In a follow-up analysis, to investigate the consistency of SPE results between the sessions, we identified the peak voxels for the SPE signal in session 2 only, and then tested for a significant SPE representation in session 1 in a reduced spherical search volume (radius: 10mm, p < 0.05, family-wise error correction for search volume). This procedure ensures that the centers for the search volumes are selected in a way that is independent of the data in session 1. We found significant effects of SPE in session 1 bilaterally in latPFC and in the right pIPS/angular gyrus (Figure 4) confirming that these areas correlate with an SPE even in the absence of any reward information (see Table 2). To test for overlapping voxels with SPE representations in both sessions we employed a conjunction analysis (Nichols et al., 2005) and found evidence that voxels in these regions were activated in both sessions at p < 0.001 uncorrected.
