tram (version 1.2-2)

Polr: Ordered Categorical Regression

Description

Some regression models for ordered categorical responses

Usage

Polr(formula, data, subset, weights, offset, cluster, na.action = na.omit, 
     method = c("logistic", "probit", "loglog", "cloglog", "cauchit"), ...)

Value

An object of class Polr, with corresponding coef, vcov, logLik, estfun, summary, print, plot and predict methods.

Details

Models for ordered categorical responses reusing the interface of polr. Allows for stratification, censoring and trunction.

The model is defined with a negative shift term, thus exp(coef()) is the multiplicative change of the odds ratio (conditional odds for reference divided by conditional odds of treatment or for a one unit increase in a numeric variable). Large values of the linear predictor correspond to large values of the conditional expectation response (but this relationship is nonlinear).

References

Torsten Hothorn, Lisa Moest, Peter Buehlmann (2018), Most Likely Transformations, Scandinavian Journal of Statistics, 45(1), 110--134, tools:::Rd_expr_doi("10.1111/sjos.12291").

Examples

  data("wine", package = "ordinal")

  library("MASS")
  polr(rating ~ temp + contact, data = wine)

  Polr(rating ~ temp + contact, data = wine)