As mentioned previously, important edge connectivity-bingeing associations might be lost in the network-level analysis. The consistency analysis, despite being exploratory, can be valuable as it highlights important edges that demonstrate significant (puncorrected<.0001) and consistent associations with bingeing (Supplementary Figure 2). Indeed, all edges identified in the consistency analysis survived the FDR correction for all partial correlation tests run across the whole sample (i.e., correlating change in each of the edge connectivity with bingeing, controlling for the aforementioned covariates).
