gMCP (version 0.3-2)

simConfint: Simultaneous confidence intervals for sequentially rejective multiple test procedures

Description

Calculates simultaneous confidence intervals for sequentially rejective multiple test procedures.

Usage

simConfint(object, pvalues, confint, alternative=c("less", "greater"), estimates, df)

Arguments

object

A graph of class graphMCP.

pvalues

A numeric vector specifying the p-values for the sequentially rejective MTP.

confint

One of the following: A character string "normal", "t" or a function that calculates the confidence intervals. If confintF=="t" the parameter "df" must be specified. If confint is a function it must be of signature ("character", "numeric")

alternative

A character string specifying the alternative hypothesis, must be "greater" or "less".

estimates

Point estimates for the parameters of interest.

df

Degree of freedom as numeric.

Value

• A matrix (or vector) with columns giving lower and upper confidence limits for each parameter. These will be labeled as lower and upper bound.

Details

For details see the given references.

References

Frank Bretz, Willi Maurer, Werner Brannath, Martin Posch: A graphical approach to sequentially rejective multiple test procedures. Statistics in Medicine 2009 vol. 28 issue 4 page 586-604. http://www.meduniwien.ac.at/fwf_adaptive/papers/bretz_2009_22.pdf

See Also

graphMCP

Examples

g <- createBonferroniHolmGraph(3)

# Estimates:
est <- c("H1"=0.9101444, "H2"=0.4485153, "H3"=1.4568271)
# Sample standard deviations:
ssd <- c("H1"=1.2, "H2"=0.8, "H3"=1.7)
pvalues <- 1-pt(est/ssd*sqrt(10),df=9)


simConfint(g, pvalues, confint="t", alternative="greater", df=9, estimates=est)

simConfint(g, pvalues, 
		confint = function(node, alpha) {
			print(paste("Node ",node,"has est=",est[node],", stderr=",ssd[node]/sqrt(10),"and alpha=",alpha,".",sep=""))
			c(est[node]+qt(alpha,df=9)*ssd[node]/sqrt(10), Inf)
		})