Premium
AN INTRODUCTION TO SMOOTHING INCIDENCE RATES BY PENALIZED LIKELIHOOD
Author(s) -
DOWNER ROBERT G.
Publication year - 1996
Publication title -
statistics in medicine
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.996
H-Index - 183
eISSN - 1097-0258
pISSN - 0277-6715
DOI - 10.1002/(sici)1097-0258(19960415)15:7/9<907::aid-sim259>3.0.co;2-d
Subject(s) - statistics , smoothing , pooling , statistic , mathematics , bayes' theorem , homogeneity (statistics) , type i and type ii errors , econometrics , test statistic , likelihood ratio test , maximum likelihood , statistical hypothesis testing , computer science , bayesian probability , artificial intelligence
Most recent developments in the smoothing of disease incidence rates have involved empirical Bayes procedures. Through the introduction of an intersite distance penalty, the pooling of information by directly penalizing the maximum likelihood estimates is proposed. A distance type statistic tests a null hypothesis of homogeneity among the rates and is applied to gastric cancer incidence of Nova Scotia municipalities. For various spatial patterns, simulation results are presented which investigate the power of the test, the total mean squared error and the distribution of the smoothed estimates.