Function to calculate the testwise type I error rate threshold
corresponding to a give familywise threshold.
Usage
cthreshold(alpha, nbtest)
Arguments
alpha
The familywise type I error threshold.
nbtest
The number of tests performed.
Value
The threshold that have to be used for individual tests.
encoding
utf8
Details
Type I error rate inflation occurs when a single hypothesis is tested
indirectly using inferences about two or more (i.e. a family
of) sub-hypotheses. In such situation, the probability of type I error
(i.e. the probability of incorrectly rejecting the null
hypothesis) of the single, familywise, hypothesis is higher than the
lowest, testwise, probabilities. As a consequence, the rejection of
null hypothesis for one or more individual tests does not warrant that
the correct decision (whether to reject the the null hypothesis on a
familywise basis) was taken properly. This function allows to obtain
correct, familywise, alpha thresholds in the context of multiple
testing. It is base on the Sidak inegality.
References
Sidak, Z. 1967. Rectangular Confidence Regions for Means of
Multivariate Normal Distributions J. Am. Stat. Assoc. 62: 626-633
Wright, P. S. 1992. Adjusted p-values for simultaneous inference.
Biometrics 48: 1005-1013
See Also
Legendre, P. and Legendre, L. 1998. Numerical Ecology. Elsevier
Science B.V., Amsterdam, The Neatherlands. p. 18