Description

1. Revisit the Bangladesh fertility data, data(bangladesh). Fit a model with both varying intercepts by district_id and varying slopes of urban (as a 0/1 indicator variable) by district_id. You are still predicting use.contraception. Inspect the correlation between the intercepts and slopes. Can you interpret this correlation, in terms of what it tells you about the pattern of contraceptive use in the sample? It might help to plot the varying effect estimates for both the intercepts and slopes, by district. Then you can visualize the correlation and maybe more easily think through what it means to have a particular correlation. Plotting predicted proportion of women using contraception, in each district, with urban women on one axis and rural on the other, might also help.
3. Modify any models from Problem 2 that contained that children variable and model the variable now as a monotonic ordered category, like education from the week we did ordered categories. Education in that example had 8 categories. Children here will have fewer (no one in the sample had 8 children). So modify the code appropriately. What do you conclude about the causal influence of each additional child on use of contraception?
1