Condition-based non-parametric analysis.
For the neuronal analyses we focused on mixed-risk trials with and without ambiguity in comparison to high- and low-risk trials. During mixed-risk trials as compared to high- and low-risk trials in the pure-risk block significant activation was observed in the left parahippocampal gyrus (Table 1, Figure 2). In the ambiguity block activations were present in the right inferior frontal gyrus IFG (BA 47) for mixed-risk trials with high ambiguity compared to high- and low-risk trials with high ambiguity (Table 2, Figure 3). For the same contrast with low ambiguity, activity in the left supplementary motor area (BA 6) and in the left posterior insula was significant (Table 2, Figure 4).
The contrast high versus low ambiguity showed a significant BOLD effect in the mixed-risk condition. High compared to low ambiguity was correlated with activations in the right inferior parietal cortex and the left DLPFC in BA 9 (Table 2, Figure 5). As noted above, mixed-risk trials under high ambiguity showed the highest amount of low-risk decisions on the behavioral level.