Presentation on theme: "Chapter 10 – Aversive Control: Avoidance and Punishment"— Presentation transcript:
1 Chapter 10 – Aversive Control: Avoidance and Punishment OutlineActive Avoidance vs. Passive AvoidanceNegative Reinforcement vs. Positive PunishmentAvoidanceOrigins of the Study of AvoidanceDiscriminated Avoidance (Signaled Avoidance)Mowrer’s Two-Factor Theory of AvoidanceExperimental Analysis of Avoidance BehaviorAcquired Drive ExperiementsExtinction of AvoidanceNondiscriminated (Free-Operant) AvoidancePunishmentWhat is Punishing?Is Punishment Effective?Rules of Punishment
2 Avoidance procedures increase the occurrence of instrumental behavior Active avoidanceNegative Reinforcement (Escape or Avoidance)Learn to make a response to avoid a negative outcomePunishment procedures suppress instrumental respondingPassive avoidancePositive PunishmentLearn to withhold responding to avoid a negative outcome.
3 Origins of the study of avoidance Vladimir Becherev (1913)Participants were asked to put their finger on a metal plate.tone (CS) was followed by a shock (US) through the metal plateParticipants quickly learned to pick up their finger (CR) when they heard the tone (CS).This was considered to be a standard example of classical conditioning.But it was differentRemoving finger cancelled the USavoidance
4 Brogden, Lipman, and Culler (1938) Directly compared classical conditioning to avoidance learningGuinea pigs in a running wheel.CS- toneUS – shockThe shock stimulated the guinea pigs to run (UR)Gp1 Gp2Classical AvoidanceUS always followed US followed the tonethe tone unless the Guinea’s ranThe Avoidance group ran much more than the classically conditioned animals.
7 Discriminated Avoidance procedure (signaled avoidance procedure) There is a warning stimulus (CS)What happens after the CS depends upon what the animal does.Escape TrialDon’t respond fast enoughUS is presented until response is madeSuccessful Avoidance TrialRespond quickly enoughCS is turned off and there is no US on that trial.
8 Figure 10.3 – Diagram of the discriminated, or signaled, avoidance procedure.
9 During early training most trials are escape trials eventually the animals learn to make the avoidance response, and then avoidance behaviors take over.These experiments have often been performed in shuttle boxes.
11 Shuttle avoidance Two-way One-way move back and forth. always start each trial on one sidemove to the other.
12 Mowrer’s two factor theory of Avoidance. The problem for avoidance tasks is that once animals are performing well, the outcome (shock) doesn’t occur anymoreLooks like extinctionSo what motivates responding?Pavlovian (emotional conditioning)SignaledCS – USInstrumental After fear is conditioned the second thing that is learned is the instrumental response (jump barrier).What motivates responding?Negative RFescape from conditioned fear.
13 Experimental Analysis of Avoidance Behavior Acquired drive experiments In the typical avoidance procedure the classical conditioning, and instrumental conditioning, occur simultaneously.But if two-factor theory is correct it should be possible to train these two kinds of learning separately.1) condition fear to a CS with a pure classical conditioning procedure.CS (tone) --> US (shock) CR (fear)2) the animals are periodically presented with the CS, but an instrumental response can prevent it.No shocks are delivered in phase 2.
14 Classical conditioning Brown and Jacobs (1949).Classical conditioningConfined rats in one side of a shuttle box – the shuttle opening was blocked. Light/tone (CS) --> shock (US).22 trials.Instrumental conditioningThe shuttle box opening was opened. CS came on, and remained on until the rats went to the other side – escape.The animal was removed until the next trial.Over time the shock conditioned animals learned to cross over to the other side sooner after the initiation of the CS.Do we need a control group?
15 Extinction of Avoidance? Solomon, Kamin, and Wynne (1953)One dog avoided shock for 650 straight trials after escaping a few.Is it possible to extinguish?Flooding or Response PreventionParticipant is prevented from making the instrumental response during CSThe US is omitted (extinction)Block off door to shuttle boxCS aloneHas clinical implicationsHelp people extinguish pathological avoidanceLearn to avoid anxiety producing situations.
16 Nondiscriminated (Free-Operant) Avoidance Executive MonkeyPress key every 15 seconds to avoid shock.S-S interval – time between shocks if you don’t respondR-S interval – period of safety created if you do respondDon’t have to be the same.S-S 8sR-S 15s.you will receive a shock every 8s unless you respond.If you do respond you buy yourself 15s in the absence of shock.
17 Chimps have been trained with several schedules to keep track of to avoid shock (Koestler & Barker, 1965).respond every 5-s on a button to avoid shock (Free operant avoidance)any key that lit up with a light (Sd) had to be pressed within one second or they would be shocked (Signaled Avoidance).a separate unlit key had depressed within one second of a second Sd ( a tone; Signaled Avoidance)The Chimps were able to keep track of all of this at oncemotivated by fear?
18 So Fear motivates responding? Kamin, Brimer, and Black ran an experiment that showed that this may not be the case.Conditioned SuppressionPhase 1Rats trained to barpress for foodPhase 2Then put into a shuttle box where a tone Sd set the occasion to jump the barrier to avoid shockThere were four groups (they differ in amount of training)group 1 = avoid shock by jumping barrier on 1 trialgroup2 = avoid shock by jumping barrier on 3 consecutive trialsgroup 3 = avoid shock by jumping barrier on 9 consecutive trialsgroup 4 = avoid shock by jumping barrier on 27 consecutive trialsTestreturned to the operant chambercontinued to barpress for food.The tone was sounded
19 Think of an air traffic controller. Rats that had reached the 1, 3, or 9 consecutive trial criterion, showed considerable suppression when the tone sounded.Fear tone.surprisingly the rats that had had 27 consecutive trials of avoidance did not suppress responding.It seems that increasing experience with the situation lessens the fear.especially because the animals have learned they have control over the situation.Think of an air traffic controller.are they motivated by negative reinforcement?are they afraid all of the time?are they only afraid when they think they have made a mistake? Nevertheless Two- Factor theory is the theory of avoidance against which all others are compared.
20 As a society we tend to be very interested in punishment As we have already discussed. Punishment is some manipulation that causes a decrease in responding As a society we tend to be very interested in punishmentCrimeChildrenIs Punishment effective?Skinner initially argued that Punishment was not very effectivePossibly came to this conclusion because he was using weak punishersElectronic slap of the pawIt has been shown that punishment can be effective if the correct conditions are metDomjan’s exampleTicket for speedingChild sticks fork in electric socket.
21 Experimental Analysis of Punishment Punishment involves decreasing a behaviorin order to measure it you must initially have a baseline of that behaviorIn real life or therapeutic situations punishment is used to decrease some maladaptive behaviorSelf injurious behaviorDrug takingIn the lab, we typically first train the animals to perform a behaviorThen introduce punishment to see if it suppresses respondingThis can make things complicatedIs responding more influenced by reward or punishment?It dependsSkinner’s paw slap
22 Characteristics of the Aversive Stimulus and its Method of Introduction Types of StimuliRemember one form of punishment (positive punishment) involves presentation of an aversive stimulusShockLoud noiseslapSquirt of lemon juice in the mouth
23 negative punishment involves taking away something the organism wants Loss of positive RFTime outRemoval of personal freedomsCan’t get typical RF
24 When is Punishment most effective? Let’s go through what could be considered the “Rules for Effective Punishment”This list is backed up by empirical findings from the animal literatureThink about how we often don’t follow these rules in real life though1) punishment must be delivered soon after a response to be most effective“Catch them in the act”Delayed punishment is far less effectiveCan work well with your dog or cat.Kids?Criminals?
25 2) Punishment must be strong as possible to be effective Mild shock causes only moderate suppression of behaviorThe effect of mild shock dissipates with repeated applicationHabituationIntense, longer lasting shocks are far more effectiveKids?Criminals?
26 3) Punishment must be delivered consistently to be effective. Punishment works best if it is applied every timeHow does this compare to the effects of partial reinforcement?Kids?Criminals?
27 4) Punishment should be as strong as possible initially It is not very effective to start off weak and then increase with repeated infractions.Kids?Criminals?
28 5) Response-Contingent versus Response-Independent Aversive Stimulation Random punishment can suppress responding.Especially if intenseLearned helplessnessTypically, however, random punishment is not as effective as contingent punishmentWith random punishment in a rat you might see some initial decline in respondingbut as the animal becomes used to it responding tends to increase
29 6) noncontingent punishment decreases the effectiveness of contingent punishment If animals are shocked at random initially they take longer to associate shock with a specific stimulus, or response.Phase 1: Random shocksPhase 2: Lever press = shockUS preexposure effectKids?Criminals?
30 7) Punishment is more effective if animal if offered an alternative response to the punished responseHerman and Azrin (1964)Human smokersTwo leversPhase1 (Training)Lever 1: cigarette VI scheduleLever 2: cigarette VI schedulePhase 2 (Test)Condition 1Lever 1: loud noise; cigarette VI scheduleLever 2: cigarette VI ScheduleCondition 2Findings?How much pressing of lever 1?Condition 1?Condition 2?
31 The findings from Herman and Azrin (1964) may seem obvious, but the implications are important If you want to discourage a particular behavior, you can punish it.But it would be more effective to punish that behavior while providing the organism alternative reinforcing/reinforced behaviorsPunish playing too much x-box, but supply lots of new books/magazines?Punish self-injurious behavior, but supply other activities (painting, music, social interaction).Providing alternative behavioral opportunities can make even mild punishment much more effective.
32 8) Punishment can come under stimulus control Discriminative punishmentResponding is punished in the presence of an S+Light indicates lever pressing will result in shockOtherwise responding is RFAnimals quickly learn not to respond to the lever when a light is on.Kids?Parent present or notClassroom effectsTeacher presentSubstitute/weak teacherCriminals?Police car speeding