Location-scale cumulative odds models for ordinal data: A generalized non-linear model approach

Research output: Contribution to journalArticle

Abstract

Proportional odds regression models for multinomial probabilities based on ordered categories have been generalized in two somewhat different directions. Models having scale as well as location parameters for adjustment of boundaries (on an unobservable, underlying continuum) between categories have been employed in the context of ROC analysis. Partial proportional odds models, having different regression adjustments for different multinomial categories, have also been proposed. This paper considers a synthesis and further generalization of these two families. With use of a number of examples, I discuss and illustrate properties of this extended family of models. Emphasis is on the computation of maximum likelihood estimates of parameters, asymptotic standard deviations, and goodness-of-fit statistics with use of non-linear regression programs in standard statistical software such as SAS.

Original languageEnglish (US)
Pages (from-to)1191-1203
Number of pages13
JournalStatistics in Medicine
Volume14
Issue number11
StatePublished - 1995
Externally publishedYes

Fingerprint

Proportional Odds Model
Ordinal Data
Odds
Nonlinear Dynamics
Nonlinear Model
Adjustment
ROC Analysis
Ordered Categories
Likelihood Functions
Social Adjustment
Statistical Software
Nonlinear Regression
Location Parameter
Scale Parameter
Goodness of fit
Maximum Likelihood Estimate
ROC Curve
Standard deviation
Regression Model
Continuum

ASJC Scopus subject areas

  • Epidemiology

Cite this

Location-scale cumulative odds models for ordinal data : A generalized non-linear model approach. / Cox, Christopher.

In: Statistics in Medicine, Vol. 14, No. 11, 1995, p. 1191-1203.

Research output: Contribution to journalArticle

@article{2950a01b653b4ed2b3b05cd39bdb033d,
title = "Location-scale cumulative odds models for ordinal data: A generalized non-linear model approach",
abstract = "Proportional odds regression models for multinomial probabilities based on ordered categories have been generalized in two somewhat different directions. Models having scale as well as location parameters for adjustment of boundaries (on an unobservable, underlying continuum) between categories have been employed in the context of ROC analysis. Partial proportional odds models, having different regression adjustments for different multinomial categories, have also been proposed. This paper considers a synthesis and further generalization of these two families. With use of a number of examples, I discuss and illustrate properties of this extended family of models. Emphasis is on the computation of maximum likelihood estimates of parameters, asymptotic standard deviations, and goodness-of-fit statistics with use of non-linear regression programs in standard statistical software such as SAS.",
author = "Christopher Cox",
year = "1995",
language = "English (US)",
volume = "14",
pages = "1191--1203",
journal = "Statistics in Medicine",
issn = "0277-6715",
publisher = "John Wiley and Sons Ltd",
number = "11",

}

TY - JOUR

T1 - Location-scale cumulative odds models for ordinal data

T2 - A generalized non-linear model approach

AU - Cox, Christopher

PY - 1995

Y1 - 1995

N2 - Proportional odds regression models for multinomial probabilities based on ordered categories have been generalized in two somewhat different directions. Models having scale as well as location parameters for adjustment of boundaries (on an unobservable, underlying continuum) between categories have been employed in the context of ROC analysis. Partial proportional odds models, having different regression adjustments for different multinomial categories, have also been proposed. This paper considers a synthesis and further generalization of these two families. With use of a number of examples, I discuss and illustrate properties of this extended family of models. Emphasis is on the computation of maximum likelihood estimates of parameters, asymptotic standard deviations, and goodness-of-fit statistics with use of non-linear regression programs in standard statistical software such as SAS.

AB - Proportional odds regression models for multinomial probabilities based on ordered categories have been generalized in two somewhat different directions. Models having scale as well as location parameters for adjustment of boundaries (on an unobservable, underlying continuum) between categories have been employed in the context of ROC analysis. Partial proportional odds models, having different regression adjustments for different multinomial categories, have also been proposed. This paper considers a synthesis and further generalization of these two families. With use of a number of examples, I discuss and illustrate properties of this extended family of models. Emphasis is on the computation of maximum likelihood estimates of parameters, asymptotic standard deviations, and goodness-of-fit statistics with use of non-linear regression programs in standard statistical software such as SAS.

UR - http://www.scopus.com/inward/record.url?scp=0029021856&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0029021856&partnerID=8YFLogxK

M3 - Article

C2 - 7667560

AN - SCOPUS:0029021856

VL - 14

SP - 1191

EP - 1203

JO - Statistics in Medicine

JF - Statistics in Medicine

SN - 0277-6715

IS - 11

ER -