Disadvantage - Pigeon may begin to form a discrimination between the yellow key and all other key colors, so that there will be progressively less generalization as training proceeds.
Measuring Generalization with Humans
· Ex: Young children listened to several presentations of a four-second tone. Children then given test trials with tones of different durations. Researchers obtained a fairly symmetrical generalization gradient with the most "yes" responses to the four second tone and fewer "yes" responses to shorter or longer tones.
Exemplar vs Prototype vs Feature Theories (natural categories)
· Exemplar Theories: A theory of concept learning that state's that one's ability to categorize objects depends on one's memory of specific examples. · Prototype Theories: A theory of concept learning that states that one's ability to categorize objects depends on forming a prototype or ideal example, to which new examples are compared. · Feature Theories: A theory of concept learning that states that one judges whether a given instance is a member of a category by checking for specific features.
Behavioral Reallocation Process
· Faster responding in the unchanged component (positive contrast) is possible because of the slower responding that occurs in the component that is changed to extinction. The slower responding in the extinction component will allow the subject to recover from fatigue, so the well-rested animal can respond faster in the unchanged component.
"Errorless" Discrimination Learning -- 2 Differences from Traditional Procedure
· First, rather than waiting for strong, steady responding to the S+, the experimenter introduces the S- early in the training procedure. · Second, a fading procedure is used to make it unlikely that the learner will respond to the S-.
Learning to Learn and Transfer from Problem to Problem
· Improvement in the rate of learning across a series of discrimination problems.
Discrimination Reversal Procedure
· In this procedure, a subject first acquires one discrimination, and then the roles of S+ and S- are periodically switched.
Study Reinforcement Schedules as Isolated Entities?
· It can be dangerous to study reinforcement schedules in this way. An individual's behavior on one reinforcement schedule may be greatly influenced by events occurring before and after the schedule is in effect.
How Kohler's and Hansen's Results Differ
· Kohler: Results support the relational theory. Chicks taught that they receive food in response to medium grey card vs. dark grey card. When presented with a light grey card vs. the reinforced medium grey card, the chicks showed a preference for the light grey card, illustrating that they learned a relationship. · Hansen: Peak shift seems to favor the relational theory. --Peak Shift Supports Absolute or Relational Position? · However, the relational position cannot explain why the responding was much lower with wavelengths between 500 and 520nm, which are the purest greens.
Fig 9-1
· Nondifferential Training: Flat generalization gradient · Presence-Absence Trainingà Peak present in generalization gradient · Intradimensional Trainingà Sharp peak present in generalization gradient
Taking Account of Other Stimuli
· Phenomena such as peak shift and transposition show that it is often impossible to predict how one stimulus will affect an individual's behavior unless we also take into account other stimuli- either those that are currently present or those that the individual has encountered in the past.
Concept Learning / Categorization
· Research that is designed to mimic more closely the types of discrimination an animal must learn in the natural environment. For example, when an animal learns to discriminate between predators and nonpredators or between edible plants and poisonous plants, (1) the stimuli will generally not be simple, idealized forms, (2) there may be countless examples from each category of stimuli, and (3) the distinction between positive and negative instances of a category may not always be easy to make.
Behavioral Contrast -- Fig 9-5
· Results from Gutman's (1977) experiment on behavioral contrasts in rats. When both the light and the noise signaled VI-30sec schedules (Phase 1), response rates were about the same for both stimuli. When the noise signaled a period extinction (Phase 2), response rates declined toward zero when the noise was present but increased substantially above those of Phase 1 when the light was present.
Terrace's Techniques and Education ("errorless" discrimination)
· Terrace first presented the S- within 30 seconds of the pigeon's first peck at the red key. The S- was only presented for 5 seconds at a time which gave the pigeon little time to respond in its presence. In addition, Terrace knew that pigeon's usually do not peck at a dark key, so at first, the S- was not an illuminated green key, but a dark key. Using a fading procedure, Terrace gradually progressed from a dark key to a dimly lit green key, and over trials the intensity of the green key was increased. In summary, in Terrace's procedure, the S- was introduced early in training, it was presented very briefly at first, and it was initially a stimulus that was unlikely to elicit responding. · Because errorless discrimination learning can produce very good stimulus control in a minimum amount of time, variations of Terrace's technique have been used in educational settings.
Lashley and Wade (Generalization Gradients)
· Theorized that some explicit discrimination training along the dimension in question (such as wavelength of light or frequency of tone) is necessary before the typical peaked generalization gradient is obtained.
Gutman's Study
· Used a type of successive discrimination procedure known as multiple schedule... Two or more reinforcement schedules are presented one at a time, in an alternating pattern, and each schedule is associated with a different discriminative stimulus. · Phase 1: Rats exposed to a two-component multiple schedule in which the component schedules were identical: One VI 30-second schedule was signaled by a noise, and a separate VI 30-second schedule was signaled by a light. The light and noise were alternately presented throughout a session. Response rates were about the same in the first condition. · Phase 2: The schedule operating during the noise was switched from VI 30-seconds to extinction. Responding became slower and slower during the noise. Surprisingly, response rates increased dramatically in the presence of light, even though the reinforcement schedule has not changed.
3 Reasons to Use Unusual Tasks (Variety of Motor Skills)
•First, these tasks are selected to be representative of a wide range of everyday movements-the knob-turning task involves a discrete, open-loop movement, and then tracking task involves a continuous closed-loop movement. •Second, these tasks are selected to be as simple as possible so that unnecessary complexities will not make the results difficult to interpret. •Third, since it is unlikely that participants will have encountered these tasks outside the laboratory, the researcher can witness acquisition of a new motor skill.
Washoe's Vocabulary vs Grammar (animal language)
•Washoe was able to learn about 130 signs, an impressive vocabulary. Although she frequently used signs in various combinations, the order in which she used the signs in a sentence was quite inconsistent. In contrast, both children and adults tend to use consistent word orders whether they are using spoken or sign language. •In short, Washoe had a good vocabulary, but poor grammar.
Acquisition of New Behaviors -- Lovaas Method
••Perhaps the best therapeutic example of the training of totally new behaviors through modeling comes from the work of Lovaas and others who have talk children with autism to speak. ••The therapy makes use of a large number of behavioral techniques, such as shaping, prompting, fading, and discrimination training, but the teacher's modeling of speech is indispensable at every stage of therapy.
Television Violence and Aggressive Behavior (Field Experiments)
••Some researchers have conducted field experiments in which the TV viewing and the measurement of aggressive behaviors occur in a more realistic setting. ••In general, the results of field experiments show a modest effect of TV violence on aggressive behavior.
