A local linear estimation procedure for functional multilevel modeling

Richard Li, Tammy L. Root, and Saul Shiffman

Li, Root and Shiffman introduce us to the field of functional multilevel modeling. This technique expands on the traditional linear mixed model by allowing coefficients to vary nonparametrically over time and introducing a local linear regression estimation procedure for the nonparametric coefficients. It is a powerful graphical tool that examines relationships in data that vary across time. The proposed model and estimation procedure are demonstrated using intensive longitudinal data, known as Ecological Momentary Assessment (EMA), on smoking cessation data.

• SAS Code for Functional Multilevel Models