期刊名称：Brazilian Journal of Probability and Statistics
印刷版ISSN：0103-0752
出版年度：2004
卷号：18
期号：1
页码：1-18
出版社：Brazilian Statistical Association
摘要：Several available approaches of analyzing bivariate/multi-variate ordinal categorical data are discussed in the present paper using theexample of an ophthalmologic data. Wisconsin epidemiologic study of dia-betic retinopathy (WESDR) is a population-based epidemiologic study car-ried out in Southern Wisconsin during the eighties of the last century. Theresulting data were analyzed by di.erent statisticians and ophthalmologistsduring the last two decades. We can primarily divide the statistical analy-sis in three categories. Some ophthalmologists and statisticians involved inthe trial analyzed the data to understand the epidemiologic behavior. Somestatisticians analyzed the data in the context of smoothing spline. Someother statisticians considered the data as a nice example of bivariate ordinaldataset, due to the very nature of the data, and analyzed the data in thatdirection. There are, of course, di.erent angles of study within a particulardirection. The present paper reviews the WESDR study and the resultingdata. Di.erent directions of studies are then described with due references
