This talk focuses on multivariate and computational approaches that are being developed in the alcohol field. There is substantial monetary support for conducting alcohol research. Alcohol use and problems are complex behaviors by individuals, across their life spans, while embedded in a number of social and economic networks. This complexity, coupled with the research support primarily from the National Institutes of Health (NIH), has led to numerous data collection and research projects, many of which require sophisticated multivariate and spatial statistical approaches. Some of the methods used to model alcohol use and prob- lems are latent growth curves, multilevel models, and latent class analysis. These techniques allow for the examination and modeling of both individual and group level factors. However, these types of models are not suitable for mining large data sets. In this paper, we exploit regional data in Erie County, NY to illustrate the use of multivariate and spatial analysis tools in alcohol studies.

This is joint work with Yasmin H. Said and Edward J. Wegman.