Presentation on theme: "Reinforcement and Extinction To recapitulate from our previous section on learning and classical conditioning: Reinforcement is rewarding or encouraging."— Presentation transcript:
Reinforcement and Extinction To recapitulate from our previous section on learning and classical conditioning: Reinforcement is rewarding or encouraging a specific behavior. Extinction is defined as: a. Presenting the conditioned stimulus alone; b. The decrease in the conditioned response or behavior.
Extinction What happens when reinforcement is stopped? Asked another way, how do you (i) minimize a specific behavior or (ii) get rid of it? Can a behavior be permanently erased? Or is it just suppressed? The following graph depicts the dynamic process of the acquisition, extinction and natural revival (spontaneous recovery) of a behavior.
Explanation of preceding graph Behavior is quickly acquired (acquisition slope) but can just as quickly weaken without reinforcement or reward (extinction slope). However, after a certain rest (extinction) period, the behavior can recover or resurface on its own (spontaneous recovery), and be as potent as before. Hence, reinforcement is not needed all the time to maintain a specific behavior.
Case of Little Albert Baby Albert had no fear of white rats until…an unethical classical conditioning experiment was performed on him in 1920 by John B. Watson: http://www.youtube.com/watch?v=0FKZA Yt77ZM http://www.youtube.com/watch?v=0FKZA Yt77ZM In the experiment, whenever a white rat was put in baby Alberts presence, a loud bang occurred. Initially, baby Albert had no fear of the rat or other animals.
Little Albert However, the sudden loud noise frightened him immensely. Poor baby Albert was startled and cried (sudden loud noises naturally scare any one). Baby Albert subsequently became traumatised whenever the white rat was presented. The experiment was proclaimed a great classical conditioning success and a new milestone in the understanding of how we learn and also how phobias might develop!
Little Alberts problems continue Little Alberts problems did not end there. He soon became fearful of furry animals. He also became fearful of Santas white beard (a mask used in the experiment). This phenomenon of reacting with similar emotion to objects resembling the original feared stimulus is called Stimulus Generalization. Similar objects are now endowed with the same potent properties as the initial CS.
Terminology White rat – NS (neutral stimulus) to CS (conditioned stimulus) Loud noise – UCS (unconditioned stimulus that naturally elicits a fright response ). Startle response – UCR unconditioned response) Fear, crying – conditioned response Stimulus Generalization - Responding emotionally to a stimulus or stimuli similar to the original conditioned stimulus or CS.
Generalization These stimuli are called CS1, CS2, CS3, and CSn, as they become less and less similar to the original stimulus but still can elicit the conditioned response. This association (close or distant) with other stimuli may have good or bad emotional consequences for the person or animal.
Generalization Hence, if you had a bad experience with say, noodles (food poisoning), you could in the future feel aversion to eating not only noodles (CS) but also food resembling noodles like pasta such as spaghetti (CS1), macaroni (CS2), lasagna (CS3). Effectiveness of generalization: there is no need to reinforce every piece of behavior. Therefore, the learning process is sped up.
Generalization – conclusion and application Generalization: From the specific stimulus to the general. From one incident such as an unpleasant experience with someone, generalization can occur to the age group or race of that person. This is the basis of prejudice. In a recent March 2011 survey, 30% of Americans believed that Justin Bieber will end up in Rehab when he is 30. This is because JB is linked by generalization with other Hollywood stars, many of whom end up with addiction issues when they become mature adults.
Operant Conditioning Classical conditioning (CC) plays a huge role in the learning process. But our understanding of the learning process is incomplete without knowing about the contribution of operant conditioning (OC) or active learning. Whereas CC occurs automatically without conscious learning, OC requires us to be active conscientious learners.
Operant Conditioning Operant learning – behaviors are associated with their consequences: 1. Rewarding a behavior will strengthen it. 2. Punishing a behavior will weaken or lessen its occurrence. 3. Pleasure (reward) is more potent for sustaining behavior than pain (punishment) is for eradicating behavior.
Biological influences on learning Animals have a biological blueprint for how certain types of conditioning will or will not occur. This is called Biological Preparedness. Certain Stimulus-Responses associations (i.e., behaviors) are more easily conditioned (acquired) than others. The animal is said to be prepared or contra-prepared to learn.
Developing a new behavior How do you develop or instil a behavior that is currently not in the person or animals repertoire of existing behavior? That is, how do you teach a baby or a puppy or even a grown adult some desired behavior? The book Pygmalion (the movie My Fair Lady) by Bernard Shaw is a good example of the socializing process of an uncultured person.
Shaping A procedure called Shaping is used to generate new behavior. Shaping is also called trial and error learning (Thorndike). The technical term is Approximations of behavior. In shaping a desired behavior, initially many random responses occur with no connection to the target behavior.
Approximations of behavior Behaviors displayed judged to lead to the desired behavior are reinforced. Through a series of chained associations the desired behavior finally emerges. A good word for shaping is manipulation and a better phrase is learning through our mistakes. This is how animal experts train animals. This is also how we learn through our errors.
Instinctive Drift Refers to the deterioration of an operant behavior despite continued training and repeated food reinforcements. Implications and Conclusion: Biological preparedness and instinctive drift do not indicate that there are no laws of learning. The laws of learning can only operate within the constraints imposed by the innate capacities of various animals.
Modeling - Imitation One very easy, no frills way of learning is through imitation. Just watching another persons actions causes us to experience or appreciate the observed persons situation. And observing the behavior being rewarded or punished is sufficient for us to either acquire or reject the behavior in question.
Vicarious Reinforcement This is known as Vicarious Reinforcement. It is also called modeling, observational reinforcement or imitation. Vicarious reinforcement is an important medium for the acquisition of good and bad behavior in childhood.
Vicarious Reinforcement This is why the so-called role models in society are important. There are many observers very willing to imitate their idols behaviors (whether good or bad) Being someones understudy or intern is a good example of learning through vicarious reinforcement. You study the role models behaviors in order to learn more and equip yourself with more and better skills.
Through vicarious reinforcement, virtuous actions can inspire: http://www.vancouversun.com/news/Virtu ous+actions+inspire/4526168/story.html http://www.vancouversun.com/news/Virtu ous+actions+inspire/4526168/story.html Next session: Memory – How and Why We Remember