Hi everyone:
I am running NBS ( Article Network-based statistic: Identifying differences in brain networks
) to compare 21x21 FC matrices between two groups. My concern is that component and differences in networks are dependent upon previous t-value thresholding. I tried several threshold values as the user guide suggests, but I am confused about what threshold should I pick to report my results. Is there any systematic approach to pick this threshold value? What criteria should I consider when choosing suprathreshold connections?