14.2 4 Steps to conduct an ANOVA
Here are the 4 steps you should follow to conduct a standard ANOVA in R:
- Create an ANOVA object using the
aov()
function. In theaov()
function, specify the independent and dependent variable(s) with a formula with the formaty ~ x1 + x2
where y is the dependent variable, and x1, x2 … are one (more more) factor independent variables.
# Step 1: Create an aov object
mod.aov <- aov(formula = y ~ x1 + x2 + ...,
data = data)
- Create a summary ANOVA table by applying the
summary()
function to the ANOVA object you created in Step 1.
# Step 2: Look at a summary of the aov object
summary(mod.aov)
- If necessary, calculate post-hoc tests by applying a post-hoc testing function like
TukeyHSD()
to the ANOVA object you created in Step 1.
# Step 3: Calculate post-hoc tests
TukeyHSD(mod.aov)
- If necessary, interpret the nature of the group differences by creating a linear regression object using
lm()
using the same arguments you used in theaov()
function in Step 1.
# Step 4: Look at coefficients
mod.lm <- lm(formula = y ~ x1 + x2 + ...,
data = data)
summary(mod.lm)