residuals.SemiParBIVProbit: Residuals

Description

residuals efficiently calculates Pearson and working residuals.

Usage

## S3 method for class 'SemiParBIVProbit':
residuals(object,...)

Arguments

object

A fitted SemiParBIVProbit object as produced by SemiParBIVProbit().

...

Other arguments.

Value

r.pIt contains three columns corresponding to the Pearson residuals from the fitted model. The first two columns are of interest; they refer to the two linear predictors in the bivariate probit model.
r.wIt contains three columns corresponding to the working residuals from the fitted model.

WARNINGS

Unfortunately residual plots from models fitted to binary data do not contain much information about the goodness-of-fit of the model. For instance, if one wants to check distributional assumptions from residuals, QQ-plots and/or half-normal plots with simulated envelope provide no useful information (this has also been checked in simulation). For binary data, it is necessary to be able to group the residuals according to some criterion. For the sample selection case, such residuals are not meaningful and alternative definitions have to be employed.

Details

Justification for these residuals is from the penalized IRLS algorithm used to fit the model (Marra and Radice, 2011).

References

Marra G. and Radice R. (2011), Estimation of a Semiparametric Recursive Bivariate Probit in the Presence of Endogeneity. Canadian Journal of Statistics, 39(2), 259-279.