Consultation: Translating SAS Multi-Level Models into R

Today at R club, we’ll help Jessica with a fun problem: she has a bunch of multi-level models written in SAS that she want’s to be able to play with in R. The motivation is that she can find R on pretty much any computer she sits down at these days, but only has access to one computer that runs SAS. For convenience, and because she is learning R, she wants to be able to work with these models using lme4 or nlme.

The data sounds really cool:

This project is looking at diurnal cortisol (4 times measurements per day per participant over the course of 2 days) in a cross sectional sample, ages 4-16 years old, specifically examining how two groups differ (previously institutionalized versus biologically reared). The second component of the project is examining how a 3rd variable (sleep) may moderate these effects.

Her data is already in long format, so we can get right into the modeling.