Page 53 - Like me, or else... - Michelle Achterberg

Social evaluation in childhood
51
Automated Anatomical Labeling (AAL) atlas (Tzourio-Mazoyer et al., 2002). The overlap between functional activation and anatomical regions were then used as regions of interest. All regions from the whole brain contrast were investigated. To mask the medial orbitofrontal cortex (mOFC) we combined the medial OFC left and right. The subcallosal cortex was masked with the subcallosal mask from the Harvard/Oxford atlas (Desikan et al., 2006).
One sample t-tests (one-sided) were used to test whether the activation was significantly different from 0. We specifically chose one sided t-tests (α=0.1), because replication is tested in the same direction as in the hypothesis- generating sample. Alpha level was Bonferroni corrected depending on the number of extracted ROIs (i.e., 0.1 divided by the number of ROIs).
Examination of pilot results in the test sample
The contrast ‘all feedback vs. fixation’ in the pilot sample resulted in activation with local maxima in the bilateral lateral occipital lobes, the bilateral fusiform cortex, the bilateral amygdala, the bilateral thalamus, the medial prefrontal cortex (PFC), and the posterior cingulate cortex (PCC), see Figure S2a and Table S1. From this whole brain contrast, we selected 8 ROIs: the right and left amygdala, the right and left fusiform cortex, the right and left thalamus, the mPFC and the PCC (Figure S2b). These ROIs were used to extract PE values from the test sample. Bonferroni corrected alpha was set at α=0.013 (0.1/8 ROIs). As Figure S2c shows, activation of the left and right amygdala, the left and right fusiform cortex, and the mPFC was significantly different from 0 in the test sample, and thus the pilot results were confirmed in the test sample (all p’s<.013, see Table S2). The test sample showed no significant activation in the left and right thalamus, nor in the PCC.
Examination of test results in the replication sample
The contrast ‘all feedback vs. fixation’ in the test sample resulted in activation with local maxima in the bilateral occipital lobes, the bilateral fusiform cortex, the bilateral amygdala, the cerebellum, the mPFC, the bilateral inferior orbitofrontal cortex (OFC), the medial OFC and the subcallosal cortex, see Figure S2d and Table S1. We selected five ROIs concerning anatomical regions that were also found and confirmed in step 1: the left and right amygdala, the left and right fusiform cortex and the mPFC. Activation in four addition regions were observed and masked as ROI: the subcallosal cortex, the medial OFC and the left and right inferior OFC (Figure S2e). In total 9 ROIs were used to extract PE values from the replication sample, therefore Bonferonni corrected alpha was set at α=0.011 (0.1/9 ROIs). Activation in all regions was statistically significantly different from 0 in the replication sample, indicating that the test results were replicated in the replication sample (all p’s<.011, see Figure S2f and Table S2). Whole brain results
2