blocking and extinction -L2 Mitchell

Question 1

what are associative links in blocking?

Answer 1

explained as failure to form a link

Question 2

what is the prediction error in blocking?

Answer 2

A&B = CS’s, + = US
A+ then AB+

US is predicted by A by the end of A+ trials – no prediction error
US is not surprising on AB+ trials (no prediction error), so not processed
No B-US association learnt (AB+ trials are the only time B appears)
Learning about B is ‘blocked’ by A

Question 3

why is blocking important?

Answer 3

Before blocking was discovered: it was believed that an association forms when two stimuli are paired
Hebbian learning (Hebb, 1949)

In blocking, Cue B -> US on AB+ trials, but there is little learning
Shows that learning isn’t just pairing of stimuli

Central to all contemporary learning theory
no prediction error on AB+ trials following A+ trials -> the US is not suprising

Question 4

how can the US become unsuprising and less potent?

Answer 4

AB+ Blocking (A+ strong)
CD+ Overshadowing (Weaker A+)

(US is surprising on initial CD+ trials)

Question 5

how can prediction error be formalised?

Answer 5

the difference between what you predict will happen and what actually happens

Question 6

what is the Rescorla and Wagner 1972 formula?

Answer 6

ΔV = α(λ –ΣV)
ΔV = learning: change in associative strength on this trial – in this case we’re interested in the blocked cue B on AB+ trials
λ = total associative strength supported by outcome
ΣV = the expected outcome (given all cues present A&B)
α = attention to cues – this doesn’t change with learning

On AB+ trials, λ = ΣV, due to presence of the pretrained ‘blocking’ cue A

captures that you only learn when there is a prediction error

Question 7

what is an alterative to r & w formula?

Answer 7

attention to CS’s
-Signals of important events are attended to especially well
The salience of the CS “α” is not fixed (contrary to R&W’s claim): α can change with experience/learning
Another way blocking might occur

Question 8

what did Mackintosh 1975 predict?

Answer 8

blocking due to changes in attention to cues
-A+ then AB+

The most predictive cues will increase in salience/α (e.g., cue A)
Relatively less predictive cues will decrease in α (e.g., cue B on AB+ trials)

Not much learning about cue B

Question 9

how can attention be measured?

Answer 9

eyetracking

Question 10

what did Beesley & Le Pelley, 2011 research?

Answer 10

eye gaze in a learning task
-No shocks
-Present stimuli on the eyetracker screen
-Cues: chemicals (Addexium, Rezaline)
-Outcomes: symptoms in Mr X (itchiness, nausea)
tested AB plus and JK plus trials

Question 11

what were the results of Beesley & Le Pelley?

Answer 11

Attention to cue A is high (it is a very good predictor of the outcome)

dwell times on pretrained cues were greater than on blocked cues across stage 2 training

Attention (eyegaze) to ‘blocked cue’ B is reduced compared to overshadowing controls J and K

Less attention/processing of Cue B would explain why there is less learning about that cue on AB-O1 trials.

Question 12

how can blocking be summarised?

Answer 12

Blocking is seen in humans (e.g., Beesley & Le Pelley, 2011) and in non-humans (e.g., Kamin, 1969)

-Prediction error – no learning because US isn’t surprising on AB+ trials
Reduced attention to cue B on AB+ trials due to higher salience better predictor cue A
-salience mechanism in humans (Beesley & Le Pelley, 2011)
Blocking is not all about prediction error

Question 12

why is extinction important?

Answer 12

Some aspects of extinction don’t fit with Rescorla Wagner
Clinical relevance: phobias and PTSD

Question 13

what do extinction curves show?

Answer 13

Initial learning curve – US becomes less surprising over trials
Increase in associative strength slows down
Extinction curve – absence of US becomes less surprising over trials
Decrease in associative strength slows down

Question 14

what are some limitations to Rescorla and Wagner?

Answer 14

Renewal
The CR returns in a different context from extinction -> extinction is context specific
Spontaneous recovery
The CR returns with a delay between extinction and test

Question 15

true or false: extinction is specific to the context in which it was learnt

Answer 15

true

Question 16

what did Bouton say on extinction?

Answer 16

Extinction is new learning of an inhibitory link
That inhibitory link only operates in the context in which it was learnt
Rescorla Wagner is wrong: The original CS-US link does not weaken

Question 17

is exposure therapy and extinction the same?

Answer 17

Exposure therapy seems to be the same as extinction
Both show renewal and spontaneous recovery

How to avoid context specificity of exposure therapy (Laborda et al, 2011):
Have therapy in lots of different contexts
Spread exposure treatments over a long period of time

Question 18

what is associative strength?

Answer 18

the strength of the connection between internal representations of the conditioned and unconditioned stimulus which determines the strength of the conditioned response

Question 19

define blocking

Answer 19

the disruption in conditioning with one element of a compond when it is accompanied by another element that has already been paired with the unconditioned stimulus

Question 20

what did Kehoe et al 1994 find?

Answer 20

in an eye-blink experiment w paired a light with a mild shock to the cheek
-measure of conditioning = percentag of trials for which a response was observed in each daily session
-at first, likelihood of a response to the light was low then rapid increase occured - following the typically learning curve!

Question 21

define associative strength

Answer 21

the strength of the connection between internal representations of the conditioned and unconditioned stimulus which determines the strength of the conditioned response.

Question 22

How does US intensity effect the rescorla-wagner model?

Answer 22

A more intense US increases λ (the maximum associative strength), leading to larger prediction errors (λ − V) early in training.
->This results in faster learning and a stronger conditioned response.
->In contrast, a weak US has a smaller λ, producing smaller ΔV per trial and slower learning.

Question 23

how does CS intensity effect the rescorla-wagner model?

Answer 23

A more intense CS increases α (the salience of the conditioned stimulus), which leads to larger changes in associative strength (ΔV) per trial.
-> This results in faster learning.
However, CS intensity doesn’t affect the maximum strength (λ), only the rate at which it’s reached.
-> influence is confined solely to determining the speed at which conditioning takes place, defined as a learning-rate parameter that reflects the associability or conditoinability of the CS.

Question 24

define conditionability

Answer 24

the ease with which the strength of a conditioned response to a stimulus can be increased by subsequent conditioning.

Question 25

What did Kamin 1969 do?

Answer 25

-conditioning was conducted with a noise prior to compound conditioning with the noise and a light -testing revealed that despire being paired with shock during compound conditioning the light virtually failed to elicit a response -according to rescorla & wagner 1972, the explanation is that the blocking effect rests with the original training with the noise, which will ensure ot has maximum associative strength when conditioning commenses with the noise-light compound.

Question 26

define overshadowing and give an example

Answer 26

the disruption of conditioning with one stimulus because of the presence of another stimulus -when animals are conditioned with two stimuli in a compound, each will gain less associative strength than if they were seperately paired with the same US, the presence of once CS is said to results in overshadowing with the other!

Question 27

what is the CS-US contingency?

Answer 27

the degree to which the us occurs during the cs, -positive: the US is more likely in the presence than the absence of the CS -negative: the US is more likely in the absence than the presence of the CS -zero: the US is equally likely during the presence and the absence of the cs

Question 28

what rescorla study represents the CS-US contingency?

Answer 28

rescorla 1968 -rats were trained to press a lever for food in a conditioning chamber before being given a number of sessions in a different chamber where a 2 minute tone and shock were presented -tone assessed by presenting the tone to the rats while they were again responding for food in the original chamber. -> strength of conditioning differed dramatically in the four groups, even though the tone shock pairings were identical for all subjects. show importance of contingency in conditioning!

Question 29

What does Hearst & Jenkins study show?

Answer 29

the probability of food was greater in the ansence than in the presence of an illuminated key -results indicate that when there is a negative CS-US contingency it will result in inhibitory learning! -later rescorla & wagner claimed that the context in which the CS, can be constructed as a stimulus.

Question 31

what are some strengths of the rescorla-wagner model?

Answer 31

-provides ideal account of the conditions of leaning but not perfect as yet to be developed, the theory is unable to explain certain results (Miller et al 1995) -good account of many facts of compound conditioning, but not all effects associated with blocking and overshadowing are consistent with it

Question 32

what are some concerns of the rescorla-wagner model?

Answer 32

-model doesn't adequately account for the role of suprise in conditioning (Holland & Kenmuir, 2005), later revealed by Dickinson et al 1976, in a blocking experiment using conditioned suppression -inhinition is conceptualised as negative associative strength as occasionaly a conditioned inhibitor can elicit its own CRs!

Question 33

what happened in the Dickenson et al 1976 experiment?

Answer 33

stage one: 2 groups received excitatory conditioning in which on every trial a light was followed immediately by two shocks sepeared by an interval of 8 seconds, both groups then given compound conditioning with the light and a clicker Group C compound followed by two shocks but for Group E only the first paie of shocks accompanied the compound. - results: Introducing an unexpected second shock 8 seconds after the initial shock reduced the blocking effect - for group C so blocking for Group E was attentunated.

Question 34

why is the rescorla-wagner model taken seriously?

Answer 34

-this theory only concern with making qualitative predictions about the relative performance of animals that are treated differently -> poss to predict a conditioned response will be stronger after one type of treatment than aother -> hypothesis deriven led experiments -example of benefits of a particular style of theorizing -seen as an important senior relative of a number of more recent theories of conditioning and associative learning. -theory extends to beyond study of animal learning! Seigal & Allan (1996) show how this theory enhanced our understanding of humans in verbal learning, category formation, casual judgements, reasoning etc.

Question 35

what did Pavlov study?

Answer 35

an oreinting response -Pavlov (1927, p. 12) noted that novel stimuli often elicit an OR or, as he referred to it, an investigatory reflex that allows the animal to investigate any changes that occur in its environment.

Question 36

define intradimensional shifts

Answer 36

the selection of two stimuli for a discrimination from a dimension that provided two stimuli for an earlier discrimination

Question 37

define extradimensional shifts

Answer 37

the selection of two stimuli for a discrimination from a different dimension to one that provided stimuli for an earlier discrimination

Question 38

what happens in an Intradimensional shift?

Answer 38

An ID shift occurs when an animal or human learns a new discrimination within the same perceptual dimension (e.g., switching from red vs. green to blue vs. yellow — all still colors). It involves applying the same attentional rule to new stimuli of the same type. Learning is usually fast because attention is already focused on the relevant dimension.

Question 39

what happens in an extradimensional shift?

Answer 39

An ED shift occurs when the relevant dimension changes entirely (e.g., from color to shape). The learner must disengage attention from the previously relevant dimension and shift to a new one. Learning is typically slower, showing the cost of shifting attention to a new dimension.

Question 40

What was the purpose of George and Pearce's (1999) pigeon discrimination study?

Answer 40

To investigate how pigeons learn to discriminate between visual features and how attention is allocated depending on which stimulus dimension (color vs. line orientation) is relevant.

Question 41

What stimuli were used in George and Pearce (1999)?

Answer 41

Each trial displayed two squares on a screen: One square was colored (either blue or yellow) The other had line orientation (forward- or backward-slanting black and white lines)

Question 42

What training did Group IDS receive in the George & Pearce (1999) study?

Answer 42

Group IDS learned a color-based discrimination: Pecking was rewarded when the colored square was yellow (+) No reward when it was blue (−) Line orientation was irrelevant.

Question 43

What training did Group EDS receive in the George & Pearce (1999) study?

Answer 43

Group EDS learned a line-orientation-based discrimination: Pecking was rewarded for forward-slanting lines (+) No reward for backward-slanting lines (−) Color was irrelevant.

Question 44

What was the key difference between Group IDS and Group EDS in George and Pearce’s experiment?

Answer 44

Group IDS: Attended to color (intradimensional shift setup) Group EDS: Attended to line orientation (extradimensional shift setup) Each group learned to focus on different stimulus dimensions.

Question 45

What is the core idea of Mackintosh’s (1975a) theory of attention in learning?

Answer 45

Mackintosh (1975a) proposed that animals pay more attention to stimuli that are better predictors of significant outcomes (e.g., food or shock). A CS gains more attention—and thus supports faster learning—if it predicts the US better than other stimuli present on the trial.

Question 46

what is a strength of Mackintosh 1975 theory?

Answer 46

- proposing that blocking is entirely a result of subjects ignoring the stimulus added for compound conditioning -The attention paid to the light will be relatively high because of its novelty, and Equation 3.11 predicts that it will gain in associative strength. Animals will also have the opportunity of discovering on this trial that the light is a much poorer predictor of the US than the tone. The theory therefore predicts that on future conditioning trials they should pay very little attention to the light and further increments in associative strength to this stimulus will be slight.