regression models for ordinal responses a review of methods

G severity of preeclampsia none mild severe are often

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: ables originally measured on an ordinal scale (e.g. severity of preeclampsia: none, mild, severe) are often categorized into several binary variables during statistical analysis. Consider, as a motivating example, the data set described in Table 1. This data is derived from a clinical * The Center for Perinatal Health Initiatives, Department of Obstetrics, Gynecology and Reproductive Sciences, University of Medicine and Dentistry of New Jersey, Robert Wood Johnson Medical School, 125 Paterson St, New Brunswick, NJ 08901–1977, USA. † Department of Epidemiology, The Rollins School of Public Health, Emory University, Atlanta, GA, USA. 1323 1324 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY TABLE 1 Ratings of drugs on a single-dose, post-operative analgesic clinical triala Drug Rating of the drugs Poor C15 & C60 Z100 & EC4 ˆ Log-odds ratioc, β Standard error se (β ) ˆˆ ˆ Odds ratios, eβ Fair Good Very goodb Total 17 10 18 4 0.7013 0.4491 2.0 20 13 1.5476 0.4016 4.7 5 34 2.6384 0.5535 13.9 61 60 1.0 a Source: ref.2 Fa...
View Full Document

This document was uploaded on 02/25/2014.

Ask a homework question - tutors are online