| Expected utility theory is the standard theory of decision-making under uncertainty. Recently, the theory has faced by many serious challenges from behavioural economics'theories. To explore applicability and effectivity for expected utility theory and choice theories of behavioural economics, respectively, to reduce or bridge the gap between them, and to promote integration of both theories, adopting ecological rationality, and in the perspective of neuroeconomics, the present study suggested "cool"-"hot" contexts frame of decision-making and conducted ERP experiments for probability variable and value variable to find different neural mechanisms of the two variables' processings in different decision-making contexts.The present study includes two parts of experiments. One is empirical study in behavioural level, the other is in ERP level.Firstly, the present study conducted behaviourally empirical study under "cool"-"hot" contexts on four axioms, which were cancellation, transitivity, dominance and invariance. The results found that, except cancellation, the other three axioms were significantly different between "cool"-"hot" decision-making contexts. The findings indicated that axioms for rational choice were obviously dependent on decision-making contexts.Secondly, the present study conducted ERP experiments under "cool"-"hot" contexts to detect the neural mechanisms of probability variable and value variable, respectively. As regards to the experiments of probability processing, under "cool" context, some significant differences of LPC's mean amplitudes between probability reasoning and counting were found in encoding stage, that is, LPC's mean amplitude of probability reasoning was smaller than counting's. This was probably a characteristic of probability encoding. Under "hot" context, no significant difference of LPC's mean amplitudes between probability reasoning and counting were found in encoding stage, but there were differences in decision-making stage, that is, there existed consistent joint-action mechanism between enlarged amplitude of N250 and longer latency of P3. Under "cool"×"hot" contexts, main featur was that the interaction of latency for P3 related to probability processing in encoding stage among experimental condition×hemisphere x electrode s was significant. This suggested that probability encoding was complex in "cool"×"hot" contexts, and that the classification between probability encoding and counting encoding was suppressed by the emotion and feeling processing primed by gain/loss. As regards to the experiments of value processing, under "cool" context, in encoding stage, the main effect of Nogo effects for N2d was significant in size of value; the interaction of mean amplitude for N2d between experimental condition×hemisphere was significant; the interaction of peak amplitude for Go-P3 between experimental condition×electrode was significant. These findings showed that the valence and size of value were binded to encode in "cool" context. In decision-making stage, the interaction of peak latency for P3 beween experimental condition×electrode was significant. In gain/loss feedback stage, the interaction of peak amplitude for P3d beween experimental condition×electrode was significant. Under "hot" context, in encoding stage, the main effect of Nogo effects for N2d was significant in size of value; several effects adjusted by valence of value were manifested by P3 waves, including amplitudes and latencies of P3d and Go-P3. These findings showed that the valence and size of value were independently encoded in "hot" context during P3 time course. In decision-making stage, P3 was insignificant for both amplitude and latency. In gain/loss feedback stage, the main effects of latencies for N2d and P3d in different sizes of value were significant. These results suggested that the gain/loss was processed in heuristic way. Under "cool"×"hot" contexts, in encoding stage, the main effect of Nogo effects for N2d was significant in size of value; the interaction of mean amplitude for N2d between experimental condition×valence was significant; the interaction of mean amplitude for P3d between experimental condition×valence×electrode was significant; the interaction of peak latency for Go-P3 between experimental condition×experimental context was significant. These findings indicated that the valence and size of value were encoded independently in "cool"×"hot" contexts, which started N2d time course and sustained P3d time course. In decision-making stage, P3 was insignificant for both amplitude and latency. In gain/loss feedback stage, the main effects of the latencies for N2d and P3d were significant between small value and large value. Meanwhile, the main effect of latency for P3d was significant in different decision-making contexts, namely, latency of P3d was shorter in "hof" context than in "cool" context. Therefore, these results suggested that value was processed in heuristic way under "hof" context, while it was processed in algorithmic way under "cool" context.In conclusion, the axioms of decision-making were obviously dependent on decision-making contexts. The neural mechanisms of probability variable and value variable in different processing stages were different under "cool" context, "hot" context, and "cool"×"hot" contexts. Meanwhile, there existed common neural mechanism of value variable in encoding stage under three different decision-making contexts. In the present study, the characteristics of ERP in "cool" context, "hot" context and "cool"×"hot" contexts were the neural mechanisms of the rational, bounded and ecological rational choices in probability/value variables'processing, respectively. Moreover, the characteristics of ERP in "cool"×"hot" contexts were the neural mechanisms of integration between the rationality and bounded rationality in probability/value variables'processing. The consistency of value encoding was the common neural mechanism of the rational, bounded and ecological rational choices in value variable's processing. All of the findings in the present study suggested that "cool" context and "hot" context had different effects on decision-making under uncertainty. Therefore, the frame of "cool"-"hot" decision-making contexts is valid for explaining the contextual suitablity of expected utility theory and behavioural economics'theories.
|
PDF Full Text Request