Probability estimation in decision making

Every day life requires numerous and fast decisions. Wrong decisions may have disastrous consequences, in personal relationships, in financial investments or in the working domain. Patients with neurodegenerative diseases, acquired brain lesions or psychiatric conditions frequently have deficits in making advantageous choices, though their intellectual functions are fully preserved. In neuropsychological research decision making has been studied with simulated gambling tasks. In these gambling tasks subjects are asked to maximise their wins and receive constant feedback about their gains and losses. Decision making with feedback and reward has been investigated intensively and emotional processing seems to play a crucial role. A further component of decision making has not been investigated so far. Uncertain decisions are based on the estimation of probability of a certain outcome. The evaluation of probability requires not only reasoning, but also cognitive estimation and approximation. We will investigate how subjects estimate the probability of an event and how their estimates are modified by feedback and reward. The project aims to define the cognitive architecture of this specific type of decision making, as well as the neuro-anatomical implementation. Importantly, the experimental design will separate estimation of probability, processing of feedback and learning. The proposed research is structured in three streams of research: In part one behavioural measures in computer based tasks (accuracy, Rts) will be used to assess age-related changes in healthy young and old subjects. Recent research suggests that in subgroups of older persons reasoning and decision-making is impaired due to disproportionate aging of the prefrontal region of the brain. Part two assesses the performance of patients with brain lesions and allows conclusions about the supporting neuro-anatomical structures. Of particular interest are patients with parietal lesions (numerical and approximation deficits), and patients with reduced semantic knowledge. In part three brain imaging (functional magnetic resonance) is used to analyze the specific brain activation patterns in the estimation of probability.

Despite their cognitive impairment, patients with acquired brain dysfunction have to face important and difficult decisions (e.g. medical treatment, retirement). Few studies so far have specifically investigated decision making in neurological patient groups. Moreover, little is known how healthy aging affects decision making abilities. In the present research project we could show that aging has an effect on decision processes under uncertainty. Healthy old adults often have difficulties to establish an advantageous decision strategy in ambiguous and uncertain situations, while they have no difficulty to make advantageous decisions when risks are explicitly defined. Difficulties in decision making are more pronounced in neurological patient groups with cognitive deficits. Patients with Alzheimer`s disease and Parkinson dementia have difficulties in all types of decision making. They shift frequently between possible options, have difficulties to assess probabilities and to recognize the utility of an option. Other patient groups, including patients affected by temporal lobe epilepsy and patients after traumatic brain injury, show difficulties in decision making as well. Results of the present study suggest that patients have pronounced difficulties in decision making when learning by feedback and emotional processing is required. The present project offers insights into the neuronal mechanisms involved in decision making and into the vulnerability of decision making by acquired brain dysfunction. Since patients very often have to face important decisions, assessment of decision abilities should be included into neuropsychological assessment. Moreover, specific interventions should be included into rehabilitation programmes. This is of special importance for patients who are still integrated into professional life or wish to take up their previous professional activities. For all patient groups reducing uncertainty in decision situations and providing complete information about risks and possible consequences of decisions significantly facilitates advantageous decision making, as shown in the present research project. In a further part of the present project we assessed by functional magnetic resonance imaging how proportions or fractions are processed in the brain. The understanding of proportions and risk is crucial in making advantageous decisions. Our imaging data suggest that not only the single parts of a fraction, but the fraction`s numerical value is accessed automatically and activates intraparietal structures. Results indicate that, within the intraparietal sulcus, a fraction might be represented by its numerical value as a whole, rather than by the numerical values of its numerator and denominator.