 |
 |
Hello everybody, I ran otherwise identical analysis with different cluster-based corrections (i.e., different cluster-forming thresholds) and e.g., also one time with an uncorrected p >= 0.005. I always thought that the zstat-maps would be identical and merely reflect the stats uncorrected for multiple comparisons. However, I noticed minor differences in zstats and varcopes (but not copes) across the different correction methods. Do they depend on the chosen correction method and why? Thanks, Stefan