Theories of Learning PDF

THEORIES OF LEARNING Dr. Kübra ÇELİK Clark Hull (1884-1952) ◦ He explained motivation and learning by scientific laws ◦ Drive theory ◦ Hull's Reinforcement Theory - In situations of deprivation, we are governed by our primary impulses (hunger, thirst). - biological needs & psychological impul...

THEORIES OF LEARNING Dr. Kübra ÇELİK Clark Hull (1884-1952) ◦ He explained motivation and learning by scientific laws ◦ Drive theory ◦ Hull's Reinforcement Theory - In situations of deprivation, we are governed by our primary impulses (hunger, thirst). - biological needs & psychological impulses - Impulses drive behavior - Homeostasis returns to its original state as impulses decrease. - Reducing impulses allows us to learn about stimulus-response relationality. The intensity of instrumental activity is determined by various factors ◦ Drive(D) ◦ Motivation (K) à Level of reward ◦ Habit power à how much the behavior was rewarded before ◦ Inhibition (I) à unrewarded behavior previously Factors that activate the organism ◦ Impulse (D) à arousal created by deprivation, or a stimulus associated with deprivation ◦ Motivation (K) àarousal elicited by the reinforcer (eg the size of the reinforcer) ◦ Habit power (H)à the strength of the link between the stimulus response ◦ Inhibition (I) à suppression of previous responses that fail to produce rewards Hull divided mice into 2 groups in his experiment: ◦ Withdrawal group (rats fasted for 3 hours and fasted for 22 hours) ◦ Training group (Highly trianed (10 trials) and less trained (5 trials) ◦ Before the experiment, the rats learned to press the lever and take food. ◦ In the experimental phase, when they press the lever, the food does not come. ◦ Result: ◦ Lever pressing behavior is lower in the low impulse state (3 hours fasting) and low habit power (5 training trials). Extrinsic stimulus and stimulant residue ◦ The external stimulus initiates an effect on the sensory nerves. ◦ The effect on the sensory nerves lasts for a few seconds after the environmental stimulus has disappeared. ◦ This effect is called the stimulus trace (Sr) Hull’s Concepts 1- Interaction of Sensory Stimulants ◦ Behavior is the result of the interactions of many stimuli. ◦ Extrinsic stimulus (S) ◦ Stimulant residues (s) ◦ Motor nerves (r) ◦ Elicited behavior (R) 2- Unlearned Behaviors (sUr) ◦ When the organism needs something, it is born with the reaction equipment that can meet this need. ◦ If one response does not reduce the need, another form of response is tried. ◦ If no form of internal response can satisfy the need, the organism has to learn new forms of response. 3- Conditions of learning; Contiguity and reduced impulse ◦ Stimulant causes a response, If the response also meets the biological need, the bond between the stimulus and the response becomes stronger à Impulse reduction ◦ Primary reinforcement à need-fulfilling stimuli ◦ Secondary reinforcement à reduces need stimulants ◦ Similar to law effect of Thorndike ◦ Habit Power (SHR) à As the number of reinforced stimuli and responses increases, the power of the habit also increases. ◦ Without drive there is no response! 4- Stimulus generalization ◦ Similar stimuli produce the same conditioned response. ◦ Previous experiences affect new learning 5- Stimulus-Drive Bond ◦ Biological deprivation in the organism creates the drive (D). ◦ Dryness of the mouth and lips with the urge of thirst; The feeling of hunger is accompanied by the urge to hunger. ◦ Reinforcement is the reduction of these symptoms of the drive. 6- Reaction Potential (sEr) ◦ The probability of a learned response occurring at any given moment is called reaction potential. ◦ For a learned response to occur, the drive must arise and activate the power of habit. ◦ Learning does not occur if S-R match reinforced without impulse. ◦ If the animal opens the door, this is highly linked to hunger 7-Reacting causes fatigue that inhibits the conditioned response ◦Fatigue prevents the organism from reacting à Reactive inhibition = Ir ◦After a period of rest, the organism begins to react again (spontaneous return of the conditioned response) ◦Spaced repetition is better than continuous repetition for performance improvement. Block Operation and Intermittent Operation ◦ A study measuring the ability to follow a moving disc ◦ 3 groups: ◦ 1 Interval run trials ◦ 2 blocks of work ◦ 3 block work rest block work ◦ The interval study group performed better than the other two groups. 8-Conditioned Inhibiton ◦ Fatigue is a negative impulse state, in which case non-reaction has a reinforcing effect. ◦ Learning not to react is called Conditioned Inhibition (sIr). ◦ Reactive inhibition and conditional inhibition prevent the learned response from occurring. ◦ Just like making your cat stay on your shoulder J 9-Oscillation Effect ◦ The level of inhibition of some factors that prevent the emergence of a learned response may vary. ◦ Some barriers a little, some a lot. ◦ This is called the oscillating effect. ◦ The occurrence of the behavior differs from one moment to the next. ◦ If the value of any of the -habit strength, -drive, -stimulus intensity, -or motivation is zero, the instantaneous reaction power is also zero. 10-Reaction Threshold ◦ A learned response can occur when the immediate effective reaction force exceeds a certain reaction threshold. ◦ Conditioning = effective reaction power > reaction threshold ◦ Even if the conditioning has taken place, inhibitory factors (SO) may arise and prevent the conditioned response from occurring. ◦ It can block the conditioned response 11- Latency ◦ It is the time elapsed between the introduction of the stimulus to the organism and the organism showing the learned response. ◦ The shorter the time between the introduction of the stimulus and the onset of the response, the higher the value of the "immediate effective reaction power". 12- Extinction ◦ The greater the number of unreinforced responses before extinction occurs, the greater the value of the "immediate effective reaction power". ◦ The higher the 'instant reaction power', the greater the number of reactions produced in the same situation. 13- First goal response ◦ The conditioned response elicited by stimuli previously encountered in the maze prior to receiving food is called the "partial, first goal response". ◦ Hull’s labyrinth learning explanation is similar to Skinner’s chain reaction explanation. ◦ A mouse is trained to solve a multi-faceted maze and the hungry animal is placed at the starting point, encountering different stimuli in the maze until the animal reaches the point where it is reinforced with food, which is the goal point. ◦ After a few trials, it is observed that the effect of the food on the animal (salivation, chewing, etc.) is also caused by the different stimuli it encounters in the labyrinth before reaching the food.  Summary of Hull’s concepts ◦ 1- Independent variables, which are stimuli that are systematically manipulated by the experimenter: amount of work, number of prior reinforcements, total impulse, amount of reinforcement, stimulus intensity ◦ 2- Intermediate variables. It is the thinking processes within the organism that cannot be observed directly: habit strength, reactive and conditioned inhibiton motivation, effective reaction potential ◦ 3- Dependent variables: effects on the manipulated independent variables. Response intensity, speed, probability, latency, amplitude of behavior, numbers of extinction Discrimination of stimulus ◦ Many learned behaviors arise due to the presence of certain environmental events. ◦ How can we decide to what extent a particular environmental feature or stimulus is responsible for a particular response?Are some stimuli more important than others in controlling a particular response? Identification of the stimulus ◦ To identify the stimulus, the behavior in the presence and absence of the stimulus must be observed. ◦ If there is not any change in the response, this stimulus has no effect on that behavior! Differential response and stimulus discrimination ◦ Reynolds (1966) ◦ Two pigeons are reinforced to peck the rotating response switch in a variable pitch pattern. ◦ Reynold wants to find out which of these controls pecking behavior. ◦ Is it because of the red circle or because of the white triangle? ◦ Results: ◦ First pigeon pecks associated with red circle, other one White triangle Stimulus generalization ◦ If the organism gives the same response to one or more similar stimuli, it generalizes the stimulus. ◦ Greater difference à Decrease in response ◦ For example : ◦ When a child first learns the word "cow", they use the word "cow" not only when they see a cow, but also when they see a bull or horse. Determinants of Stimulus Control ◦ sensory capacity ◦ sensory orientation ◦ Stimulus intensity and salience ◦ motivational factors ◦ Pigeons are particularly sensitive to visual cues when they are hungry and driven to find food. Hull’s theory vs Avoidance ◦ In avoidance à classical and operant conditioning are together. ◦ Shock à reinforcer that increases the (pain) ◦ Food à (hunger) reducing reinforcer ◦ Learning occurs as a result of increased impulse as well as decreased impulse. ◦ When the ringing tone comes before the food, it is hopeful, or when it comes just before the removal of the food, it is disappointing. ◦ Clear responses that lead to success create hope, responses that lead to failure create fear! Learning discrimination ◦ Choosing a stimulus and not choosing the other. ◦ According to Gestalt psychologists ◦ Discrimination is establishing a relational link between two stimuli. Implicit Learning ◦ Spence : Implicit learning is without reinforcement ◦ Mouse are trained to find food and water when they were NOT HUNGRY. ◦ Next experiment, half of them hungry and other half thirsty placed in the maze. ◦ They could find what they need easily. ◦ According to Spence, the more the two events take place together, the stronger the bond between the two. ◦ Even if there is no impulse, the response reinforced many times before can occur easily. à learned response Intrinsic motivation in implicit learning - Spence ◦ S- R connection is determined by intrinsic motivation! ◦ Reinforced response occurs many times, even if there is no motivation. ◦ Even if there is no intrinsic motivation, if the organism is in the state of impulse (hungry, thirsty, etc.), it shows the learned response. ◦ For example, a person who eats lunch at a certain place at 12 every day may eat when he is in the same place at 12, even though he is not hungry. Extinction ◦ Hull : Response produces fatigue (IR) and fatigue inhibits response. ◦ Spence: Withdrawal of reinforcement creates anger in the organism, the behavior caused by anger becomes dominant and causes the learned response to extinct. ◦ WHO IS RIGHT? Visceral conditioning, Biofeedback ◦ Neal Miller (1909-2002) ◦ We can control our heart beat, blood pressure or body temperature ◦ 20 college students are trained to control their blood pressure by showing a pornographic picture. ◦ Studies show that biofeedback is useful for cognitive treatments and hipertension. ◦ With biofeedback trainings patients are able to control muscle tension or body temperature and reduce migraine. Neurofeedback Gestalt Theory of Learning ◦ The whole is different from the sum of the parts. ◦ The music that comes out of the orchestra is different from the sum of the notes played by the various musicians.‘ ◦ Nature does not accept a hole. ◦ The visual system is formed by combining parts. ◦ The learner thinks about all the elements necessary to solve a problem and brings them together first in one direction and then in many ways (cognitively). Wolfgang Köhler ◦ It has created problem situations for monkeys that involve reaching for food that they cannot solve at first. ◦ He used stimuli such as sticks and boxes in his experimental setup https://www.youtube.com/watch?v=lD2_RU1o_Po Insight Learning ◦ Pre-solution period à The period in which the organism develops a set of hypotheses to solve the problem ◦ Insight ◦ 1- Immediate and complete transition from preliminary solution to solution ◦ 2-2 Performance on solution gained by insight is generally error-free ◦ 3 The solution to a problem gained through insight is remembered for a long time. ◦ 4-The principle gained through insight is easily applied to other problems. https://www.youtube.com/watch?v=mDntbGRPeEU ◦ Absolute Theory à Behaviorists' theory. Learning is learning the U T connection. ◦ Relational Theory à Gestalt learning. comparing two stimuli learning. Education according to Gestalt psychology à Creative thinking ◦ Learning without understanding and only with memorizing is limited and easily forgotten. ◦ Learning with Gestalt principles is based on understanding the essence of the problem. ◦ Such learning comes from within the individual and is not imposed by anyone else. ◦ It can be generalized easily and remains in the memory for a long time. Reinforcement according to Gestalt psychology ◦ The solution is as satisfying to the person who solves it as a hamburger does to a hungry person. ◦ Reinforcement à reducing uncertainity ◦ John Holt (1967): ◦ A person who really needs to know something does not need to say it, repeat it, or test it many times. ◦ The new piece of information fits into the void like a missing piece of a puzzle. Once in place, it cannot fall again. ◦ Understand! Don’t memorize to keep it in your memory for a long time! Tolman unites behaviorism and Gestalt ◦ Sign Theory ◦ It is necessary to study holistic behaviors, not small units of behavior. ◦ Behavior is purposeful. ◦ Behavior can be changed and adapted according to environmental conditions in line with the purpose. ◦ Behavior is driven by the product, the goal, that the organism wants to achieve. ◦ Learning is discovering the environment. à the organism tracks the signs and uses them Tolman’s learning theory ◦ Expectations à when you ring a doorbell you expect it to open. ◦ Trial à It can be true or not. It can open or not. ◦ Confirmation of the trial à Not a reinforcement. We still can guess what is going to come up even we don’t have any purpose. ◦ The organism acquires knowledge as a result of paying attention to and exploring environmental events, for which reward is not necessary. ◦ Only the contiguity of experiential events is sufficient. ◦ DO YOU AGREE WITH THIS? Cognitive Map of Tolman ◦ If the organism is given the opportunity to explore its environment, it can acquire many stimulus associations. ◦ And then we can keep that information organized. à this is called ‘cognitive map’ ◦ Just as we use the city map to find a place in the city,Tolman says, higher-level organisms use information that includes the spatial plan of roads and objects to get them to their desired ends. Navigation cells of brain: Place cells and Grid cells ◦ Orange point à place cells in hippocampus ◦ The gray area is where the rat moves freely. ◦ Continuously activated neurons were observed when the mouse was in one area of the room. It was observed that some other nerve cells were also activated in different parts of the room. à map of the room ◦ Blue point à grid cells in entorhinal cortex ◦ When it comes to a certain area in the field, the grid cells in the rat's brain are fired. Thus, it can identify the corners in a space. Tolman’s principle of least effort ◦ After the organism is trained, it tries to reach the goal by using alternative ways, the way that will require the least effort and in the shortest time. ◦ Mental trial and error. Reaction learning vs. Goal learning ◦ Behaviorists à The organism learns to react to certain internal or external stimuli. Reinforcement conditions are not represented. ◦ Tolman In order to know which stimulus-response combinations to follow, the organism learns the goal (goal learning) ◦ The organism acts by knowing the reward or reinforcement conditions and learns the ways to reach it (goal learning) Summary ◦ According to Tolman expectatitons are everything! ◦ If you don’t get what you expect you can’t learn in contrary you become angry. à reward expectations ◦ Extinction occurs when learned expectations are not satisfied. à implicit extinction Social Cognitive Learning Theory Mimicking à Involuntary reaction same up to 3 months Miller and Dollard ◦ Learning by imitation is a form of instrumental conditioning. ◦ Imitated behaviors become stronger when they are reinforced. ◦ The imitated behavior is reinforced intermittently. ◦ Imitation becomes a habit. ◦ Tendency to imitate the behavior of others à generalized imitation ◦ observation à response à reinforcement Bandura ◦ Similar to Tolman’s theory. ◦ Learning occurs even there is no reinforcement ◦ Motivation affects learning. ◦ Learning and performance are different concepts. Bandura’s Bobo Doll Experiment ◦ Group 1: agressive model was reinforced ◦ Group 2: agressive model was punished ◦ Group 3: agressive model – a neutral approach ◦ Group 1 was the most agressive ◦ Although they watched and learned the aggressive behavior, the behavior was not revealed. ◦ https://www.youtube.com/watch?v=dmBqwWlJg8U Social learning theory - reinforcers ◦ Social reinforcers: Approval, smiling, showing interest, praise, and encourage ◦ Self reinforcers: rewarding yourself after a positive behavior à I will eat chocolate after I study one hour. internally speaking positively and feeling positive feelings ◦ Indirect reinforcement: Seeing reinforcement of the behavior increases the frequency of the behavior and shortens the learning time by imitation. ◦ Indirect punishment: Seeing the behavior punished prevents the observer from engaging in similar behavior. ◦ Indirect motivation: If the observed behavior results in a valued product, the observer has a desire to do that behavior. ◦ Indirect emotion: Many emotions are gained through observation. For example, phobias. Models ◦ The more the features of the model are similar to the features of the observer, the more similar the behavior of the observer to the behavior of the model. ◦ Higher status à higher modeling ◦ If the observer fails to imitate the behavior of a successful and effective model, he/she will stop doing the behavior and the effect of the model on the observer will decrease. Impacts of different models on children ◦ Bandura’s experiment: ◦ 3 models:real life people ◦ movie characters ◦ cartoon heroes ◦ Conclusion:Cartoon heroes and movie characters are more effective. Principles of social cognitive theory ◦ 1- Mutual Determination ◦ An individual's future behavior is determined by the environment, behavior and individual characteristics. ◦ There are reinforcements and punishments in the potential environment, and their emergence is determined by our behavior. ◦ 2- Symbolizing Capacity: The external world consists of cognitive representations. ◦ Instruments of thinking and language ◦ Thoughts, which are symbols or cognitive representatives of the past and the future, are the materials that influence or cause subsequent behaviors. ◦ It is the process of creating symbols that enables a person to think of future events cognitively as if they were occurring at the moment. Principles of social cognitive theory ◦ 3- Foresight Capacity: ◦ Human behavior is purposeful and regulated by foresight. ◦ Foresight is the capacity of a person to motivate himself and direct his actions by predicting. ◦ Prior experiences create expectations (predictions) about what will happen as a result of that behavior, before the behavior is performed. ◦ Expectations affect behavior. An individual's capacity to regulate their behavior based on expectations provides a mechanism for predictive behavior. ◦ 4- Indirect Learning Capacity: Indirect learning indicates that people can learn by observing others as well as learning through direct experience. Principles of social cognitive theory ◦ 5-Self-regulation capacity: Accepting that the individual has the potential to shape himself is one of the basic principles of social learning theory. ◦ 6-Self-judgment capacity: The capacity of people to think about themselves, make judgments, and reflect on themselves. Observational Learning Processes 1- Attention Process ◦ Factors affecting the attention process ◦ Adequacy of sense organs ◦ Relevance of the observed activity ◦ Reinforcements received in the past. 2-Retention Process ◦ Information stored imaginatively or verbally needs to be mentally repeated, or applied immediately after being observed, and transformed into behavior. ◦ Most of the cognitive processes that regulate behavior are verbal rather than visual. ◦ Do you remember the vision of the word or the meaning ? Forming behavior process ◦ Transforming learning into performance ◦ Required factors: ◦ physical andnpsychomotor features ◦ self-efficacy capacity ◦ Mental repetitions during the process of creating the behavior ◦ It enables the behavior to be performed more accurately and skillfully. ◦ Motivation process à Motivation is the process of transforming learning into performance. Stages of observational learning What can we learn from models? ◦ psychomotor skills ◦ prohibitions ◦ Values, beliefs ◦ Learning how to use the environment and stuff ◦ Expressing emotions ◦ According to Bandura non-functional learning brings wrong expectations à psychopathology ◦ Bandura etc. (1967) ◦ They showed a child with a fear of dogs a peer who was fearlessly interested in dogs. ◦ The control group is a group of phobic children who have not experienced modelling. ◦ Children who saw their peers interacting with the dog fearlessly gave a significantly more approach to the dog response than the children in the control group. ◦ Direct modeling and symbolic modeling are also effective in reducing fears. Brain activity in social learning ◦ Mirror Neurons ◦ In a 1992 study, it was determined that the nerves of monkeys, which are activated when holding an object, are also activated when observing someone else holding the same object. ◦ The mirror neuron system underlies the neurobiologically rapid repetition of behavior by others. ◦ Yawning? Mirror Neurons ◦ In humans, the mirror neuron system is capable of immediately imitating what others do. ◦ WHY NOT IN MONKEYS? ◦ The monkey understands active intent in purposeful behavior, but cannot encode all the details. ◦ learning by imitationàprefrontal lobe function ◦ These regions occupy a much larger space in the human brain. The involuntary response is the same in all infants up to three months of age. What if mirror neurons don’t work? ◦ Autism ◦ Mirror neurons are active when children with autism do or think about simple movements, ◦ but mirror neurons are not active when children with autism observe the movement of others. D. Hebb – Neurophysiological theory According to Hebb, there are two types of learning: ◦ Childhood learning à where cell clusters and cascades are formed ◦ Adult learning à Insightful and creative learning ◦ Factors that affect childhood learning do not affect adult learning in the same way. Childhood learning forms the framework for subsequent learning. ◦ In childhood àwhile learning a language, cell clusters and successive stages are formed, ◦ In adulthood à existing ones are rearranged and organized. Sequential phase/ cascade ◦ Footstep cell cluster à face cell cluster à cuddle cell cluster ◦ Footsteps à arousal ◦ A temporally related series of cell groups is called a cascade. ◦ When the successive phase is stimulated, the events and objects that are related to each other flow as ideas. Sensory Deprivation and Enriched Environment ◦ The absence of stimuli inhibits the development of the neurophysiological network representing objects and events. ◦ Experiment with deprivation of all sensory stimuli (1957) ◦ According to Hebb, receiving normal stimuli in different environments is a basic need. Lack of these stimulants impairs mental functions and personality. ◦ Isolated subjects stated that they could not think logically, became incapable of solving simple problems, and began to hallucinate. à Sczhizophrenia ◦ Enriched environment; The stimulant-enriched environment improves the organism's ability to learn, especially problem-solving. ◦ Sensory diversity helps the organism to produce more cell clusters and more complex cascades. ◦ After the organism has created these neural circuits, these circuits are used to a great extent in new learning. ◦ Enriched environment, although not at early ages, increases the weight of the cerebral cortex. ◦ According to Hebb, general talent is not determined by heredity, it is a product of experience. ◦ Hebb claims that every environmental object we experience stimulates a complex set of neurons. He calls this complex group of neurons as cell assemblies. ◦ Cell assemblies are neuron units associated with environmental objects and events. Hebbian Learning ◦ If A and B neurons connects and fires each other repeatedly à learning ◦ If A neuron is not enough to fire B neuron, but other neurons can fire B à extinction ◦ "The representation of a sensory experience in the brain, once formed as a pattern by cells that fire together, is fully revealed even when one or more of these cells does not fire."