Page 131 - Like me, or else... - Michelle Achterberg

Longitudinal changes of brain and behavior
129
wave 2 including the VIPP-SD group, the control group and the reference group (n=360). Results were False Discovery Rate (FDR) cluster corrected (pFDRcc<0.05), with a primary voxel-wise threshold of p<.005 (uncorrected) (Woo et al., 2014). We computed a full factorial ANOVA with three levels (positive, negative and neutral feedback) to investigate the neural response to the social feedback. Similarly to the whole brain analyses at wave 1 (reported in Achterberg et al. (2018b)), we first explored the general valence effects of social evaluation, by calculating a conjunction (using the “logical AND” strategy, see Nichols et al. (2005)) of positive vs neutral and negative vs neutral social feedback. Next, we calculated the contrasts negative vs positive and positive vs negative to investigate brain regions that were specifically activated for social rejection or social acceptance.
Brain-behavior analyses
In addition to neural responses to social feedback, we also examined whole brain- behavior relations in late childhood (wave 2). Similar to previous brain-behavior analyses in adults (Achterberg et al., 2016b) we conducted a whole brain regression analysis at the moment of receiving negative social feedback (negative vs neutral), with the difference in noise blast duration after negative and neutral feedback as a regressor. In this way, we tested how initial neural responses to feedback were related to subsequent aggression. The difference in noise blast was computed by:
C!%6!%D/ E2 = !%62/#3% -"#$% H<2$/ E2 − !%D/42< -"#$% H<2$/ E2.
To investigate brain-behavior associations across time, we computed the difference over time in noise blasts duration for the contrast negative-neutral and for brain activation in this contrast. A total of 293 participants had behavioral and brain data available at two waves and were included in the analyses regarding brain-behavior associations over time. Difference scores over time for behavior and brain were computed as follows:
C!%6!%D/ H%h23#"4
= (!"#$% H<2$/ -%62/#3% E2 − !"#$% H<2$/ -%D/42< E2)
− (!"#$% H<2$/ -%62/#3% E1 − !"#$% H<2$/ -%D/42< E1)
C!%6!%D/ H42#- = (!%D42< 2,/#3#/A -%62/#3% E2 − !%D42< 2,/#3#/A -%D/42< E2) − (!%D42< 2,/#3#/A -%62/#3% E1 − !%D42< 2,/#3#/A -%D/42< E1)
Behavioral genetic analyses
To examine genetic and environmental influences on brain and behavior, we calculated Pearson within-twin correlations for mono- and dizygotic twin pairs. Similarities among twin pairs can be due to additive genetic variance (A) and
5