Ordinal regression is a relatively new statistical method developed fo
r analyzing ranked outcomes. In the past, ranked scales have often bee
n analyzed without making full use of the ordinality of the data or, a
lternatively, by assigning arbitrary numerical scores to the ranks. Wh
ile ordinal regression models are now available to make full use of ra
nked data, they are not used widely. This article, directed to clinica
l researchers and epidemiologists, provides a description of the prope
rties of these methods. Using ordinal measures of back pain in a follo
w-up study of adolescent idiopathic scoliosis, we illustrate the advan
tages of these methods and describe how to interpret the estimated par
ameters. Comparisons with binary logistic regression are made to show
why a single dichotomization of the ordinal scale may lead to incorrec
t inferences. Two ordinal models (the proportional odds and the contin
uation ratio models) are discussed, and the goodness-of-fit of these m
odels is examined. We conclude that ordinal regression is a tool that
is powerful, simple to use, and produces an interpretable parameter th
at summarizes the effect between groups over all levels of the outcome
. Copyright (C) 1997 Elsevier Science Inc.