Final Exam Notes PDF

Preview

Summary

Final exam notes from learning classes...

Description

Chapter 5→ Underlying Processes in Classical Conditioning Preparatory-Response Theory ● The CR prepares the organism for the coming of the US Compensatory-Response Model ● When drugs are the US the CR is known as a conditioned compensatory response ● CCR is in opposition to the effects of the drug ● This is so the body can attempt to maintain homeostasis ● Does not always occur, sometimes the CR to the CS associated with the drug is drug-like ● US = Drug ● UR = drug’s effect ● CS = cues associated with drug ● CR = conditioned compensatory response (CCR) ● Helps explain withdrawal and tolerance effects ● Withdrawal = when give CS but don’t give US Rescorla Wagner Model Triangle V → change of V (associative change→ how much associative strength has the CS gained) Magnitude of US→ how intense is US V = what was the condition responding coming into this trial (will be 0 when the CS has never been paired before) When V is lambda delta V will equal 0 Biggest delta V occurs when trial 1 occurs ● Explains overshadowing by saying the more salient CS acquires more associative value than the less salient CS ● Explain blocking effect by saying the increase in associative value of the CS depends on the “surprisingness” of the US ● This model makes the surprising prediction that pairing two CSs together can lower their conditioned association with the US ○ Overexpectation effect ○ Phase 1 pair CS with US1 many times, CS2 and US1 paired together many times, both have maximum associative strength (excitatory conditioning) ○ Pair CS1 and CS2 and presenting US as well, should be no change in associative value ○ They start to lose associative strength → CR goes down ○ Delta V will be negative → proves they lose associative strength

Factors Affecting Phobic Conditioning ● Observational Learning→ the US (display of fear by others) results in the UR (fear) ● Temperament→ a person’s emotionality and reactivity ● Preparedness→ innate to fear some objects more than others -- these objects serve as better CS’s for fear conditioning ● Locus of Control→ external or internal, internal locus of control after brief exposures to the CS ● Incubation→ strengthening of fear conditioning after brief exposures to the CS ● US Revaluation→ being exposed to a stronger US after conditioning has taken place can increase the CR Treating Phobias ● Counterconditioning→ pairing CS with something positive ○ The positive CR creates a reciprocal inhibition to the previous negative CR ● Systematic Desensitization (most effective) ○ Training the patient to physically relax ○ Establishing an anxiety hierarchy of the stimuli involved ○ Counterconditioning relaxation as a respond to each feared stimulus beginning first with the least anxiety-provoking stimulus and moving them to the next least anxiety-provoking stimulus until all of the items listed in the anxiety hierarchy have been dealt with successfully ■ Biofeedback instrumentation has often been used to ensure that the patient is truly well relaxed before going the next higher item in the anxiety hierarchy ■ Examples; pulse rate, respiration rate, and electro-dermal responses. Patients can monitor their own biofeedback in an effort to better control it ■ Could be counterconditioning or could be extinction ● Flooding ○ Response prevention technique ○ Relies of extinction ○ Could aggravate the condition if not done correctly ● Exposure Therapies ○ Patient approaches feared object as closely as possible ○ Stays there until anxiety fades ○ Approaches object closer ■ Often employs participant modeling where the therapist touches the object and models the correct behaviour ● Aversion Therapy ○ Pairing a desired object/action with an aversive stimulus to make it less desired ○ See clockwork orange

○ Ie. addiction counselling Medical Applications ● Immune response ● Asthma response (hissing sound (CS) paired with an allergen (US) that caused asthma response (UR), animal had an asthma attack after pairings of CS and US when CS was just presented) ● Placebo Effect

Chapter 6→ Instrumental and Operant Conditioning ● A category of learning that involves situations in where the presentation of the US depends on the individual’s actions ○ Certain response will lead to a certain outcome ○ S D → contextual/discriminative stimuli ○ R→ response ○ S R → reinforcing stimulus ○ S P → punishing stimulus What’s the Difference? ● Since the response is instrumental in producing the outcome it is referred to as instrumental behaviour. The consequence is referred to as the reinforcer/reward or punisher ● Sometimes instrumental behaviour is called operant behaviour. This is the case if the response is defined in terms of a particular operation or manipulation of the environment Thorndike and Skinner ● Discrete Trial Method→ the animal has the opportunity to perform the instrumental response only at certain times, as determined by the experimenter (Thorndike) ○ Straight alley and T-mazes are often used in the discrete-trial method ● Free Operant Method→ the Skinner box “frees” the researcher of having to continually handle the animal (Skinner) ● The rat presses a lever to receive a reward. Whenever he wants. ● The rate of responding becomes the primary measure of behaviour in experiments using the free-operant method Establishment of an Instrumental Response ● There are three main ways to train an instrumental response ○ Using an old response in a specific situation ○ Learning a new response from old components ○ Learning a new response (shaping)

Shaping New Responses ● Instrumental conditioning can be used to construct entirely new responses ● How can a new response be reinforced that never occurs naturally? ○ Due to the variability of behaviour ○ We get the new response to occur by reinforcing naturally occurring behaviours that are nearest to the final behaviour we want to participant to perform ■ This procedure is called shaping ○ You must have ■ A goal ■ Starting criteria ■ Reinforcer ■ Training steps ■ Target response ○ Shaping is defined as the reinforcement of successive approximations ■ When you move onto next step you stop reinforcing last step

Name

Response Produces

Result

Positive Reinforcement (Reward)

Appetitive stimulus

reinforcement

Positive Punishment

Aversive Stimulus

suppression

Negative Reinforcement (avoidance learning/escape behaviour)

Prevents aversive stimulus

reinforcement

Negative Punishment (omission training)

Takes away positive stimulus

Suppression

Example→ want child to increase likelihood to clean room ● Positive reinforcement → more tv time if clean room more ● Negative reinforcement→ if you clean your room you do not have to do the dishes OR if you clean your room you won’t get yelled at Example→ want son to decrease the likelihood of playing rough with baby sister ● Negative punishment→ when you treat sister roughly take away gameboy ● Positive punishment→ play roughly with sister get hit by parents

● A stimulus that is associated with a primary reinforcer is called a conditioned (secondary) reinforcer ● A marking stimulus can also be used to help shape a behaviour in the absence of a reinforcer What is Learned? ● At first the most important relationship was thought to be R-S*. S* = ( S R or S P ) ● Thorndike Law of Effect→ Thorndike proposed that the relationship b/w the S D and R is learned ○ The S* simply serves as a catalyst for learning the S D − R association ● Later, S D − S association was thought to play a large part in the learning ● It is also possible to think of it as higher order learning S D (R-S*). In this way of thinking the S D acts as an occasion setter for the R-S* to occur in ● Experimental investigations have provided evidence for all 4 types of associations Constraints on Learning ● As with classical conditioning there are innate mechanisms that predetermine which responses can be most readily associated with a reward ● Some responses are hard to pair with a reward ● Some learned responses tend to change over time (instinctive drift) → animals have an innate behaviour they want to perform but it interferes with the learned behaviour ○ Pig that was trained to take token from machine and put it in a slot then food would come out of machine ○ What would happen over time it would get token and instead of putting it in slot they would put it on ground and push it around and would not do the trick b/c it started to treat the token as it was food; association of food and token was getting very strong ● Autoshaping → a light comes on then food is given, sometimes light would be on other end of cage what eventually would happen is light would come on at opposite end of

cage where food is given, bird would go over to light and peck at it then go over to the food Latent Learning ● Rats were put in a maze and one third were given a food reward when they reached the end of the maze and two third were not ○ Second group started to get rewarded ○ All groups learned the maze but some groups were not reinforced for their behaviours so no incentive to go through maze quickly ● This is a perfect example of the learning performance distinction ● The unrewarded animals had learned about the maze (latent learning) but the ‘test’ for learning did not allow them to demonstrate this knowledge Reinforcer ● Quality - better quality more effective reinforcer ● Quantity - higher quantity better reinforcer ● Temporal contiguity - it is important that the reward follows the response closely ○ Immediate vs delayed reinforcement Intrinsic vs Extrinsic Reinforcement ● Intrinsic reinforcement: the behaviour itself is reinforcing ● Extrinsic reinforcement: t he behaviour brings about a consequence that is reinforcing ● Giving extrinsic rewards can decrease intrinsic motivation when: ○ Reward is tangible (ie. not praise but money) ○ Reward is expected ○ Reward is not based on quality of performance Shifts in Reinforcer Quality ● Prior experience with a reward determines a reward’s effects ● Positive contrast (S → L) means the reward will be more effective ● Negative Contrast (L → S) means the reward will be less effective Superstitious Behaviour ● Skinner → birds were in experimental chambers that delivered food every 15 seconds no matter what they did ● Came back and found 6 out of 8 birds performing very defined behaviours ● Seems that the birds were performing terminal responses. Responses that occurred just before they got food Learned Helplessness

● Interference with the learning of a new instrumental response as a result of exposure to an inescapable aversive stimulus ● Previous experience where their behaviour did not affect the outcome. Previous experience interferes with them learning in the future where their response could affect the outcome ● Experiment for learned helplessness ○ 3 groups of animals ○ EYR → Escapable Yolked Control group ■ Yolked: Whatever happens to another animal happens to you. EX: another animal presses lever, you get food too but you don't have a lever ○ E → escapable shock group. When shock came on, they knew if they went to other side of cage the shock would stop ○ Y → yolked inescapable shock. Were not the ones controlling when the shock stopped. ○ R → control group. No shock ○ Conditioning phase → all three groups got an escapable shock. Response they had to get rid of shock is not the same as the first behaviour. ○ DV → how fast they learned to avoid the shock ○ Y group learned slow. Before nothing they did prevented shock ○ Group E learned rapidly ○ Group R was fast ● Learned helplessness research teaches us that reinforcers may reinforce not only if they are inherently pleasant but also because they provide information that allows the learner to associate the actions with outcomes that are under their control Chapter 7→ Reinforcement: Schedules and Theories ● Perfect contingency b/w response & reward rarely occurs in the real world ● Schedule of Reinforcement→ i s the set of rules that determines how & when a response will be followed by a reinforcer (4) ○ Influence how an instrumental response is learned and how it is maintained ○ This is studied in free-operant chambers (skinner box) where the rate and pattern of responding can be observed ● Cumulative record → graph that looks at how the responding takes place, how does it break down? ○ Y-axis represents the cumulative # of responses ○ How much have you responded since we started? ○ Slope of responding = how fast they response → steep = fast, shallow = slow ○ X-axis represents how much time has past (seconds, minutes, hours) Ratio Schedule

● The number of responses a subject has performed since its last reinforcer determines if it is rewarded. ● How much time it takes to make the responses does not matter ○ Fixed Ratio (post reinforcement pause) → fixed # of responses before you get outcome, happens every single time, does not matter how much time has passed, typically see post reinforcement pause ■ Ie. FR4, 4 responses have to be done before you get outcome ■ FR1 → would mean every time you respond you would get reinforced ○ Variable Ratio → can vary from reinforcement to reinforcement, average, randomly generated by researcher, varies every time, never a certainty of ratio, uncertainty, do not normally see post reinforcement pause we see they same amount performed every time, produces highest steady state of responding, sharpest slope, animal has uncertainty → not sure how many responses they need to do ■ Ie. VR4, on average an animal has to respond 4 times ■ Example → slot machine, do not know how many times you must put in money before you win Interval Schedules ● Reinforced by when you respond, only have to respond once but have to respond at the right time, certain amount of time must pass ○ Fixed Interval (scallop) ■ Increased rate of response just before reinforcement ■ Ie. FI 3mins, 3 minutes has to pass before you get reinforcement ○ Variable Interval ■ Steady rate of response ■ Ie. VI 3min, average of 3 minutes reinforcement is given, animal is never sure when it will actually come ● Interval schedules produce slower rates of responding ● Limited Hold ○ A window of opportunity for animal to complete task ○ If completed after this window, they will have to wait a whole interval before it has another opportunity for reinforcement Chained Schedules ● Involves a sequence of responses ● The primary reinforcer is NOT presented until all the components of the chain have been performed ● Good for behaviours that are made up of linked behaviours put together ● Training response chains → which is the most effective way to train behaviour?

○ Backward Chained Schedules→ we learn the last behaviour first then we work to the first behaviour ○ Forward Chained Schedules→ we learn the first behaviour first then work our way to the last behaviour ○ People think backward chained schedules are more effective than forward chained Extinction of Reinforced Behaviour ● Researchers are interested in how various reinforcement schedules affect persistence of responding when extinction procedures are introduced ● How long will the animal keep doing behaviour after you stop reinforcing them ● Persistence of Responding → lasts longer if the animal was trained on a partial reinforcement schedule (PRF) as opposed to a continuous reinforcement schedule (CRF) ● Partial reinforcement extinction effect (PREE) Why? ● Discrimination Hypothesis→ extinction is more difficult to detect after PRF than CRF (wrong) ● Sequential Theory→ PRF training allows the animal to associate non-rewarded behaviour to be associated with reward ● Frustration Theory→ states PRF training taught animals to respond in the face of frustration Theories of Reinforcement ● A theory of reinforcement has to answer 2 questions about instrumental conditioning: ○ What makes something an effective reinforcer? ○ How does a reinforcer exert its effects and increase the probability of the reinforced response? Thorndike’s Law of Effect ● He just named the reinforcer “satisfier” w/o answering the question ● A reinforcer establishes an association b/w the S and the R ● Said that the S-R association made it so the S elicited a R Drive Reduction Theory (Hull) ● Reinforcers are effective in reducing the drive state and returning the animal to homeostasis ● Primary and conditioned reinforcers ● Primary and acquired drives ● Accepted Law of Effect ● There are reinforcers that do not follow this prediction

The Premack Principle ● “Good theories” ● Premack stated the issues of reinforcement in terms of responses, not in terms of stimuli ● Also called the differential probability principle ● Focuses on question of what makes something a reinforcer ● The reinforcing response is defined as the behaviour that is most likely to occur ● What the animal most wants to do at that time Response Deprivation Hypothesis ● The instrumental response is reinforced if it allows the animal access to the reinforcer response ● The reinforcer response is any behaviour that the animal has been deprived of performing as much as it normally would ● Anything can be reinforcer response if animal or human is not doing it as much as it fulfills them ● Predicts (accurately) that low probability responses can act as reinforcers Behavioural Regulation Approach ● This theory borrows from the idea of biological homeostasis and creates the idea of behavioural homeostasis ● There is a preferred distribution of activities for each organism (behavioural bliss point) that is defended The Six Tools of Compliance Experts Tools of Compliance 1. Reciprocation 2. Commitment and Consistency 3. Social Proof 4. Liking 5. Authority 6. Scarcity Why do tools of compliance work? a) We need shortcuts to conserve out cognitive energies b) We have certain societal rules we feel obligated to follow Reciprocity ● Rule: we will feel the desire to repay someone who has done something for us ● Even if ○ We did not ask for the favour ○ We do not value the gift

● Also ○ The repayment can be MORE than the gift ○ It is difficult to refuse a gift freely given ○ Still powerful motivator even if you don’t like the person ● Why is it so powerful Door In The Face Technique ● When someone is more likely to do something if we first request something larger ○ Works b/c other person feels like you made a concession and they should now reciprocate ○ There is a contrast from the large to the small request ○ Example ■ Coca-cola for raffle tickets ■ Krishnas in the airport ■ The watergate scandal ■ Dividing up the money ■ Boy scouts and chocolate bars Commitment and Consistency ● Once we make a choice we will encounter personal and societal pressures to behave consistently with our commitment ● This holds true even ○ If the original reasons for our commitment change or are taken away ○ Our automatic commitment is now not to our benefit ● Follows from cognitive dissonance theory Foot in Door Technique → consistency ● We are more likely to commit to a larger request if we first agree to a smaller request ● Works because the agreement to have a smaller request commits us to have a certain self perception about ourselves Examples ● The hot new christmas toy → method to the madness, couldn’t get tickle me elmo so parent may get something bigger and better, get the toy in January ● Car sales techniques → having unnecessary features added when buying a car which adds a lot of money to total car price ● Chinese prisoner of war camps → sign statement denouncing government of America, would give them more food if they signed it, made them read it out loud and put it on video, attitudes against America actually changed after this ● Quitting smoking → publicly committing will stay more committed than people who don’t publicly commit to it

● Juries → publicly announcing how they will vote are likely to not change opinion ● Energy conservation → people would commit to energy conservation if name was on a list Social Proof ● We determine what is the correct behaviour by finding out what other people are doing ● More likely to work ○ When more ppl are doing it ○ When the ppl doing it a...