This book is an attempt to formulate a sociological theory of the phenomena which is otherwise known as substance abuse and substance dependency in the medical field. The first step toward such a theory is a critique of how DSM-IV presents diagnostic criteria for substance abuse and substance dependency as if these criteria are valid and reliable ways of measuring mental illness and bodily dysfunction. I point to the fact that these criteria neither are valid or reliable variables for measurement nor theoretically grounded defi nitions of the phenomena. My point is that diagnosis is an elaborate way of disguising moral judgments as medical assessments of illness or dysfunction. The second step towards the formulation of my theory goes via my empirical study of nine men and women who call themselves sex and love addicts. By conducting deep interviews with them I conclude that it is impossible to talk about sex and love addiction without constantly referring to stereotypes and widely held assumptions of alcoholics and drug addicts. Moreover it is apparent that these men and women are making moral judgments about themselves rather than pointing to some form of genuine disruption which they cannot control. From my critique and my empirical study I am able to identify a number of areas that a theory of addiction, as opposed to the medical view where this phenomena are being broken down in two sub categories (abuse and dependency), must be able to handle to explain addiction generally and specifically in relation to alcohol, drugs, sex and other forms of social deviance. A number of philosophical hallmarks of medicine is also identifi ed and seen as part of the problem, therefore I mean that a successful theory of addiction must transcend the mind body dualism of Descartes as it is a cornerstone in the medical view upon and understanding of humans. By reconstructing the habitus and field concepts in Pierre Bourdieus theory of practice I mean that it is possible to understand addiction from an action theory point of view. This discussion is also broadened by a pragmatist discussion of the mind and a neo meadian theory of the emotional self. My conclusion is that addiction can be understood and explained within an action theory which focus upon individual as well as collective action and understand this as habitual practice that is partly embodied and therefore not discursive – habits of which we cannot easily speak is seen as one of the compelling components in addiction rather than a mythical loss of control which is nowhere to be found but as a rhetoric grip stemming from the AA view upon addiction as a disease.

Available statistical methodologies focus more on accommodating continuous variables, however recently dealing with count data has received high interest in the statistical literature. In this dissertation, we propose some statistical approaches to investigate linear and nonlinear dependencies between two discrete random variables, or between a discrete and continuous random variables. Copula functions are powerful tools for modeling dependencies between random variables. We derive copula-based population version of Spearman’s rho when at least one of the marginal distribution is discrete. In each case, the functional relationship between Kendall’s tau and Spearman’s rho is obtained. The asymptotic distributions of the proposed estimators of these association measures are derived and their corresponding confidence intervals are constructed, and tests of independence are derived. Then, we propose a Bayesian copula factor autoregressive model for time series mixed data. This model assumes conditional independence and shares latent factors in both mixed-type response and multivariate predictor variables of the time series through a quadratic timeseries regression model. This model is able to reduce the dimensionality by accommodating latent factors in both response and predictor variables of the high-dimensional time series data. A semiparametric time series extended rank likelihood technique is applied to the marginal distributions to handle mixed-type predictors of the high-dimensional time series, which decreases the number of estimated parameters and provides an efficient computational algorithm. In order to update and compute the posterior distributions of the latent factors and other parameters of the models, we propose a naive Bayesian algorithm with Metropolis-Hasting and Forward Filtering Backward Sampling methods. We evaluate the performance of the proposed models and methods through simulation studies. Finally, each proposed model is applied to a real dataset.