Reward Representation and Learning Flashcards
Which of the following does O’Doherty (2004) say are necessary for learning to increase rewarding and decrease punishing experiences?
A. Represent the reward value of stimuli, update the value of stimuli over time, and encode these values in long term memory.
B. Represent the reward value of stimuli, establish predictions of future reward and punishment, and use the predictions to guide behavior.
C. Estimate the intrinsic value of stimuli, establish predictions of future reward and punishment, and use the predictions to guide behavior.
D. Estimate the value of previous rewards and punishments, update the value of stimuli in real time, and encode these values in long term memory.
B. Represent the reward value of stimuli, establish predictions of future reward and punishment, and use the predictions to guide behavior.
A subject participates in an fMRI study before and after a snack of strawberry ice cream. Which of the following patterns suggests that the activity reflects the reward value of strawberry ice cream to the participant?
A. Before the snack, the smells of apple pie and strawberry ice cream evoked equally greater activity than the smell of car wax, but afterwards the difference between apple pie and car wax was disproportionately reduced
B. Both before and after the snack, the smells of apple pie and strawberry ice cream evoked equally greater activity than the smell of car wax.
C. Before the snack, the smells of apple pie and strawberry ice cream evoked equally greater activity than the smell of car wax, but afterwards the difference between strawberry ice cream and car wax was disproportionately reduced.
D. Both before and after the snack, the smell of apple pie but not strawberry ice cream evoked greater activity than the smell of car wax.
C. Before the snack, the smells of apple pie and strawberry ice cream evoked equally greater activity than the smell of car wax, but afterwards the difference between strawberry ice cream and car wax was disproportionately reduced.
According to O’Doherty, a proposed role of neural systems that generate a prediction error is:
A. to enhance fear conditioning by facilitating error processing in dopamine circuits.
B. to improve the accuracy of the reward valuation of specific stimuli
C. to improve reward predictions by gating plasticity of neural representations.
D. to improve concentration by enhancing the reward value of reinforcers
C. to improve reward predictions by gating plasticity of neural representations.
According to O’Doherty, neuroimaging studies suggest that activity in which of these areas most consistently resembles the “prediction error”?
A. ventral striatum
B. dorsal thalamus
C. anterior cingulate
D. posterior cingulate
A. ventral striatum
Salience vs. reward prediction: Activation of the ventral striatum has often been reported during reinforcement learning tasks. Which of the following is taken as support for the idea that stimulus salience, not reward prediction error, is indexed by ventral striatal activity during these tasks?
A. Activity in the ventral striatum increased when expected intensity of an upcoming shock increased and decreased when expected intensity of the shock decreased.
B. Activity in the striatum was increased when the subject had to actively respond relative to when the task was passive, even when the same reward was expected by the participant.
C. Ventral striatal activity increases when reward exceeds expectation and decreases when reward falls short of expectation.
D. Ventral striatal activation by unexpected stimuli with no apparent reward value in the task are never observed.
B. Activity in the striatum was increased when the subject had to actively respond relative to when the task was passive, even when the same reward was expected by the participant.
As reviewed by O’Doherty, what would “prediction error” accounts of midbrain dopamine response predict about the response to a monetary reward that was less than a person expected?
A. Dopaminergic neurons would increase their firing rate.
B. Dopaminergic neurons would decrease their firing rate.
C. The change in dopamine firing rate would be the same whether the reward was less or more than expected.
D. There would be no change in dopaminergic neuron firing rate.
B. Dopaminergic neurons would decrease their firing rate.
Activity in which part of the striatum was observed by O’Doherty and his colleagues to respond to violations of expected reward in an instrumental, but not a Pavlovian, reward learning task?
A. ventral striatum
B. dorsal striatum
C. anterior cingulate
D. posterior cingulate
B. dorsal striatum
Fill in the blanks: According to O’Doherty, more dorsal striatal areas may play a larger role in prediction of ____ than of _____.
A. stimulus-stimulus associations, response-outcome associations
B. stimulus-reward associations, response-reward associations
C. response-outcome associations, stimulus-stimulus associations
D. response-reward associations, stimulus-reward associations
D. response-reward associations, stimulus-reward associations
O’Doherty describes one view that the orbitofrontal cortex plays which of the following roles in decision making?
A. It facilitates risk taking behavior by highlighting the reward value of outcomes
B. It maintains flexible representations of reward associations needed to make choices
C. It compares consequences of outcomes to resolve response conflict.
D. It reduces the salience of distracting sensory information.
B. It maintains flexible representations of reward associations needed to make choices
O’Doherty (2004) identified a limitation in the research investigating reward processing and decision making, which is:
A. There are no neuroimaging studies of prediction of aversive, as opposed to rewarding, stimuli.
B. That the individual roles of particular neural structures need to be studied more intensively.
C. Almost no studies have investigated interactions between different neural structures in reward processing.
D. Human subjects are not very responsive to the kinds of reward stimuli administered in reward processing research.
C. Almost no studies have investigated interactions between different neural structures in reward processing.
