1

I done a GLMM with the function glmer from lme4, the variable paternity is binomial (yes or no)(model <- glmer(paternity~factor1, factor 2 ..., family=binomial).

My model run correctly, but after I want plot the effect. So I used the package effects. When I plot my effect, the Y-axis is numeric but I do not understand what it mean because this axis is binomial (paternity yes or no) ...

How I must interpreted this ?

kjetil b halvorsen
  • 77,844
anais
  • 11
  • The results are probably on the log-odds scale. (They might be on a continuous probability scale.) – Ben Bolker Apr 01 '14 at 01:05
  • yes, but how I interpret this ? Does it mean that the probability of paternity rise with the log scale ? Or that the probability of losing paternity rise with the log scale ? – anais Apr 01 '14 at 09:34
  • Please, show us the graph you are talking about and the code to reproduce it. Tks. – Andre Silva Sep 08 '16 at 17:23
  • Perhaps https://stats.stackexchange.com/questions/133623 now answers your question? – whuber Jun 21 '22 at 13:20

0 Answers0