resid_conditional: Conditional residuals

In HLMdiag: Diagnostic Tools for Hierarchical (Multilevel) Linear Models

Description

Calculates conditional residuals of lmerMod and lme model objects.

Usage

## Default S3 method:
resid_conditional(object, type)

## S3 method for class 'lmerMod'
resid_conditional(
  object,
  type = c("raw", "pearson", "studentized", "cholesky")
)

## S3 method for class 'lme'
resid_conditional(
  object,
  type = c("raw", "pearson", "studentized", "cholesky")
)

Arguments

object	an object of class lmerMod or lme.
type	a character string specifying what type of residuals should be calculated. It is set to "raw" (observed - fitted) by default. Other options include "pearson", "studentized", and "cholesky". Partial matching of arguments is used, so only the first character needs to be provided.

Details

For a model of the form Y = X β + Z b + ε, four types of marginal residuals can be calculated:

raw

e = Y - X \hat{beta} - Z \hat{b}

pearson

e / √{diag(\hat{Var}(Y|b)})

studentized

e / √{diag(\hat{Var}(e))}

cholesky

\hat{C}^{-1} e where \hat{C}\hat{C}^\prime = \hat{Var}(e)

Value

A vector of conditional residuals.

References

Singer, J. M., Rocha, F. M. M., & Nobre, J. S. (2017). Graphical Tools for Detecting Departures from Linear Mixed Model Assumptions and Some Remedial Measures. International Statistical Review, 85, 290–324.

Schabenberger, O. (2004) Mixed Model Influence Diagnostics, in Proceedings of the Twenty-Ninth SAS Users Group International Conference, SAS Users Group International.

HLMdiag documentation built on May 2, 2021, 9:06 a.m.