Tukey.HSD {NormData}

Conducts Tukey's Honest Significance Difference test

Description

This function conducts Tukey's Honest Significance Difference (HSD; Tukey, 1949) test that allows for making post hoc comparisons of the group means. Tukey's HSD can only be conducted when the mean structure of the Stage 1 model only contains qualitative independent variables (i.e., when the fitted regression model is essentially an ANOVA).

Usage

Tukey.HSD(Stage.1.Model, ...)

Arguments

Stage.1.Model

A fitted stage one model that only contains qualitative variables.

...

Arguments to be passed to the plot function of the Tukey HSD procedure.

Value

No return value, called for side effects.

Author(s)

Wim Van der Elst

References

Tukey, J. (1949). Comparing individual means in the Analysis of Variance. Biometrics, 5, 99-114.

Van der Elst, W. (2024). Regression-based normative data for psychological assessment: A hands-on approach using R. Springer Nature.

See Also

plot.Tukey.HSD

Examples

data(Personality)
Model.Openness <- Stage.1(Dataset = Personality, Model = Openness ~ LE) 
# conduct post hoc comparisons for the levels of education
Tukey.Openness <- Tukey.HSD(Model.Openness)
summary(Tukey.Openness)
plot(Tukey.Openness)

# conduct post hoc comparisons for the levels of education by education combinations
data(Substitution)
Model.Substitution <- Stage.1(Dataset = Substitution, Model = LDST ~ LE*Gender)
Tukey.Substitution <- Tukey.HSD(Model.Substitution)
summary(Tukey.Substitution)
plot(Tukey.Substitution)