In everyday situations, individuals are required to make choices between actions with respect to current behavioural goals, and in the context of information relating to the existing contingencies between particular actions and relevant outcomes. An important project for biological psychology is to specify and describe the neural bases of the psychological processes that allow such choices to be made.
Traditionally, much of the existing research into human-decision making has centred around findings that subjects often make sub-optimal choices in comparison to those predicted by normative theories arising out of economic and statistical research, especially in the context of decisions about probabilistic events. For example, seminal work by Tversky and Kahneman demonstrated that subjects’ judgements about the probable truth of hypotheses can be inaccurate, relative to Bayesian estimates (see BAYESIAN STATISTICS) because of failure to take full account of the relative frequency of events (‘base rates’) as well a tendency to focus on information that is either more readily available to the subject, or more consistent with pre-existing prototypes used in the decision-making process (the so-called availability and representiveness heuristics). Other research has shown that the factors which determine the precise strategies used by subjects to arrive at a decision include the number of choices and their attributes, as well as the type and format of material involved in the choice (see Slovic et al., 1988 for review).
The major impetus for research on the neurobiology of decision-making stems from the observation that certain patient popula tions with neurological or neuropsychiatric conditions, such as those with focal damage to the PREFRONTAL CORTEX, DRUG ABUSE problems, OBSESSIVE COMPULSIVE DISORDER and sociopathic behaviours, often exhibit marked changes in the quality of decision-making in their everyday lives. Of particular interest are those patients who have sustained damage to ventromedial sectors of the prefrontal cortex (VM-PFC) (see MEDIAL PREFRONTAL CORTEX). A classic example is patient EVR (see Damasio, 1994) whose surgery for a tumour left him with a bilateral LESION of this cortical area. Before his operation, EVR was an effective, successful individual in both professional and personal terms. However, after his operation, his behaviour was characterized by a pattern of personally disastrous decisions whose consequences included the failure of his first and subsequent marriages, a series of ill-judged investments and profoundly disrupted interpersonal behaviour. However, his neuropsychological profile showed little sign of gross cognitive decline (as evidenced by a performance IQ of 135 points), and essentially normal performance on instruments known to be sensitive to other aspects of prefrontal dysfunction. Thus, the scientific challenge posed by patients like EVR is to account for their specific inability to make effective choices in social and real-life settings, in the absence of neuropsychological deficits in other functions (such as short-term retention of information in WORKING MEMORY systems, the ability to control mental and motor sets, and basic linguistic competence) widely thought to underpin the complex cognitive operations required to arrive at adaptive, successful decisions.
One possibility is that the real-life decisions which VM-PFC patients find so difficult are characterized by a need to reason about uncertain or poorly understood contingencies between actions and outcomes. In these cases, the available propositional knowledge about the alternative choices and their relative attributes, over which subjects might have been able to deploy a number of cognitive strategies (see above), is insufficient for arriving at a decision and needs to be supplemented by additional, perhaps more implicit, information about the consequences of choices in similar situations encountered in the past. Evidence that VM-PFC patients are not able to use such information to assist their decision-making is seen in the finding that, in a simulated gambling paradigm involving uncertain action-outcome contingencies, VM-PFC patients make choices that earn high REWARD but still higher penalties (resulting in eventual losses) at the expense of choices that earn smaller rewards but still smaller penalties (resulting in eventual gains). Additionally, these risky choices are associated with reduced autonomic responses, suggesting that the basis of this behaviour might be a failure to mark disadvantageous choices with an appropriate somatic marker (see SOMATIC MARKER HYPOTHESIS) that could guide future behaviour.
It is important to emphasize that, at the time of writing, very little is known about the neural basis of decision-making. For example, it is certain that the ventromedial prefrontal cortex is only one station in the wider circuitry mediating the various cognitive processes that underpin decision-making cognition. This circuitry will probably incorporate additional cortical areas (for example, the SOMATOSENSORY CORTEX) as well as those LIMBIC SYSTEM structures (such as the ventral STRIATUM and AMYGDALA) that influence the selection of actions via the provision of reward or emotion-related information. Moreover, precise modulation of this circuitry by the ascending MONOAMINE projections is also likely to be extremely important in the efficiency and quality of decision-making. For instance, extensive clinical evidence indicates that impaired decision-making culminating in violent, impulsive behaviour is associated with reduced levels of SEROTONIN metabolites in the CEREBROSPINAL FLUID of sociopathic individuals. Thus, a full account of the NEUROPSYCHOLOGY of decision-making will need to integrate neuro-anatomic and neurochemical information, as well as psychological information about the way subjects reach their decisions.
Slovic P., Lichtenstein S. & Fischhoff B. (1988). In Steven’s Handbook of Experimental Psychology, 2nd edn, ed. Atkinson R.C., Hernnstein R.J., Lindzey G. & Luce R.D., pp. 673–738, Interscience-Wiley: New York.
ROBERT D.ROGERS
This is the complete article, containing 889 words
(approx. 3 pages at 300 words per page).
