boot.lavan: Nonparametrical bootstrapping of a SEM model fit by 'lavaan'.

In stremo: Functions to help the process of learning structural equation modelling

Description

This function does very crude nonparametrical bootstrapping of a fitted sem model using lavaan.

Usage

boot.lavaan(fitted.model, n)

Arguments

fitted.model	A model fitted by sem or cfa.
n	Number of replications.

Details

Steps:

1. Fit a model normally using the arguments 'sample.cov' and 'sample.nobs' instead of 'data';

2. Get the sigma hat from the fitted model and build an empirical dataset with 'sample.nobs' observations;

3. Get 'sample.nobs' observations from this new dataset with replacement;

4. Fit a new model using the sample taken from the simulated dataset;

5. Repeat 3 and 4 'n' times.

Value

X2	Bootstrapped maximum likelihood chi-squared
p.X2	P value of the bootstrapped maximum likelihood chi-squared
est	Estimated coefficients
se	Bootstrapped standard errors
z	Bootstrapped z values
p.est	P values of the bootstrapped coefficients
coefs	Matrix with the values of all coefficients in all runs

Examples

## Not run: 
data(albert)
fit <- sem(albert.model, sample.cov = albert.litho.cov, sample.nobs
           = 107)
booted.fit <- boot.lavaan(fit, 1000)

## End(Not run)

stremo documentation built on Jan. 15, 2017, 7:33 p.m.