Spatial prediction of counts and rates

In this paper we provide both theoretical and empirical comparisons of marginal and conditional methods for analysing spatial count data. We focus on methods for spatial prediction developed from a generalized linear mixed model framework and compare them with the traditional linear (kriging) predictor. Prediction methods are illustrated and compared through a case study based on real data and through a detailed simulation study. The paper emphasizes a better understanding of the strengths and weaknesses of each approach.

[1]  Noel A Cressie,et al.  Statistics for Spatial Data. , 1992 .

[2]  S. Zeger,et al.  Longitudinal data analysis using generalized linear models , 1986 .

[3]  Subhash R. Lele,et al.  A Regression Method for Spatial Disease Rates: An Estimating Function Approach , 1997 .

[4]  Derek G. Cook,et al.  Analysing geographic variation in cardiovascular mortality: methods and results. , 1982 .

[5]  M. Lindstrom,et al.  A survey of methods for analyzing clustered binary response data , 1996 .

[6]  P S Albert,et al.  A latent process regression model for spatially correlated count data. , 1997, Biometrics.

[7]  Lancaster Ho,et al.  Some geographical aspects of the mortality from melanoma in Europeans. , 1956 .

[8]  D. Clayton,et al.  Empirical Bayes estimates of age-standardized relative risks for use in disease mapping. , 1987, Biometrics.

[9]  N. Breslow,et al.  Approximate inference in generalized linear mixed models , 1993 .

[10]  J. N. Morris,et al.  Mortality and hardness of local water-supplies. , 1968, Lancet.

[11]  A. Goldberger Best Linear Unbiased Prediction in the Generalized Linear Regression Model , 1962 .

[12]  Noel A. C. Cressie,et al.  Statistics for Spatial Data: Cressie/Statistics , 1993 .

[13]  Ecological analyses and case-control studies of gastric cancer and leukaemia in relation to DBCP in drinking water in Fresno County, California. , 1989, British journal of industrial medicine.

[14]  G. Melvyn Howe,et al.  National Atlas of Disease Mortality in the United Kingdom. , 1964 .

[15]  David A. Harville,et al.  Decomposition of Prediction Error , 1985 .

[16]  F. Carrat,et al.  Epidemiologic Mapping using the “Kriging” Method: Application to an Influenza-like Epidemic in France , 1992 .

[17]  G Christakos,et al.  A study of the breast cancer dynamics in North Carolina. , 1997, Social science & medicine.

[18]  J. Bithell An application of density estimation to geographical epidemiology. , 1990, Statistics in medicine.

[19]  Assessment of forecasts and forecast uncertainty using feneralized linear regression models for time series count data , 2001 .

[20]  S. Zeger A regression model for time series of counts , 1988 .

[21]  T. Louis,et al.  Approximate Posterior Distributions for Incomplete Data Problems , 1982 .

[22]  G. Robinson That BLUP is a Good Thing: The Estimation of Random Effects , 1991 .

[23]  Forbes Ja Infectious diseases, today and tomorrow. , 1966 .

[24]  D. Bates,et al.  Nonlinear mixed effects models for repeated measures data. , 1990, Biometrics.

[25]  J. Fraumeni,et al.  Geographic patterns of oral cancer in the United States: etilogic implications. , 1977, Journal of chronic diseases.

[26]  J. Lindsey Models for Repeated Measurements , 1993 .

[27]  Dale L. Zimmerman,et al.  A comparison of spatial semivariogram estimators and corresponding ordinary Kriging predictors , 1991 .

[28]  W. W. Stroup,et al.  A Generalized Linear Model Approach to Spatial Data Analysis and Prediction , 1997 .

[29]  George Christakos,et al.  Bayesian Maximum Entropy Analysis and Mapping: A Farewell to Kriging Estimators? , 1998 .

[30]  Mervyn G. Marasinghe,et al.  SAS System for Linear Models , 1991 .

[31]  M. Kenward,et al.  Small sample inference for fixed effects from restricted maximum likelihood. , 1997, Biometrics.

[32]  N. Cressie,et al.  Spatial Modeling of Regional Variables , 1993 .

[33]  R. Wolfinger,et al.  Generalized linear mixed models a pseudo-likelihood approach , 1993 .