Skip to main content
. 2018 Jul 9;19(Suppl 7):199. doi: 10.1186/s12859-018-2173-7

Fig. 3.

Fig. 3

Simulation Study: Overall Type I and Type II error. Under a B-H control of the FDR: a Boxplots showing the distribution of the percentage of responses (within a dataset) falsely identified as having non-constant mean response and b Boxplots showing the distribution of the percentage of responses (within a dataset) falsely identified as having a constant mean response over all treatment groups