Fitting a Poisson Log-Linear Model to the Incidence of Schizophrenia in the State of Bahrain

Document Type : Original Article

Author

Central Statistics Organization, The State of Bahrain

Abstract

The paper deals with the application of a categorical data analysis methodology to investigate the socio-demographic correlates of incidence of schizophrenia in the State of Bahrain in the period 1988-1996. For this purpose, the Poisson log-linear model was devised: a concise summary of underlying theory was reviewed. The response variable is the number of new schizophrenic cases while the covariates are calendar year; sex; age and marital status. The maximum likelihood method, with the application of the iterative procedure of the ridge stabilized Neton-Raphson algorithm, was used to estimate model parameters. The model has proven to fit the data very well.

Keywords