residuals.polr: Residuals of a Binary or Ordered Regression In regr0: Building regression models

Description

Calculates quartiles and random numbers according to the conditional distribution of residuals for the latent variable of a binary or ordinal regression, given the observed response value. See Details for an explanation.

Usage

 ```1 2``` ```## S3 method for class 'polr' residuals(object, ...) ```

Arguments

 `object` the result of `polr` or of `glm(,family=binomial)` with binary data.
 `...` unused

Details

For binary and ordinal regression, the regression models can be described by introducing a latent response variable Z of which the observed response Y is a classified version, and for which a linear regression applies. The errors of this "latent regression" have a logistic distribution. Given the linearly predicted value eta[i], which is the fitted value for the latent variable, the residual for Z[i] can therefore be assumed to have a logistic distribution.

This function calculates quantiles and random numbers according to the conditional distribution of residuals for Z[i], given the observed y[i].

Value

a data.frame with the variables

 `median` medians of the conditional distributions `lowq` lower quartiles `uppq` upper quartiles `random` random numbers, drawn according to the conditional distributions `fit` linear predictor values `y` observed response values

Author(s)

Werner A. Stahel, ETH Zurich

References

See http://stat.ethz.ch/~stahel/regression

`condquant`, `plot.regr`
 ```1 2 3 4 5``` ```require(MASS) house.plr <- polr(Sat ~ Infl + Type + Cont, weights = Freq, data = housing) house.resid <- residuals(house.plr) head (house.resid) summary(house.resid) ```