CODEWITHSUNDEEP
rmodeloutput
Max Liebergesell
Max Liebergesell

Reputation: 1

In a binomial GLMM how do I include all levels in my model output?

I'm a little new to this so bare with me but basically I'm currently running this model:

fit.glmm_A_B_C_D = glmer(cbind(success, failure) ~ treatment_letter + (1|trial_rep), family = binomial, data=Con_GLMM_A_B_C_D_Data)

In this model my fixed factor (treatment letter) has 4 levels to it: Treatments A, B, C, and D. Each treatment has 12 subjects within my data so everything is balanced. I'm having issues with the model output because it is only showing me the intercept, treatment B, treatment C, and treatment D when I asked for the model summary. Here is the output:

     AIC      BIC   logLik deviance df.resid 
   136.1    145.5    -63.1    126.1       43 

Random effects:
 Groups    Name        Variance Std.Dev.
 trial_rep (Intercept) 0        0       
Number of obs: 48, groups:  trial_rep, 12

Fixed effects:
                  Estimate Std. Error z value Pr(>|z|)   
(Intercept)         0.5390     0.2746   1.963  0.04965 * 
treatment_letterB  -1.1054     0.3874  -2.854  0.00432 **
treatment_letterC   0.1788     0.3870   0.462  0.64401   
treatment_letterD  -0.1335     0.3888  -0.343  0.73125   

Correlation of Fixed Effects:
            (Intr) trtm_B trtm_C
trtmnt_lttB -0.709              
trtmnt_lttC -0.709  0.503       
trtmnt_lttD -0.706  0.501  0.501
convergence code: 0
boundary (singular) fit: see ?isSingular

How do I get the model to show me treatment A in the output? Thanks!

Upvotes: 0

Views: 52

Answers (1)

Ann Brush
Ann Brush

Reputation: 11

You can use the emmeans package, the syntax of which goes something like this:

fit.glmm_A_B_C_D_MainEffect <- emmeans(fit.glmm_A_B_C_D, "treatment_letter", type = "response")

That will provide mean probabilities and uncertainty estimates for the overparameterized model main effect of treatment on the data scale (back transformed from the logit scale).

Upvotes: 1

Related Questions