Compute the empirical detection rate for Type I errors and Power

Description

Computes the detection rate for determining empirical Type I error and power rates using information from p-values.

Usage

1
EDR(p, alpha = 0.05)

Arguments

p

a numeric vector or matrix/data.frame of p-values from the desired statistical estimator. If a matrix, each statistic must be organized by column, where the number of rows is equal to the number of replications

alpha

the nominal detection rate to be studied (typical values are .10, .05, and .01). Default is .05

See Also

ECR

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
rates <- numeric(100)
for(i in 1:100){
   dat <- rnorm(100)
   rates[i] <- t.test(dat)$p.value
}

EDR(rates)
EDR(rates, alpha = .01)

# multiple rates at once
rates <- cbind(runif(1000), runif(1000))
EDR(rates)

Questions? Problems? Suggestions? or email at ian@mutexlabs.com.

All documentation is copyright its authors; we didn't write any of that.