From the course: Descriptive Healthcare Analytics in R

Unlock the full course today

Join today to access over 22,600 courses taught by industry experts or purchase this course individually.

Designing confounders: Other variables used in analysis

Designing confounders: Other variables used in analysis - R Tutorial

From the course: Descriptive Healthcare Analytics in R

Start my 1-month free trial

Designing confounders: Other variables used in analysis

“

- [Instructor] Finally, we've made it to the last section of Chapter two where we will finalize the design of our confounders. So referring to my web of causation, there are only a few confounders left to include: self-reported general health, whether or not the respondent has health insurance or a health plan, highest education level of the respondent, respondent's household income and obesity status, and respondent's exercise habits. I added these to the dictionary so let's go take a look. Here's the general health question. This is a helpful question for adjusting for a chronic disease. Rather than putting all the diseases in which can be a problem in modeling, the general health question tends to take care of all that variation. Let's look at the GENHLTH tab. You will see that the grouping variable I designed, GENHLTH2, is almost like the native variable but groups together the Refused and unknown but you will see that I only make indicator variables for the FAIRHLTH level and the…

Contents