asd (version 2.2)

combn.test: Combination Tests for ASD

Description

Implements weighted inverse normal and Fisher combination tests for combining p-values for adaptive seamless designs.

Usage

combn.test(stage1, stage2, weight = 0.5, method = "invnorm")

Arguments

stage1

Output from function dunnett.test from stage 1 of an ASD

stage2

Output from function dunnett.test from stage 2 of an ASD

weight

Weight indicating how p-values from stages 1 and 2 are combined; default weight is 0.5 indicating equal weighting between stages (0<weight

method

Select combination test method; available options are “invnorm” or “fisher”, with default “invnorm”

Value

method

Selected method of combining p-values

zscores

Z-scores for each hypothesis

hyp.comb

A list of matrices indicating the structure of the intersection hypotheses

weights

Weights used for each stage

Details

The basic ideas of the combination test approach were proposed by Bauer and Kieser (1999) and make use of a combination function (Bauer and Kohne, 1994) to combine stagewise p-values to allow for interim adaptations and the application of the closed test principle (Marcus et al., 1976) to control the overall test size across multiple hypotheses.

References

Bauer P, Kieser M. Combining different phases in the development of medical treatments within a single trial. Statistics in Medicine 1999;18:1833-1848.

Bauer P, Kohne K. Evaluation of experiments with adaptive interim analyses. Biometrics 1994;50:1029-1041.

Marcus R, Peritz E, Gabriel KR. On closed testing procedures with special reference to ordered analysis of variance. Biometrika 1976;63:655-660.

Lehmacher W, Wassmer G. Adaptive sample size calculations in group sequential trials. Biometrics 1999;55:1286-1290.

See Also

treatsel.sim, dunnett.test, hyp.test, select.rule, simeans.binormal

Examples

stage1 <- dunnett.test(c(0.75,1.5,2.25))
stage2 <- dunnett.test(c(0.15,1.75,2.15))
combn.test(stage1,stage2,weight=0.5,method="invnorm")