All maps of parameter estimates are misleading.

Maps are frequently used to display spatial distributions of parameters of interest, such as cancer rates or average pollutant concentrations by county. It is well known that plotting observed rates can have serious drawbacks when sample sizes vary by area, since very high (and low) observed rates are found disproportionately in poorly-sampled areas. Unfortunately, adjusting the observed rates to account for the effects of small-sample noise can introduce an opposite effect, in which the highest adjusted rates tend to be found disproportionately in well-sampled areas. In either case, the maps can be difficult to interpret because the display of spatial variation in the underlying parameters of interest is confounded with spatial variation in sample sizes. As a result, spatial patterns occur in adjusted rates even if there is no spatial structure in the underlying parameters of interest, and adjusted rates tend to look too uniform in areas with little data. We introduce two models (normal and Poisson) in which parameters of interest have no spatial patterns, and demonstrate the existence of spatial artefacts in inference from these models. We also discuss spatial models and the extent to which they are subject to the same artefacts. We present examples from Bayesian modelling, but, as we explain, the artefacts occur generally.

[1]  B. Efron,et al.  Data Analysis Using Stein's Estimator and its Generalizations , 1975 .

[2]  S. Jarcho.  Atlas of Cancer Mortality for U.S. Counties , 1976 .

[3]  I. Leck,et al.  Atlas of Cancer Mortality For U.S. Counties: 1950-1969 , 1976, British Journal of Cancer.

[4]  Donald B. Rubin,et al.  Using Empirical Bayes Techniques in the Law School Validity Studies , 1980 .

[5]  John M. Chambers,et al.  Graphical Methods for Data Analysis , 1983 .

[6]  Beat Kleiner,et al.  Graphical Methods for Data Analysis , 1983 .

[7]  T. Louis.  Estimating a population of parameter values using Bayes and empirical Bayes methods , 1984 .

[8]  D. Rubin.  Bayesianly Justifiable and Relevant Frequency Calculations for the Applied Statistician , 1984 .

[9]  R. Tsutakawa,et al.  Empirical Bayes estimation of cancer mortality rates. , 1985, Statistics in medicine.

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

[11]  Roger A. Sugden,et al.  Multiple Imputation for Nonresponse in Surveys , 1988 .

[12]  C. Muir.  Cancer mapping: overview and conclusions. , 1989, Recent results in cancer research. Fortschritte der Krebsforschung. Progres dans les recherches sur le cancer.

[13]  D. Rubin,et al.  Multiple Imputation for Nonresponse in Surveys , 1989 .

[14]  M. Woodbury,et al.  Empirical Bayes procedures for stabilizing maps of U.S. cancer mortality rates. , 1989, Journal of the American Statistical Association.

[15]  A. Mollié,et al.  Empirical Bayes estimates of cancer mortality rates using spatial models. , 1991, Statistics in medicine.

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

[17]  L H Moulton,et al.  Potential Pitfalls in Interpreting Maps of Stabilized Rates , 1994, Epidemiology.

[18]  T. Louis,et al.  Empirical Bayes methods for stabilizing incidence rates before mapping. , 1994, Epidemiology.

[19]  Noel A Cressie,et al.  Statistics for Spatial Data, Revised Edition. , 1994 .

[20]  F F Nobre,et al.  Feasibility of contour mapping epidemiological data with missing values. , 1995, Statistics in medicine.

[21]  L. Pickle,et al.  Effects of the choice of age-adjustment method on maps of death rates. , 1995, Statistics in medicine.

[22]  P Schlattmann,et al.  Covariate adjusted mixture models and disease mapping with the program DismapWin. , 1996, Statistics in medicine.

[23]  L. Bernardinelli,et al.  Bayesian methods for mapping disease risk , 1996 .

[24]  Xiao-Li Meng,et al.  POSTERIOR PREDICTIVE ASSESSMENT OF MODEL FITNESS VIA REALIZED DISCREPANCIES , 1996 .

[25]  D. Rubin.  Multiple Imputation After 18+ Years , 1996 .

[26]  P N Price,et al.  Bayesian prediction of mean indoor radon concentrations for Minnesota counties. , 1996, Health physics.

[27]  J. Estève,et al.  Practical approaches to disease mapping , 1996 .

[28]  Jack Cuzick,et al.  Geographical and environmental epidemiology : methods for small-area studies , 1997 .

[29]  P N Price,et al.  Predictions and maps of county mean indoor radon concentrations in the mid-Atlantic states. , 1997, Health physics.

[30]  Bradley P. Carlin,et al.  BAYES AND EMPIRICAL BAYES METHODS FOR DATA ANALYSIS , 1996, Stat. Comput..

[31]  A. Rukhin.  Bayes and Empirical Bayes Methods for Data Analysis , 1997 .