compensateAlpha: Significance level compensation for multiple comparisons.

In tsieger/tsiMisc: Miscelaneous utility functions, mostly related to exploratory statistical analysis

Significance level compensation for multiple comparisons.

Description

compensateAlpha compensates statistical significance level (alpha) for multiple comparisons/tests.

Usage

compensateAlpha(alpha, 
    n, method = c("bonferroni", 
        "sidak"), r = 0)

Arguments

alpha	significance level to be compensated
n	comparison count
method	compensation method, 'bonferroni' is more conservative, but does not assume independence of the tests, 'sidak' is more liberal and assumes independence of the tests
r	optional mean "correlation between the tests made", the extreme value of 0 leads to full compensation, the extreme value of 1 leads to no compensation (as all the tests are considered equal). See the SISA help for explanation.

Value

Compensated significance level. It equals alpha/n for the 'bonferroni' method, and 1-(1-alpha)^(1/n)) for the 'sidak' method.

Author(s)

Tomas Sieger

References

SISA, http://www.quantitativeskills.com/sisa/calculations/bonhlp.htm

See Also

compensatePValue, p.adjust

Examples

compensateAlpha(.05,2,method='sidak')

tsieger/tsiMisc documentation built on Dec. 14, 2024, 3:23 p.m.