qanova: Quantile ANOVA

In WRS2: A Collection of Robust Statistical Methods

Quantile ANOVA

Description

One-way ANOVA based on quantiles. Only known method to work well when tied values are likely to occur.

Usage

Qanova(formula, data, q = 0.5, nboot = 600, ...)

Arguments

formula	an object of class formula.
data	an optional data frame for the input data.
q	quantile (or vector of quantiles) to be used.
nboot	number of bootstrap samples
...	currently ignored.

Details

Test global hypothesis that J independent groups have equal quantiles (default: median) using the Harrell-Davis estimator. Performs well when there are tied values.

Value

Qanova an object of class "qanova" containing:

psihat	value of the test statistics
contrasts	contrasts
p.value	p-value
call	function call

References

Wilcox, R. (2012). Introduction to Robust Estimation and Hypothesis Testing (3rd ed.). Elsevier.

See Also

qcomhd, t1way, discANOVA

Examples

## median comparison
set.seed(123)
fitmed <- Qanova(libido ~ dose, viagra, nboot = 200)
fitmed

## 1st, 3rd quartile comparison
set.seed(123)
fitquart <- Qanova(libido ~ dose, viagra, q = c(0.25, 0.75), nboot = 200)
fitquart

WRS2 documentation built on March 19, 2024, 3:08 a.m.