Regression, Discrimination and Measurement Models for Ordered Categorical Variables

Abstract

Summary Regression models for the analysis of ordered categorical variables are discussed with particular reference to the logistic model. Maximum likelihood estimation procedures are established for three common sampling plans including the case where sampling is conditional on the ordered variable. This was previously not available. The use of these models in discrimination is discussed and an example given. An original method for the establishment of rating scales based on a coarse direct assessment and related variables is introduced. Mention is made of some difficulties in the application of the models and of their potential use for the analysis of designed experiments with ordered response variables.

Statistical Analysis of Risk Factors Affecting the Prognosis of Biliary Atresia in Infants

“…In neonates, low birthweight might be a contributing factor for early onset of hyperbiliruenemia but it doesn't significantly affect to disease burden in case of biliary atresia. Work by Madan and coworkers reported 31% incidence of ABO incompatibility in terms of prolonged jaundice in neonates which strong confirms results of our study [8,14,15]. Remarkable ratio and strong association of mother child blood group incompatibility has been reported in our analyzed data.…”

“…Huge percentage difference of males (64%) and females (35.9%) showed the gender biasness/male disadvantage theory but Pearson chi square test results doesn't conform to this theory. Although a study reports high incidence of biliary atresia in females and associate the delay of cognitive skills in females to their high bilirubin level due to biliary atresia but in our data male peers were found more vulnerable to this disease [5,14]. For variable birthweight; 38% neonates were less than or equals to 2.5 kg, 36% neonates weigh between 2.6-2.7 kg while 32% cases were reported to have more than 2.8 kg weight at the time of birth.…”

Environmental policy evaluation from the perspective of farmers' perception in arid oasis ecosystem: Experiences of Minqin, Northwest China

“…In this study, we employed the proportional odds model (POM) to explore the potential influencing factors that affect the farmers' perception of the policy effectiveness. According to Anderson and Philips (1981) the model can be formulated as follows:…”

Ideal Point Discriminant Analysis and Ordered Response Categories

“…The probability of a response (a column) in a particular row is assumed equal to that of the row random variable falling into the interval corresponding to the column. The implementation of this approach is rela tively new in statistics (Anderson & Philips, 1981 ;Cox, 1970;McCullagh, 1980), but it has a long tradition in psychometrics (Torgerson, 1958). It is called the successive categories scaling method.…”