0

I'm running an Anova for unbalanced design with the Anova function of the car package using the type III. I can't find the way of extracting the denominators degrees of freedom in order to describe the results. Here is the example of the code and the output is in the attached picture. Any suggestion?

mod <- lm(PP1meancenter~Partner.type * Drug.1 * Gender1M2F, data=Data_PPIlight, contrasts=list(Partner.type=contr.sum, Drug.1=contr.sum,Gender1M2F=contr.sum))

Anova(mod, type="III")

output image

Machavity
  • 30,841
  • 27
  • 92
  • 100
user310106
  • 1
  • Please make a [reproducible example](https://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example). – Martin Gal Oct 18 '21 at 16:14
  • 1
    Can you please post your results as text (e.g. in a code block) rather than as an image? Images are unsearchable and inaccessible ... – Ben Bolker Oct 18 '21 at 16:42
  • 1
    Your denominator degrees of freedom (ddf) are the "residual" degrees of freedom listed in the last line of the output. In your example, per the "Residuals" line, each F test is based on 600 ddf. Is that what you are asking for? You can test this by calculating the p-values manually based on the F stat and the numerator and denominator df. For "Drug.1" see, e.g., `pf(1.93, 2, 600, lower.tail = FALSE)`. – aosmith Oct 18 '21 at 18:54
  • @aosmith this is exactly what I was looking for, it was just weird to have the same ddf for each of the main effect and interaction and I need confirmation of that. Thank you! – user310106 Oct 19 '21 at 10:13

0 Answers0