Check ANOVA Assumptions

Produces a 2x2 panel of diagnostic plots for an ANOVA or linear model: residuals vs fitted, Q-Q plot, Shapiro-Wilk test result, and Levene's test result (if a grouping variable is provided).

Usage

check_assumptions(model, data = NULL, group = NULL)

Arguments

model

A fitted model object (from aov or lm).

data

The data frame used to fit the model (required for Levene's test).

group

A character string naming the grouping variable in data for Levene's test. If NULL, Levene's test is skipped.

Value

Invisibly returns a list with the Shapiro-Wilk p-value and (optionally) the Levene's test p-value.

Examples

data(crd_data)
mod <- aov(Yield ~ Variety, data = crd_data)
check_assumptions(mod, data = crd_data, group = "Variety")