Information Technology Reference
In-Depth Information
Fig. 6.
Accuracies of significance classifier and p-threshold for different number of used p values
for randomly generated distributions
Fig. 7.
Root mean squared error for replication prediction (normal distributions with mean 20,
stdev 2 vs. mean 22, stdev 2)
estimate the needed sample size (e.g., [12]). We use the implementation of R Project
(
power.t.test
) with the estimated mean difference of the corresponding number of