Let's take data :
binary_var <- sample(0:1, 100, replace = T)
indep_var <- rnorm(100)
And let's consider model output of logit regression :
glm(binary_var~indep_var, family = quasibinomial(link = 'logit'))
Call: glm(formula = binary_var ~ indep_var, family = quasibinomial(link = "logit"))
Coefficients:
(Intercept) indep_var
0.04441 -0.09845
Degrees of Freedom: 99 Total (i.e. Null); 98 Residual
Null Deviance: 138.6
Residual Deviance: 138.3 AIC: NA
As you can see, AIC is calculated as 'NA'. I've read about this and it seems that AIC cannot be calculated easily for quasibinomial models. I also found package AICcmodavg which should allowed me to calculate wanted AIC. However I couldn't find proper function in this package which would allowed me to do so. Could you please help with me with calculating AIC for quasibinomial models ?