How to extract coefficients' standard error from an “aov” model

问题

I did a aov model and I just want to extract the standard errors of each coefficient.

model <- aov(Molecule ~  Comorbidity + Age + BMI + Sex, data = mydata)

I can see the estimated marginal mean with:

allEffects(model)

But I can't extract standard error of each coefficient. I saw on the internet the function se.coef() but it doesn't work. summary(model)$coefficients[, "Std. Error"] does not work either.

I've read effects and other packages, but I don't find what I want. Any idea?

回答1:

Use lm method for summary:

coef(summary.lm(model))

This will give a coefficient table / matrix of 4 columns (mean, standard error, t-value, p-value) for all identifiable coefficients. Then you can extract the 2nd column for standard error.

aov returns object of primary class "aov" but secondary class "lm", hence both summary.aov and summary.lm apply but gives different things. When you simply do summary(model), the former is called as the result of S3 method dispatching.

来源：https://stackoverflow.com/questions/41022222/how-to-extract-coefficients-standard-error-from-an-aov-model