We investigated the consequences of using normalization ii in terms of type I and type II error rates (percentage of false positives and false negatives, respectively ...