Determining the number of interactions between the independent variables

Question

I am trying to use GLMMs models to analysis the morbidity status of child (yes or no) with mother’s demographic and environmental factors like

1. Wealth with factors ("Lower quartiles”,"Middle quartiles") and “Higher quartiles"
2. Residence (Urban" and Rural")
3. Education ("Never attended”, “Primary education", “Secondary education", "Above secondary education")
4. Marital status (“Married or with partner", Widowed or divorced”, ” Never married"))
5. Age (Age between 15-24”, “Age between 25-34", "Age above 34")
6. Parity (“Zero Parity", “Parity between 1 and 2", “Parity between 3 and 4”, Parity above 5")

The following is r code formula that I try to fix determine the model

Fit1<-glmer (Morbidity~Wealth+Residencey+education+Marital+Age+Parity +(1|region)+(1|participant_id),family=binomial,data=M)

Dear respected professionals, I try to use GLMMs model, Then I face a bit more serious problems on the following issues like

Q1. How many interaction is possible between the given independent variables and how can specify it? If I try to test the interaction between certain variables, it is honestly insignificance.

Q2. The appropriate R code to plot the interaction effect between variables two variable and the responsice variable, morbidity. If somebody can help me on it, I can’t thank you enough your quick response.

Answer 1

Welcome to CV. You ask

Q1. How many interaction is possible between the given independent variables and how can specify it? If I try test the interaction between certain variables, it is honestly insignificant.

I don't know what "honestly insignificant" means, but the number of two way interactions among $p$ variables is $p(p-1)/2$. If you want to look at higher-way interactions, there are a lot more.

Q2. The appropriate R code to plot the interaction effect between variables two variable and the response variable, morbidity.

There is no such thing as a interaction between the independent variables and the response variable. Interactions are among independent variables.

You didn't ask about this, but categorizing quantitative variables (wealth and age) is usually a bad idea, unless you have some strong, substantive reason for doing so.

I'm also not clear why you need a GLMM (that is a mixed model). There doesn't seem to be any clustering, either in time or space or otherwise, and the errors would seem to be independent (but maybe there is more going on that you have not told us).