First, using the WVS dataset, run a regular regression model using ETHDIVORCE as the dependent variable and RELIGIONIMP as the independent variable. Then, imagine you didn’t have the ETHDIVORCE variable, and instead had only the DIVORCE2CAT dichotomous variable.

Run a logistic regression model using DIVORCE2CAT as the dependent variable and RELIGIONIMP as the independent variable. Interpret both models by making predictions for a country where 20% of the population says religion is very important and then for a country where 90% of the population says religion is very important.