Pavlovian conditioning and association

Question 1

What is learning?

Answer 1

Lecture 1, slide 9

Question 2

What experiment did Cross et al conduct in 1967? What was its hypothesis, purpose and results?

Answer 2

Lecture 1, slide 10-11, 14-15

-the rats showed preference for Mozart over Schoenberg

Question 3

What is the difference between Pavlovian learning (classical conditioning) and Instrumental learning?

Answer 3

• Pavlovian learning: learning about the relationship between different stimulus events
• Instrumental learning: learning about the relationship between an organisms own responses and stimuli

Question 4

What did Pavlov’s experiments show? What are the two explanations for a conditioned response?

Answer 4

Lecture 1, slide 24, 27

Question 5

How can we measure learning with behaviour?

Answer 5

Lecture 1, slide 29-30

Question 6

Give three examples of excitatory conditioning.

Answer 6

Lecture 1, slide 31

Question 7

What is extinction and spontaneous recovery?

Answer 7

Lecture 1, slide 36-37

Question 8

What effect does changing the intensity/salience of the US have on learning/behaviour?

Answer 8

Lecture 1, slide 38-40

-increasing intensity/salience: stronger and quicker learning

Question 9

What is congruency in learning? Give an example of a study that shows that some stimuli are resistant to association.

Answer 9

Lecture 1, slide 44

-some associations are harder to learn than others

Question 10

Describe the Pavlovian eye-blink conditioning in humans.

Answer 10

Lecture 1, slide 46

Question 11

What is habituation and what is sensitisation? What animal is used to model them and why?

Answer 11

Lecture 1, slide 47-48

Question 12

What is filled traced conditioning? What is an occasion setter?

Answer 12

Lecture 1, slide 50

-stonger learning in filled trace conditioning (experimental) than trace conditioning (control)

Question 13

What is higher-order learning? What is sensory preconditioning?

Answer 13

Lecture 1, slide 51-52, 54

Question 14

What is conditioned inhibition? How can we test whether a CS has been inhibited?

Answer 14

Lecture 1, slide 55-59

Question 15

What are the temporal constraint for eliciting a CR?

Answer 15

Lecture 1, slide 28

Question 16

How does the probability and rate of US presentation affect learning?

Answer 16

Lecture 2, slide 3

Question 17

How does contingency affect learning? What evidence is there for this effect?

Answer 17

Lecture 2, slide 5-8

Question 18

What is a conditioned stimulus? What is a conditioned response?

Answer 18

Lecture 2, slide 10

Question 19

What study did Siegel carry out in 1982? What was the method and what were the results

Answer 19

Lecture 2, slide 11-16

-context seems to be a/the CS

Question 20

What is patterning?

Answer 20

Lecture 2, slide 17-18

Question 21

What are the different theories of hippocampal function? What aspects of relational learning may the hippocampus be required for?

Answer 21

Lecture 2, slide 25, 27

Question 22

What was the lesion study and results of the study conducted by Iordanova in 2011? Can episodic memories be explained by associative learning?

Answer 22

Lecture 2, slide 28-31
-can episodic memories be explained by associative learning: maybe not, because in this study, hippocampal lesions don’t destroy the rats’ ability to learn associations, but seems to be important for configuring more complex episodes.

Question 23

What is Pavlovian conditioning used to study?

Answer 23

Lecture 2, slide 33

Question 24

Explain the gradual learning curve?

Answer 24

Lecture 2, slide 34

Question 25

Is learning gradual or a result of insight?

Answer 25

Lecture 2, slide 35-43

Question 26

What is Hull’s computational theory of error correction in learning? How can it be used to explain acquisition and extinction?

Answer 26

Lecture 2, slide 44-50

Question 27

What is overshadowing?

Answer 27

Lecture 2, slide 51

-more salient cue interferes with the learning of weaker cues

Question 28

What is Kamin’s blocking effect?

Answer 28

Lecture 2, slide 52

-prior learning of CS1 interferes with learning of CS2

Question 29

What is the Rescorla-Wagner theory? What learning effects/phenomena does it explain that Hull’s theory cannot?

Answer 29

Lecture 2, slide 53-64

Question 30

What is relative value and its importance?

Answer 30

Lecture 2, slide 65

Question 31

What study did Wagner conduct in 1968? What is relative validity?

Answer 31

Lecture 3, slide 3-11

• in the True Discrimination group, CS2 and CS3 are more valid than CS1. They reliably signal occurance and nonoccurance of US.
• in the PseudoDiscrimination group all 3 cues are equally valid (50% chance of US)
• if one cue is mor valid it interferes with the learning of another cue

Question 32

What is over expectation?

Answer 32

Lecture 3, slide 11-12

Question 33

What is super conditioning? What is protection from extinction?

Answer 33

Lecture 3, slide 14-20

Question 34

How is Schizophrenia related to prediction error correction and selective learning?

Answer 34

Lecture 3, slide 21, 27

Question 35

What study was conducted by Serra et al in 2001?

Answer 35

Lecture 3, slide 28-29

Question 36

What effect does increasing the strength of the US have on blocking? What was the study carried out by Dickinson et al (1976)? What are positive and negative prediction errors?

Answer 36

Lecture 3, slide 32-33

-if the strength of the US is unexpected then there will be learning

Question 37

What is latent inhibition?

Answer 37

Lecture 3, slide 34

Question 38

What role does dopamine have in learning? What type of DA responses can you have in learning? What is reward substitution?

Answer 38

Lecture 3, slide 41-44

Question 39

What study was conducted by Waelti et al in 2001?

Answer 39

Lecture 3, slide 45-47
-RPE: reward error prediction
In their study, Kamin blocking was used to generate prediction errors. Monkeys were trained to know visual stimuli A and B (control) were followed by juice (reward) and not followed by a reward respectively. This conditioning was followed up with compound training where BY+ triggered more behaviour from the monkey than B- (which triggered no behaviour after conditioning) whereas AX+ triggered the same amount of as behaviour as A+. DA neurons of the VTA showed more activity when the monkey was presented with BY+ than just B-, but activity remained similar for both AX+ and X+. X- minus has been blocked by A+ and so it cannot be learnt about. This is a failure to form new cue-reward associations due to the lack of RPE signal. The monkey has already been conditioned to expect a reward when presented with A+, so the presentation of a reward with AX+ does not deviate from what is expected and nothing is learnt from the experience. On the other hand, the presentation of BY+ was not predicted by either stimulus and therefore receiving a reward should have generated a positive prediction error.

Question 40

What study was conducted by Tobler et al in 2003?

Answer 40

Lecture 3, slide 48-49

-X is learnt to be inhibitory

Question 41

What is the difference between latent inhibition, overshadowing and blocking?

Answer 41

classical conditioning is undermined by preexposure to the CS (latent inhibition) or the concurrent presentation of either a more salient CS (overshadowing) or one that has already been associated with the US (blocking).

• latent inhibition: preexposure to CS prevents learning/acquisition of a CR
• blocking: CS1 has already been associated with the US. This blocks learning of CS2
• overshadowing: CSa is more salient than CSb. Concurrent presentation of CSa and CSb, reduces learning of the less salient CSb.

Question 42

What are some causal manipulations to determine the function of a structure?

Answer 42

Lecture 4, slide 2

Question 43

What are optogenetics? What are the steps to optogenetics?

Answer 43

Lecture 4, slide 4

Question 44

What study was carried out by Steinberg et al., 2013?

Answer 44

Lecture 4, slide 6-7
-RPE: reward error prediction
-they utilised optogentics to demonstrate that DA released by dopaminergic cells of the VTA behave as positive RPE signals.
-In the first stage of the experiment two groups of rats (a blocking group and a control group) were conditioned with a single cue—tones A and B respectively. This was followed by compound testing where AX (tone and light) was presented to both groups followed by a sucrose drink (reward). A test was subsequently performed to determine how much the rats had learnt about the light stimulus (X). As expected, the blocking group had not learnt much about the light due to blocking and the lack of a RPE, while the control group were able to learn about the light. The differences in learning were significantly different at a 99.9% confidence level since p=o.oo3.
-The aim of the next stage of the experiment was to determine whether increasing the DA firing rate of VTA dopaminergic cells at the point of US administration would be sufficient to unblock the learning of light in the blocking group. There are three groups involved in this method: PairedCre+, UnpairedCre+ and PairedCre-. The Cre- group consists of wildtyepe rats. The two Cre+ groups consitst of transgenic rats which express Cre recombinase under the control of the tryosine hydoxylase promoter. All of three groups received identical injections of a Cre-dependent virus expressing channelrhodopsin-2 into their VTA. The Cre- individuals did not take up the channelrhodopsin due to their lack of Cre, whereas the channelrhodopsin’s were inserted into the VTA dopaminergic neural membranes of the Cre+ groups, enabling their DA neurons to increase their activity in the presence of light.
All 3 groups were trained in a way that normally causes blocked learning. However, during the compound training, the two paired groups received the light/optical stimulation paired with the US, while the unpaired group received the stimulation during the inter-trial interval.
The results of this experiment succeeded in showing that VTA DA is involved in learning through acting as a RPE signal. The PairedCre+ group showed learning of the light cue unlike the other two groups. Typically due to blocking and the lack of an unexpected stimuli, the DA neurons would not fire, however the artificial firing of DA in response to light mimics the naturally occurring prediction error signal, allowing the rats to learn.

Question 45

What are the limitations of the studies that have proposed dopamine as a learning signal?

Answer 45

Lecture 4, slide 8