Linear mixed models and semi-parametric smoothing in fisheries research with implementation of winBUGS

Linear mixed models (Fisheries Research Vol. 70, 2004) and semi-parametric models (such as generalized additive models, GAM, and spline smoothing) have growing important in fisheries researches.  The analysis of complex correlated data structure is a typical application of linear mixed models, which can incorporate semi-parametric smoothing of the relationship between outcome and covariates.  This report shows the connection of mixed models and smoothing and the implementation of Markov Chain Monte Carlo (MCMC) using WinBUGS.