I have two models binomial logit:
modelo_logit_viv <- glm( SAP ~sexo + edad + peso + niv_est + enf_cron + sit_lab + frec_act_fis + GHQ_12 + ingreso_eq + n_dormitorios + cont_indus + delincuencia, # variables de estudio data = datos_modelo, family = binomial(link = "logit"),na.action = "na.omit")
- Model B (nested model A):
modelo_logit <- glm(SAP ~ sexo + edad + peso + niv_est + enf_cron + sit_lab + frec_act_fis + ingreso_eq + GHQ_12, data = modelo_logit_viv$model, family = binomial(link = "logit"))
c(edad, peso, ingreso_eq, GHQ_12) These are continuous variables, the others are categorical (factor) variables.
I want to analyze whether the characteristics of housing (study variables) have an influence on self-rated health status (SAP). All variables are significant in both models. However, I would like to perform an analysis of variance (ANOVA) between these two models to check that model B is better than model B. I therefore execute:
And I get the following table:
Resid. Df Resid. Dev Df Deviance 1 16805 15439 2 16802 15420 3 18.644
Can the interpretation of this table tell me if there is an influence between the characteristics of the accommodation and the state of health?
Another way to ask this question is:
18.644 the F test statistic that I need to compare with the Snedecor F table to determine whether or not to accept the null hypothesis that there are mean differences between the two models?
How else to compare these two models with an ANOVA in R?