Page 54 - Like me, or else... - Michelle Achterberg

Chapter 2
from the ‘all feedback vs. fixation’ contrast in the replication sample are shown in Figure S3 and Table S1.
Consistency in neural activation in the general contrast
The whole brain analyses resulted in robust activity in the extended face processing network (Scherf et al., 2012), including the FFA and amygdala. Interestingly, these findings were consistent across pilot, test, and replication samples, showing that the task elicits reliable responses in 7-10-year-old children. Even though most activated regions in the pilot sample could be confirmed in the test sample (i.e., bilateral amygdala, bilateral fusiform cortex, and the mPFC), not all regions were confirmed: the PCC and bilateral thalamus were not significantly activated in the test sample. The smaller pilot sample has a reduced chance of detecting a true effect, but a small sample also reduces the likelihood that a significant result reflects a true effect (Button et al., 2013), which shows the need to replicate findings in small samples. This is especially important in developmental neuroimaging studies, since the use of fMRI in children remains a challenging undertaking due to both practical and methodological issues such as more biological noise and motion (Kotsoni et al., 2006; O'Shaughnessy et al., 2008; Thomason, 2009). We therefore repeated the procedure with the test and replication sample and showed that all activated brain regions that were found in the test sample - which was somewhat larger than the pilot sample - could be replicated in the replication sample. Taken together, these findings indicate that the SNAT elicits reliable and consistent neural activation for the general contrast all feedback > fixation.
Brain-behavior correlations
To test for brain-behavior correlations, we correlated the significant meta- analytical brain results with the subsequent behavior. Negative>positive amygdala activation and negative>positive noise blast duration were not significantly correlation in the separate samples (pilot: r=-.02, p=.921; test: r=.28, p=.152; replication: r=-.03, p=.892), nor when tested in a meta-analyses (d=0.14, 95% CI: -0.48-0.76, p=.664). Negative>neutral insula activation and negative>neutral noise blast duration were not significantly correlation in the separate samples (pilot: r=.05, p=.848; test: r=.32, p=.096; replication: r=.04, p=.856), nor when tested in a meta-analyses (d=0.27, 95% CI: -0.35-0.90, p =.394). Lastly, Negative>neutral mPFC/ACCgyrus activation and negative>neutral noise blast duration were not significantly correlation in the separate samples (pilot: r=.17, p=.485; test: r=-.10, p=.600; replication: r=.13, p=.530), nor when tested in a meta-analyses (d=0.14, 95% CI: -0.48-0.76, p =.659). Thus, no significant brain- behavior correlations were found.
52