Fitting non-linear structural relationships using SAS procedure NLMIXED

Both linear and non-linear structural relationships can be formulated explicitly in terms of random-effects models. When the random effects are normally distributed, the SAS NLMIXED procedure can be used to fit such models to data by using maximum likelihood estimation. Any distribution for the data conditional on the random effects can be used although certain common distributions are built into the procedure. Techniques for detecting parameter unidentifiability are discussed. Examples covering a range of applications are given.