When we make economic decisions, for example the purchase of a good or a service, our brain has to perform at least three computations. First, it has to assess the goal value of the good: in economic terms, our maximal willingness to pay. Second, it has to assess the decision value of the good: the goal value minus the unavoidable costs. Third, there is a prediction error, which indicates the deviation from one's expectations of reward; the prediction error is positive when something better than expected happens and negative when the opposite occurs. Unfortunately, these three related quantities are intermingled and are often highly correlated, making it challenging to isolate the neural regions performing these computations.Here is a summary figure from the paper:
Hare et al. have attempted to measure goal value, decision value, and prediction error in a single neuroimaging task so that they could dissociate these parameters. They found that ventral striatum activation reflected prediction error and not goal or decision value. However, activity in the medial orbitofrontal cortex and the central orbitofrontal cortex correlated with goal value and decision value, respectively.
Figure - Combined activation maps for goal values (GVs), decision values (DVs), and prediction errors (PEs). Activity correlated with GVs in the mOFC is shown in red, activity correlated with DVs in the cOFC is shown in yellow, and activity correlated with PEs in the ventral striatum is shown in green.