In short: how to extract the different sum of squares from a mixed-effect model built using "lme" function in R (package nlme)?

In long: I am trying to finish my first experimental study. One reviewer asked me to report effect sizes for my mixed-effects model. I used mixed-effects models because I have a one-way repeated-measure design with one categorical predictor including different number of observations for each category (something the ANOVA does not like at all).

I found in Bakeman (2005) and Olejnik & Algina (2003) that generalized eta squared is particularly suited, and should be easy to compute. For that I need the sum of squares of my model. Unfortunately, as said before, I am new to this kind of statistics. I am completely struggling to understand all the posts I find on the internet, and may have missed my answer somewhere. It may even exist a different estimate of effect sizes for my design.

Nevertheles, for the moment I go on the only way I found. Would anyone know how to extract sum of squares from the model built with "lme"?

Thank you for your time.

  • $\begingroup$ Have you tried using the aov_4() function in the afex package? It should compute generalized eta squared (ges) values. See singmann.org/… for details. $\endgroup$ Commented Aug 15, 2018 at 2:18
  • $\begingroup$ @IsabellaGhement thank you for this suggestion! I did not know this package, which sounds really convenient for multiple uses. I will take a look at it. $\endgroup$
    – Pyxel
    Commented Aug 15, 2018 at 8:53


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.