- main document
Download (6MB) | Lizenz: Print on Demand
The current thesis had one central aim: contributing to the understanding of age differences in decision making under uncertainty. Three studies were dedicated to investigate age differences in uncertain decisions in individual contexts. A fourth study aimed at developing a framework for the study of age differences in uncertain decisions with a social dimension. In the first study, we have shown the existence of age differences in ambiguous decisions, but not in risky decisions. More specifically, we can show that young adults are more ambiguity averse than older adults, and that young adults are more ambiguity than risk averse, whereas no such effect appears in older adults. In the second study, I could reproduce the results of the first study. Further, I provide the first direct experimental evidence that whereas no age differences appear for risky decisions with a priori probabilities, there are age differences when it comes to risky decisions with statistical probabilities, with older adults being more risk averse than young adults. I also have shown that feedback on decision outcomes increases risk- and ambiguity seeking in both young and older adults, even if this feedback cannot help predicting future decision outcomes. Finally, I also have shown that the level of education may counteract age effects on decision making. The aim of the third study was to investigate the link between age differences in decision making under uncertainty and age differences in its underlying brain mechanisms. Unlike in the two precedent studies, we do observe a higher level of risk aversion in older adults than in young adults, and no age difference in ambiguity aversion. However we still reproduce the finding that whereas young adults are more ambiguity- than risk-averse, no difference appears in older adults. Within the brain, we observe clear age differences in brain regions responsible for different mechanisms involved in uncertainty processing. We also observe an age group x uncertainty condition interaction effect in connectivity between two brain regions. In the fourth study, we have developed a framework for studying decision making under uncertainty with the possibility to delegate the decision to an expert, or to seek an advice from that expert. We developed a computerized task to investigate these decisions. In this task, it appears that decision makers valuate their own information much higher than the experts’ information. The level of uncertainty of the task has an effect on the decision whether to consult the expert or whether to make a decision that does not involve the expert. We also observe gender differences on the task, with male participants choosing more often to seek an advice or to delegate the decision than female participants. Choice behavior on this task also correlates with the risk preferences of the decision maker. Based on the literature on aging and delegation or advice, we make predictions about the results we might expect when administrating the task to an older population. In a greying society, understanding the influences of age on decision making becomes increasingly important. In fact, with increasing life expectancy, it is likely that retirement age will increase, and that people need to work and make decisions until a higher age. Many everyday life decisions involve some amount of uncertainty, and understanding how age affects these decisions can contribute to the conservation and improvement of older adults’ of decision making abilities. The aim of this thesis was to contribute to the understanding of age effects on uncertain decisions. Contrary to the common stereotypes of older adults being generally more uncertainty averse than young adults, I could show throughout this thesis that various factors influence whether young and older adults differ with respect to decision making under uncertainty. It appears for example that the type of uncertainty involved in the decision has an impact on whether age differences appear, and that whereas young adults behave differently under risk and under ambiguity, older adults do not make a difference and have similar risk and ambiguity preferences. Hence, in everyday life decisions where one might expect people to behave differently if a problem is framed as risky or as ambiguous, our results point in the direction of older adults not differentiating between risk and ambiguity and behaving similarly in both frames. Furthermore, in two of the three experiments comparing young to older adults, we have shown that older adults are less ambiguity averse than young adults. These findings can for example be beneficial in work settings where a low level of ambiguity aversion is required, or where it is preferable if decision makers behave similarly under risk and ambiguity. In risky decisions, the type of risk in the decision plays a role too. If probabilities of winning or losing in a decision are explicitly shown to older adults, it is likely that no age differences in decision making will appear. If however older adults need to learn from experience about the riskiness of a choice, one might expect older adults to be more risk averse than young adults would be. We also have shown that the learning requirements are not the underlying factor of this effect. Rather it seems that young adults are less “careful” and rather use strategies that allow making many decisions in a short amount of time. Hence one possibility to minimize age differences in risky decisions is to make the probabilities as explicit as possible. Another possibility might be (in conditions where a higher risk aversion is disadvantageous) to act on older adults’ decision strategies, and teach them to use younger adults’ strategies. Two additional factors that have been shown in this thesis to influence uncertainty preferences are feedback and education. Feedback per se acts by reducing subjective uncertainty of the decision, even if it does not influence the objective uncertainty. Therefore, providing feedback on decisions decreases the uncertainty aversion of the decision maker, and led in our experiment to a higher amount of bets. Hence, if a reduced uncertainty aversion of the decision makers is beneficial for the decision outcome, feedback can be used both in young and in older adults to reduce their uncertainty aversion. Education also has, at least in self-reported risk preferences, an effect on uncertainty preferences; the more years of education, the less risk averse the decision maker. Further, the level of education of the decision maker counteracts the age effects in self-reported risk preference; whereas age increases risk aversion, education decreases it. Besides describing behavioral age differences, we also looked at the processes underlying these differences. When understanding the underlying processes, influencing behavior gets much easier. We therefore investigated brain mechanisms underlying age differences in decision making under uncertainty. We observed that brain regions involved in various processes important in decision making under uncertainty, such as risk and value processing, and cognitive and emotion processing, show age differences. This suggests that the processes underlying age differences in decision making change with age. It is at this point however not trivial to draw strict conclusions on how the age differences in brain functioning influence behavior. In fact, this study was the first at investigating age differences in brain mechanisms not only under risk but also under ambiguity, and further research will be needed to clarify how exactly the effects observed in the brain relate to behavior. What we can tell at this point is that various processes are underlying the age differences in behavior, and that not only age differences in overall brain activation appear, but also that the way the brain regions connect to each other changes. One of the potential downsides of the studies on age differences performed within this thesis is the impossibility to rule out cohort effects. That is, to account for cohort effects and the most accurately describe development of the observed effects, one would need to observe the same subjects over a prolonged time (in the best case over the entire lifespan). However, besides the non-feasibility of such an effort within a Ph.D. project, our societies are rapidly getting older, and there is an immediate need to describe and understand age differences in decision making behavior. How these differences develop certainly is also an interesting question, but at the moment, describing the age differences and understanding their underlying processes is a more urgent question. The aim of the last study of this thesis was to develop a framework for the study of decision making under uncertainty with the possibility to seek an advice or to delegate a decision. In fact, research results point in the direction of age differences in both, delegation and advice. We developed a novel task which we validated with a small sample of young participants, and reviewed the literature relating to age differences that might be expected on the task. The next step will be to administrate the task to a group of older participants and to increase the size of the young sample to test whether our hypotheses about age differences in decision making with delegation and advice hold. As the task is general enough to allow for various sophistications, we expect it to be a first building block in a larger series of studies.
In summary, throughout this thesis we observed age differences in decision making under uncertainty in a number of different settings. As the age effects not always appear in the way intuition predicts, one should always in a first step think of the type of uncertainty and the mechanisms involved in the decision, and not just start from a deficit perspective in decision making when dealing with older adults. Many questions about the nature of the age differences in decision making have been answered, and future work will contribute to a full understanding of the processes that lead to age differences in decision making.
|Supervisor:||Schwieren, Prof. Dr. Christiane|
|Date of thesis defense:||8 January 2013|
|Date Deposited:||01 Feb 2013 07:27|
|Faculties / Institutes:||The Faculty of Economics and Social Studies > Alfred-Weber-Institut for Economics|
600 Technology (Applied sciences)
|Uncontrolled Keywords:||neuroeconomics, behavioural economics, ageing, decision making, uncertainty|