Open jalvord1 opened 6 years ago
After the four main variables we chose for the model, we should also have more confounding variables to make sure our model is accurate. Here are some potential variables:
Generalhealth, physicalhealth, mentalhealth, poorhealth, healthplan, medicalcost, checkup, sexfactor, sexualorientation, doctorvisits, medicationcost, bmi
Variables to add to our data set: age, marital status, smoking
substance abuse
Age as a numeric variable is not available so we are going to add the age as a categorical variable with 6 levels
Alcohol consumption: combine the heavy drinker and binge drinker categories to create a binary variable for unhealthy drinking practices
data is too large to upload to repository. I selected certain variables based on what I thought we wanted and filtered out those missing info for "trnsgnr" var, but we should go through and decide. Codebook is already uploaded and can be used to do this.