adaperm_DR: Permutation based decision rule for adaptive designs

In floatofmath/adaperm: Permutation tests for adaptive designs

Description

User friendly wrapper to adaptive_permdr

Usage

adaperm_DR(x, g = NULL, n1, n, m1 = n1, m = n, test_statistic,
  alpha = 0.025, cer_type = "non-randomized",
  atest_type = "non-randomized", permutations = 10000, stratified = TRUE)

Arguments

x	Observations
g	Treatment assignments (if NULL) a one-sample test will be performed
n1	First stage control-group sample size
n	Pre-planned overall sample size
m1	First stage treatment-group sample size (will be ignored if g is NULL)
m	Pre-planned overall treatment-group sample size (will be ignored if g is NULL)
test_statistic	Test statistic @param alpha0 Early rejection boundary for group sequential trial
alpha	Significance level
cer_type	what type of conditional error rate function should be used (see permutation_cer) for detailcer_type
atest_type	if 'CER' compute only conditional error rate, else type of adaptive test should be performed (see perm_test for details)
permutations	Number of permutations to use
stratified	should permutation be stratified by stage

Details

Group sequentialism assumes that the trial does not stop at the first stage - otherwise second stage data would not be available.

Value

Decision TRUE if null hypothesis is rejected

Author(s)

Florian Klinglmueller

floatofmath/adaperm documentation built on May 16, 2019, 1:18 p.m.