ch9.pdf
Transcript
PSY381-Assoc Prof Hande Kaynak 10/10/2023 Escape, Avoidance and Punishment Chapter 9 1 Escape vs. Avoidance 2 Escape behavior performance of the behavior terminates the aversive stimulus, Avoidance behaviorperformance of the behavior prevents the aversive stimulus from occurring. e.g. We escap...
PSY381-Assoc Prof Hande Kaynak 10/10/2023 Escape, Avoidance and Punishment Chapter 9 1 Escape vs. Avoidance 2 Escape behavior performance of the behavior terminates the aversive stimulus, Avoidance behaviorperformance of the behavior prevents the aversive stimulus from occurring. e.g. We escape from the rain when we run indoors after it has started; we avoid the rain when we head indoors before it has started. One first learns to escape from an aversive stimulus and then to avoid it. A shuttle avoidance procedure Escape vs. Avoidance-Negative punishment 3 4 Latency for jumping behavior Safe Shock Time A rat is placed in a chamber divided by a low barrier. A stimulus such as a light, is presented for 10 seconds, followed by a mild electric shock. The rat can escape from the shock by climbing over the barrier to the other side of the compartment. Crossing the barrier is then negatively reinforced by the removal of shock. 1 PSY381-Assoc Prof Hande Kaynak 10/10/2023 Escape vs. Avoidance Escape vs. Avoidance 5 6 Shock is preceded by the presentation of a light. It means that light is a warning which signals that a shock is about to occur. As the rat learns to associate the light with the shock, it will begin crossing the barrier whenever the light is presented and before the shock begins. Light Shock Escape vs. Avoidance Escape vs. Avoidance 7 8 In similar fashion, we might first learn to escape from an upsetting conversation with a racist acquaintance by inventing an excuse for leaving. After a few experiences, however, we might begin actively avoiding that individual before any encounter. By doing so, we avoid having to endure any exposure to that person’s racist views. http://www.youtube.com/watch?v=fwzEbzgPszk Behavior that terminates an aversive stimulus is called _______behavior, whereas behavior that prevents an aversive stimulus from occurring is called _____ behavior. Typically, one first learns to __________ from an aversive stimulus, and then to _______________ it. Julio initially takes vitamin C whenever he has a cold, in the hope that it will shorten the duration of his symptoms. Feeling that this is effective, he begins taking it daily in the hope that it will keep him from contracting a cold. Julio initially took the vitamin C to (avoid/escape) _________ the symptoms of a cold; he later took it to ___________ the symptoms of a cold. In the shuttle avoidance procedure described previously, the rat first learns to ____________ from the shock. The rat later learns to ________ the shock, with the ____________ acting as the SD for the behavior. 2 PSY381-Assoc Prof Hande Kaynak 10/10/2023 Two-Process Theory of Avoidance Two-Process Theory of Avoidance 9 10 Escape behavior is relatively easy to understand. The rat Two-process theory of avoidance (also known as the two- moves from a clearly aversive situation to a nonaversive situation. But the motivation underlying avoidance behavior is less apparent. When climbing over a barrier to avoid shock, the rat seems to be moving from one nonaversive situation (no shock) to another nonaversive situation (no shock). So, no aversive stimulus! where is the reinforcer here? Is it possible to occur conditioning without reinforcer? factor theory of avoidance) proposed by Mowrer (1947, 1960) tried to explain this. 2 processes are involved in learning an avoidance response. The first process is classical conditioning of a fear response to a CS. From a behaviorist perspective, it is impossible to change a behavior without a consequence. Two-Process Theory of Avoidance Two-Process Theory of Avoidance 11 12 Once this conditioned fear has been established, it then forms the basis of an operant conditioning procedure. If the CS generates a conditioned fear response, then moving away from the CS should result in a reduction of fear. This reduction of fear should in turn serve as a negative reinforcer for the response. Two distinct processes: (1) classical conditioning, in which a fear response comes to be elicited by a CS, and then (2) operant conditioning, in which moving away from the CS is negatively reinforced by a reduction in fear. So, avoidance response is not reinforced by ‘nothing’! It is reinforced by ‘reduction of fear’! This explanation is indeed very similar to what Hull says (drive reduction). According to Mowrer, I go out of my way to avoid bees because behaving this way has been (positively/negatively) _____ (reinforced/punished) ______ by a _____ in fear. 3 PSY381-Assoc Prof Hande Kaynak 10/10/2023 Two-Process Theory of Avoidance Two-Process Theory of Avoidance 13 Some problems about this theory: Avoidance responses are often extremely persistent. One can make more than 600 avoidance responses in a shuttle box. Yet, avoidance was no longer necessary because shock was not presented. If the animal repeatedly encounters the CS in the absence of the US, then fear of the CS should eventually extinguish—meaning that the animal should eventually stop jumping over the barrier. But no! it did not stop jumping. 14 A possible answer comes from ‘anxiety conservation hypothesis’. Avoidance responses usually occur so quickly that there is insufficient exposure to the CS for the conditioned fear to fully extinguish—that is, a good deal of the conditioned fear is conserved because exposures to the CS are too brief for extinction to take place. 2nd problem about the theory: after repeated avoidance trials, animals appeared to show no evidence of fear but continued to make the avoidance response anyway. If the animals were no longer afraid of the CS, how could avoidance of the CS have been negatively reinforced by a reduction in fear? Two-Process Theory of Avoidance Two-Process Theory of Avoidance 15 16 Levis (1989) argued that although animals in avoidance experiments may become significantly less fearful with experience, there is no evidence that they become completely nonfearful. Since it is too persistent, how can we extinguish it? Presenting the shock in the ‘safe’ part. Blocking the avoidance response. (a long barrier works.) In fact, evidence suggests that if an animal completely loses its fear of the aversive CS, then, just as two-process theory predicts, the avoidance response stops. But as long as some fear remains, the avoidance response continues, suggesting that fear reduction is still functioning as a negative reinforcer for the behavior. 4 PSY381-Assoc Prof Hande Kaynak Avoidance Learning 17 Avoidance learning appears to be a fundamental process in 10/10/2023 Escape, Avoidance and Punishment the development and maintenance of phobic behavior. In human phobias, people avoid the aversive CS. Chapter 9 (continued) 18 E.g. A person who has a fear of elevators because he was once trapped in an elevator does not simply avoid being trapped in an elevator; he avoids elevators altogether. Not only avoid the dog itself, avoids the possibility of even encountering a dog. in class When the organism experiences little or no exposure to the aversive CS, what happens? it is hard to extinguish. Exposure to the aversive stimulus is so minimal that the avoidance response is extremely resistant to extinction. It is not surprising that phobic behaviors are often extremely persistent. Types of Punishment 19 When the cat sat at your feet and meowed annoyingly during breakfast one morning, you sprayed it with water. As a result, the cat did not come near the table or meow the next time you sat down for a meal. The consequence for the cat’s begging consisted of the (presentation/removal) _____ of a stimulus, and the cat’s behavior subsequently (decreased/increased) _________ in frequency. Therefore, this is an example of _______ _______. On the other hand, negative punishment involves the (presentation/removal) _____ of a stimulus following a response that subsequently results in a (increase/decrease) ______ in the likelihood of that response occurring again. Types of Punishment 20 Various types of punishment, the application of punishment. There are two basic types of negative punishment: time- out and response cost. Time-out involves the loss of access to positive reinforcers for a brief period of time following the occurrence of a problem behavior. Unfortunately, time-out procedures have little effect on the problem behavior depending on the time-out setting. 5 PSY381-Assoc Prof Hande Kaynak 10/10/2023 Types of Punishment 21 Another problem is that parents often use time-outs that are too long. The main aim of time-out is to facilitate the development of more appropriate behaviors. Those appropriate behaviors need to be reinforced, which cannot be done if the child is sitting in his room for hours. So, time-out periods should be quite brief to effectively suppress the unwanted behavior. (as short as 1 minute) As soon as the child is returned to the normal setting, more appropriate behaviors should be reinforced. Types of Punishment 22 Response cost is the removal of a specific reinforcer following the occurrence of a problem behavior. E.g. Receiving a fine (which leads to loss of money) for speeding or taking a child’s toys away for playing too roughly. The disadvantage is, you must clearly identify a reinforcer which probably will have an impact on behavior. So, it requires a more careful analysis of the situation than a timeout procedure does. Moreover, the advantage to use it is one can easily modify the severity of the punishment to suit the behavior. E.g. Slight aggression with a younger sibling could result in the loss of dessert, while more severe aggression could result in the loss of dessert and the opportunity to watch television that evening. Types of Punishment Types of Punishment 23 24 When Bobbi started to scream very loudly, her mother turned off the television program that Bobbi was watching. Bobbi’s mother is attempting to apply a (response cost/time-out) _______ procedure. When Bobbi started to scream very loudly, Bobbi’s mother made her sit in the corner for a minute. Bobbi’s mother is attempting to apply a (response cost/time-out) _________ procedure. A(n) (advantage/disadvantage) ________ of a time-out procedure is that one (does/does not) ________ have to clearly identify a specific reinforcer before implementing the procedure. An (advantage/disadvantage) _________ of a response cost procedure is that one (can/cannot) ________ easily modify (adjust) the severity of the punishment to suit the behavior. 6 PSY381-Assoc Prof Hande Kaynak 10/10/2023 Types of Punishment 25 Other differentiations (1): Intrinsic punishment is punishment that is an inherent aspect of the behavior being punished. In other words, the activity itself is punishing, such that the person performing the behavior is now less likely to repeat it. E.g. Watching an upsetting television show is intrinsically punishing if you stop watching such shows in the future because of their upsetting nature. Exercising to the point of exhaustion is for many people likely to be an (extrinsically/intrinsically) ________ punishing event. Types of Punishment 26 Other differentiations (1): Extrinsic punishment is punishment that is NOT an inherent aspect of the behavior being punished. Yet, the activity is followed by a separate event that serves to punish the activity if it subsequently reduces how frequently you do the behavior. E.g. after lighting up a cigarette, being accused. Looking at an old photo album reminds you of your loneliness as a child, the loss of a favorite pet, and a childhood friend who died. As a result, you stop looking at it. Looking at the old photo album is (intrinsically/extrinsically) ______ punishing. Types of Punishment Other differentiations (2): Problems with the Use of Punishment 27 Primary (or unconditioned) punisher is an event that is innately punishing. E.g. Electric shock, intense heat, and loud noise. Secondary (or conditioned) punisher is an event that has become punishing because it has in the past been associated with some other punisher. 28 1. 2. 3. 4. The bad taste of rotting food will likely, for most people, function as a (primary/secondary) _____ punisher, while a restaurant that has served such food will function as a ___ punisher. 5. Punishment of an inappropriate behavior does not directly strengthen the occurrence of appropriate behavior. It may even result in a general suppression of behavior. The person delivering the punishment could become an SD for punishment, with the result that the unwanted behavior is suppressed only when that person is present. Punishment might simply teach the individual to avoid the person who delivered the punishment. Punishment, especially pos punishment, is likely to elicit a strong emotional response. Punishment can sometimes elicit an aggressive reaction directed toward the punisher, might be directed toward a substitute target… 7 PSY381-Assoc Prof Hande Kaynak 10/10/2023 Problems with the Use of Punishment Problems with the Use of Punishment 29 30 6. The use of punishment, through the process of modeling, could teach the person that punishment is an acceptable means of controlling behavior. e.g. children who are abused will sometimes (not always) begin to abuse others. 7. Because punishment often has an immediate effect in stopping an unwanted behavior, the use of punishment is often strongly reinforced. e.g. If hitting one’s children has the immediate effect of getting them to stop making noise, then the behavior of hitting them has been strongly reinforced. So the parents use it more and more frequently. Yelling at your dog for chewing your slippers might teach the dog to avoid _______ rather than the slippers. Yelling at your dog for chewing your slippers might also teach your dog not to chew the slippers only when ______________. Effective Use of Punishment Brown & Martin 31 32 There are some criteria to meet for an effective punishment. 1. As much as possible, punishment should be immediate rather than delayed. 2. Punishment should consistently follow each occurrence of the unwanted behavior. 3. Punishment should be intense enough to suppress the target (unwanted) behavior. 4. Occurence of punishment shouldn’t be signalled by any other stimuli in the environment. 5. Punishment of inappropriate behavior should be combined with positive reinforcement for appropriate behavior. 6. Punishment is more effective when accompanied by an explanation. Start Shock Runway (2 m.) Goal No Shock! What kind of reinforcement? Shock Result: Rat learns to run faster and faster gradually and reaches the goal box. Question: How can we extinguish this behavior? Start Runway (2 m.) Shock Goal Safe! No shock! Result: instead of returning back to SB, animal runs to GB. Safe! Extinction occurs very slowly and in the long run No Shock 8 PSY381-Assoc Prof Hande Kaynak 10/10/2023 Brown & Martin 33 Conditioned by punishment. Organism continues running in spite of shock! Although rat receives shock, the resistance to extinguish is high, so it takes so much time to extinguish the behavior. This is the underlying mechanism of masochism. Getting conditioned through punishmentbasic of masochism One of the reasons: ‘species-specific defense reaction’ ‘running when encountered a dangerous situation’ is a species-specific response. It is impossible to teach just the opposite behavior to organism. (fight-or-flight) Running away from danger species-specific behavior. Tinsley & Renner Tinsley & Renner 34 Conducted with university students Light SD Stimulus electric shock given to the finger of participants. Behavior pressing the key in front of themselves for 60 times.shock is off. Escape from shock by pressing key. Result: by presenting the light, participants start to press the key with no taking the shock. And then experimental setting’s been changed: no shock will be given to participants, they do not know it though. So, as soon as they see the light, they start to press the key without waiting for the shock. However, if they do not press key anymore, they will not be presented the shock. When they press, they are given. Muenzinger: Punisher used as a SD 35 36 1st Group (learn discrimination) If the participant does not press the key, s/he would not receive shock. However, s/he avoided shock by pressing the key before, so whenever s/he sees the light, s/he performed the behavior without being sure whether shock will be presented or not. Result: Brown & Martin’s rat = Human-being This is also a sort of species-specific response for humans. This mechanism is innate. Punishment results in primitive behaviors. Using punishment is dangerous. It pushes the organism to perform primitive mechanisms. Goal No Reward Goal Reward Simple discrimination 9 PSY381-Assoc Prof Hande Kaynak 10/10/2023 Muenzinger Punisher used as a SD 37 Muenzinger Punisher used as a SD 38 3rd Group 2nd Group (reinforcement+punishment) Goal Goal Electric shock Goal Goal Reward Reward NO Reward NO Electric shock Correct response reward Wrong response punishment More effective learning compared to 1st group Muenzinger Punisher used as a SD For example: 39 Spoiled kid. Reaction of mother during day: beating.. Reaction of father when get home at evening: getting angry at mother & receiving chocolate. What is SD here? Result: Although the kid is beaten by the mother, spoiled behaviors are reinforced by the father. Beaten by mother is the SD for kid. Mother is beating father is rewarding Systematic pairing! Mother is NOT beating father is NOT rewarding In this sense, moderate punishers are not functional! They do not eliminate unwanted behaviors. They only work like discrimination, not like a punishment. Electric shock SD Electric shock Correct response punisher+reward “I’ll turn right, get shocked, Wrong response nothing and then get the food.” Punisher used as a SD Shock becomes a SD for reward. Result: more effective learning than 1st groups. Equal to 2nd group. Systematic pairing: Shock-Food (right box) NO Shock-NO Food (left box) General Conclusion 40 With mild/moderate punishers, behavior is recovered earlier. With severe punishers, behavior is recovered later. With much much severe punishers, behavior is NOT recovered, do not come back! When you apply punishment, Rate of you obtain response supression. response Mild Severe Very very severe Time 10 PSY381-Assoc Prof Hande Kaynak 10/10/2023 Theories of Punishment Theories of Punishment 41 42 Conditioned Suppression Theory (Skinner, 1938). Although punishment can quickly suppress a behavior, the behavior often quickly recovers when the punishment is withdrawn. It is because punishment generates an emotional response that tends to suppress any ongoing appetitive behavior. e.g. When the rat is shocked for pressing a lever that produces food, it becomes so upset that it loses interest in the food and therefore does not press the lever to obtain it. If, however, the shock is withdrawn, the rat resumes lever pressing as soon as it calms down. Punishment does not weaken a behavior but instead produces an emotional response that interferes with the occurrence of the behavior. Theories of Punishment 43 E.g. if Ahmet no longer teases his sister after being scolded for doing so, it is simply because he is too upset to pay much attention to his sister. Skinner concluded that punishment is an ineffective means for producing a lasting change in behavior. However, he used a relatively weak form of punishment. Subsequent research revealed that more intense forms of punishment, such as strong electric shocks, are capable of suppressing behavior for much longer periods of time. Theories of Punishment 44 Avoidance Theory of Punishment Dinsmoor, 1954. Avoidance Theory of Punishment. Punishment actually involves a type of avoidance conditioning. So, similarly, Ahmet no longer teases his sister after being The avoidance response consists of any behavior other than scolded for doing so because any behavior he carries out other than teasing his sister is negatively reinforced by the absence of a scolding, This means that punishment procedures are actually a form of negative reinforcement. the behavior being punished. Behavior of doing “anything other than lever pressing” reinforced by shock avoidance in a punishment-of-lever-pressing situation. This theory also assumes that punishment does not directly weaken a behavior. It simply replaces the punished behavior with an avoidance response of some sort. Any behavior other than lever pressing is negatively reinforced by the absence of shock 11 PSY381-Assoc Prof Hande Kaynak 10/10/2023 Theories of Punishment Effects of Noncontingent Punishment 45 46 The Premack Approach to Punishment. A low probability behavior can be used to punish high probability behavior. A rat that is both hungry and also tired of exercising. The rat is much more likely to eat food (an HPB) than to run in a wheel (an LPB). This means that the behavior of eating can be used as a reinforcer for the behavior of running in a wheel. The experimenter can also use running in a wheel to punish the response of eating. In a punishment procedure, if the organism does not make the response, then it will not be punished. There is a contingency here! What happens if contingency were absent in another setting? What if the aversive event was essentially uncontrollable (and even unpredictable), such that whatever you do, you are unable to influence your exposure to that event? If Sally rarely washes dishes and often bites her nails, then the behavior of washing dishes can be used to punish the occurrence of chewing nails. Learned Helplessness 47 48 Seligman and Maier (1967). 12 PSY381-Assoc Prof Hande Kaynak 10/10/2023 Learned Helplessness 49 Learned Helplessness They are neighbours but can not see 50 each other! D 2 different settings: in the first setting, there is a panel who prevents escaping. In the second one, there is a barrier! 3 groups of dogs: 1st group. Naive group (no shock. control group. Added to experiment in the 2nd setting) 2nd group. Escape group. escapable-shock condition. (they are acquired the escape and avoidance learning in 1st and 2nd setting, respectively) 3rd group. Yoked group. inescapable-shock condition (yoked to 2nd group, paired up with them. so they have to do what 2nd group do!) Learned Helplessness 51 Latency for jumping behavior S : light LightShock Pressing the paneloffset of shock (3rd group can not learn this association) When shock was turned off for the dog itself in 2nd condition, it also turned off the shock for its partner dog in the other condition (3rd condition). The dog in the escapable-shock condition had control over the shocks while the dog in the inescapable-shock did not. In the 2nd experimental setting, there was a barrier. Shuttle avoidance procedure in which the task was to learn to avoid shock by jumping over a barrier, however 3rd group can not learn avoidance learning. They made no effort to escape the shock. Due to the uncontrollable punishment for 3rd group. lack of contingency Learned Helplessness 52 It is a headache for behaviorists. Even the 3rd group get the reinforcer, why doesn’t avoidance learning occur? Answer: no connection is formed. LACK of CONTINGENCY! Time In summary, the prior exposure to inescapable shock seemed to impair the dogs’ ability to learn to escape shock when escape later became possible. LEARNED HELPLESSNESS What they learned is ‘any attempt to escape was useless’ dogs which were yoked to the 2nd group (they were unable to do anything about E.S.) 2nd group which learned how to avoid. Naive group (in their learning history, there is NO any uncontrollable, unpredictable learning) Because of what behavior, the electric shock is gone? They can not learn this. Maybe at that time point, the dog was standing and looking at the ceiling or doing anything else when the elecktric shock was gone. Absence of shock is related to what? This was not clear for the 3rd group. Loss of control over one’s environment, even when temporary, is very stressful and anxiety producing. Aversive events that occur outside of one’s control are much more stressful and arousing. Such uncontrollable and unpredictable events may be the most common source of neurotic behaviors. 13 PSY381-Assoc Prof Hande Kaynak 10/10/2023 Learned Helplessness 53 54 Is this only valid for punishment? Nope. For reward, it’s called ‘learned indolence’ (insensitivity) Punishment: a child who is beaten constantly. Reinforcement: a mom who always brings chocolate for her child every night. Learned Helplessness 55 The original experiments on learned helplessness that dogs that had first been exposed to inescapable shock had (no difficulty/diffi culty) ________learning an escape response when later exposed to (escapable/inescapable) ________ shock. It seemed as though these dogs had learned that there (is/is not) _________a contingency between their behavior and the offset of shock. This effect can be overcome by (forcing/enticing) _________ the dogs to make an escape response. As well, dogs that have had previous exposure to escapable shock are (more/less) _________ susceptible to becoming helpless when later exposed to inescapable shock. Learned helplessness may account for various difficulties in humans, including the clinical disorder known as d__________. 14
