mutoss (version 0.1-9)

multiple.down: Benjamini-Krieger-Yekutieli (2006) Multi-Stage Step-Down

Description

A p-value procedure which controls the FDR for independent test statistics.

Usage

multiple.down(pValues, alpha)

Arguments

pValues

A numeric vector of p-values

alpha

The FDR error rate to control

Value

• A list containing:
• rejected
  A logical vector indicating which hypotheses are rejected
• criticalValues
  A numeric vector containing critical values used in the step-up-down test.
• adjPValues
  A numeric vector containing adjusted p-values.
• pi0
  An estimate of the proportion of true null hypotheses among all hypotheses (pi0=m0/m).
• errorControl
  A Mutoss S4 class of type errorControl, containing the type of error controlled by the function and the level alpha.

Details

A non-linear step-down p-value procedure which control the FDR for independent test statistics and enjoys more power then other non-adaptive procedure such as the linear step-up (BH). For the case of non-independent test statistics, non-adaptive procedures such as the linear step-up (BH) or the all-purpose conservative Benjamini-Yekutieli (2001) are recommended.

Examples

pvals<- runif(100)^2
alpha<- 0.2
result<- multiple.down(pvals, alpha)
result 
plot(result[['criticalValues']]~pvals)
plot(result[['adjPValues']]~pvals)
abline(v=alpha)