AbstractThe aim of this study was to develop the best possible method for providing estimates of
relative risk for cancers by utilizing a multivariate approach through shared component
model to produce specific and shared spatial patterns of cancer disease in Libya. The
methodology in this study is statistical model to development to jointly analyze the spatial
variation in rates of several cancers with their four spatial components without any
conditions. The data for incidence of breast, lung, prostate, bladder, stomach, colon and
liver in Libya districts, were used in this study. Four risk factors representing smoking, low
physical activity, overweight and alcohol consumption were considered. A general
framework for jointly modelling the variation of seven cancers, some of which share latent
spatial fields, is considered. Finally, a new Multivariate Shared Component model is
successfully developed. The procedure for predicting relative risks based on the model is
presented in ord.....