bootDominanceAnalysis: Bootstrap analysis for Dominance Analysis

In dominanceanalysis: Dominance Analysis

Description

Bootstrap procedure as presented on Azen and Budescu (2003). Provides the expected level of dominance of predictor X_i over X_j, as the degree to which the pattern found on sample is reproduced on the bootstrap samples. Use summary() to get a nice formatted data.frame

Usage

bootDominanceAnalysis(
  x,
  R,
  constants = c(),
  terms = NULL,
  fit.functions = "default",
  null.model = NULL,
  ...
)

Arguments

x	lm, glm or lmer model
R	number on bootstrap resamples
constants	vector of predictors to remain unchanged between models. i.e. vector of variables not subjected to bootstrap analysis.
terms	vector of terms to be analyzed. By default, obtained from the model
fit.functions	list of functions which provides fit indices for model. See fit.functions param in dominanceAnalysis function.
null.model	only for linear mixed models, null model against to test the submodels. i.e. only random effects, without any fixed effect.
...	Other arguments provided to lm or lmer (not implemented yet).

Examples

lm.1<-lm(Employed~.,longley)
da.boot<-bootDominanceAnalysis(lm.1,R=1000)
summary(da.boot)

dominanceanalysis documentation built on Jan. 13, 2021, 3:47 p.m.