Instrumental conditioning Flashcards

You may prefer our related Brainscape-certified flashcards:
1
Q

Thorndike

A

If in the presence of a stimulus a response is followed by a satisfying state of affairs, the connection between the stimulus and the response will be strengthened. If the response if followed by an annoying state of affairs, the connection between the stimulus and response will be weakened

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Goal directed actions

A

Stimulus –> response –> outcome decisions

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Habitual actions

A

stimulus –> response associations

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Outcome devaluation

A

can demonstrate that SR and SOR both contribute to instrumental conditioning

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Adams and Dickinson 1981

A

Devaluation of sucrose (lever 1) leads to pressing of lever 2 (pellets) after limited training (SOR learning) but after extended training there is no devaluation effect and both levers are pressed evenly

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Clinical relevance of outcome devaluation

A

OCD – patients show an insensitivity to outcome devaluation (Gillan et al., 2011)
Stress results in less sensitivity to outcome devaluation (Schwabe & Wolf, 2011)
Actions learnt under the influence of alcohol are less sensitive to outcome devaluation (Hogarth et al., 2012, Biological Psychology)
Alcohol seeking behaviour becomes insensitive to devaluation if over-trainined (Corbitt et al., Biological Psychiatry, 2012)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Likely candidate for instrumental conditioning?

A

Striatum (caudate putamen and nucleus accumbens). It receives information from a variety of cortical and subcortical areas and then sends information back to the motor cortex.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Yin et al 2005

A

Lesions of the dorsomedial striatum in rats impairs sensitivity to outcome devaluation (i.e. lesioned rats behave in a habitual manner).

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Basolateral Amygdala

A

The Basolateral Amygdala is necessary
for the outcome representation. Balleine B W et al. 2003. Lesioned rats had far fewer actions at an extinction test for both devalued and valued stimuli

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Killcross & Coutureau 2003

Infralimbin vs Prelimbic

A

Damage to these two areas provides a double dissociation. When given limited training prelimbic FC lesions impair outcome devaluation, but infralimbic FC lesions are without effect. However, when given extended training infralimbic FC lesions fail to show a reduction in sensitivity to outcome devaluation; they remain goal-directed when sham rats display habitual behavior. Prelimbic FC lesions continue to show habitual responding when overtrained.
The double dissociation between prelimbic and infralimbic regions of the FC the outcome devaluation task demonstrates that S-R and S-O-R learning occurs independently of one another.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Prelimbic

A

Damage after acquisition of the behaviour does not impair outcome devaluation suggesting that the prelimbic region is important for encoding associations, but its not where the memory is stored (Ostlund & Balleine, 2005). Prelimbic necessary for acquiring SOR association but not necessary once this behaviour is acquired

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Infralimbic

A

Inactivation after overtraining results in habitual behaviour becoming goal-directed again. Therefore, the infralimbic region is not important for learning S-R associations, but for suppressing the influence of S-O-R associations on behaviour (Coutureau & Killcross, 2003). The fact that infralimbic FC lesions stops rats from displaying habitual responding suggests that it may be necessary acquiring S-R associations. This is not the case. If the infralimbic region is inactivated once habitual responding has been acquired, the rats return to showing goal-directed behavior. The infralimbic region is not necessary for acquiring S-R learning, but is necessary for suppressing the influence of S-O-R associations on behavior.

See slides for diagram

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Olds 1958

A

Electrical stimulation to the ventral tegmental area (VTA) leads to the release of a neurotransmitter called dopamine. In rats, electrical stimulation to the VTA appears to be rewarding. Rats will show a preference for spatial locations in which the stimulation occurred, and moreover, they learn to press a lever for stimulation.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Mobbs et al 2003

A

Humour, a pleasurable emotion, activates the nucleus accumbens, an area that receives dopaminergic input from VTA leads to strong evidence that dopamine is the brain’s pleasure chemical.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Anhedonia hypothesis

A

Pimozide treated rats stop responding for food reward. Wise 1982 - the food stopped tasting good, lost its pleasure. But maybe not. Parkinson’s patients rate foods as being as pleasant as do healthy subject (Travers et al. 1993)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Berridge and Robinson 1998

A

Incentive salience. dopamine is necessary for ‘wanting’ but not ‘liking’ of rewards. E.g. Rats with dopamine depletion will eat flavoursome foods as much as normal rats, but will they will not work for the food (Salamone et al. 2002)

17
Q

Kiyatkin, 2002

A

The dopamine response is a maximal at the time point for the lever press rather at the time point of receiving heroin.

18
Q

Schultz et al. 1997

A

Studies in monkeys have shown that dopamine also appears to be important for predicting reward. If a reward is predicted by a particular cue then the dopamine signal occurs when the cue is presented, but not when the reward occurs. The dopamine signal appears to be driven by ‘prediction error’: the difference between what was expected to occur and what did occur.

19
Q

Recorla - Wagner model

A

animal learns from discrepancy between what they expect to happen and what actually happens

20
Q

Kamin Blocking effect

A

The association between the CS and US is blocked if during the conditioning process the CS is presented together with a second CS that has already been associated with the unconditioned stimulus. Rescorla wagner model explains this The model says, essentially, that if one CS already fully predicts that the US will come, nothing will be learned about a second CS that accompanies the first CS. dopamine comes to signal how well reinforcement is predicted