Knowledge and Data Management White Papers
Using PROC NLMIXED and PROC GLMMIX to Analyze Dyadic Data With a Dichotomous Dependent Variable
Overview In the social and health sciences, data are often hierarchical (subjects nested in groups). One kind of hierarchy is the dyad, or couple, where each group consists of two subjects. Dyadic data pose particular problems for statistical analysis for several reasons: First, variation may occur at the individual or dyadic level. Second, the data are not independent. Third, the small group size poses special difficulties. Multilevel models have been used for dyadic data; the paper demonstrates the use of PROC NLMIXED and PROC GLIMMIX, and discuss the strengths and weaknesses of this approach in general, and these SAS procedures in particular. The authors illustrate this with data on predictors of viral Hepatitis C among heterosexual couples in Harlem in New York City.
| Publisher | SAS Institute | File Format | |
|---|---|---|---|
| Date Published | May 2007 | Downloads | 6 |
| Format | White Papers | ||
| Topics | |||
-
Did you find this white paper useful? 6 out of 7 users found this white paper useful
White Paper RSS
Download
Register
