t1way function

A heteroscedastic one-way ANOVA for trimmed means.

The t1way function computes a one-way ANOVA on trimmed means. Homoscedasticity assumption not required. It uses a generalization of Welch's method. Corresponding post hoc tests can be performed using lincon.

t1way(formula, data, tr = 0.2, alpha = 0.05, nboot = 100, ...)
lincon(formula, data, tr = 0.2, alpha = 0.05, method = "hochberg", ...)

Arguments

• formula: an object of class formula.
• data: an optional data frame for the input data.
• tr: trim level for the mean.
• alpha: alpha level for CI computation.
• nboot: number of bootstrap samples for effect size CI computation.
• method: method to correct the p-value (see p.adjust)
• ...: currently ignored.

Details

In the post hoc computations, confidence intervals and p-values are adjusted to control FWE. The default for the p-values is to use Hochberg's 1988 sharper Bonferroni procedure.

Returns

t1way returns an object of class "t1way" containing:

• test: value of the test statistic (F-statistic)

• df1: degrees of freedom

• df2: degrees of freedom

• p.value: p-value

• effsize: explanatory measure of effect size

• effsize_ci: boostrap effect size CI

• call: function call

lincon returns an object of class "mcp1" containing:

• comp: inference for all pairwise comparisons

• fnames: names of the factor levels

linconv2 returns an object of class "linconv2" containing:

References

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

See Also

med1way,t1waybt

Examples

set.seed(123)
t1way(libido ~ dose, data = viagra)

## post hoc tests
lincon(libido ~ dose, data = viagra)