combine_rubin: Combine results from different imputations using Rubin's...

Description Usage Arguments Value References

View source: R/mifa-ci.R

Description

Applies Rubin's rules to combine estimates and variance-covariance matrices from different imputations.

Usage

1
combine_rubin(param_imps, cov_imps)

Arguments

param_imps

Matrix containing estimated parameters in each imputation as its rows.

cov_imps

List of estimated covariance matrices for each imputation.

Value

A list:

param_est

Vector of combined parameter estimates with the same length as columns in param_imps.

cov_param

Combined variance-covariance matrix of size n x n, where n is the number of columns in param_imps.

between_cov

Between imputations variance-covariance matrix of size n x n, where n is the number of columns in param_imps.

References

Rubin D. B. Multiple imputation for nonresponse in surveys (2004). John Wiley & Sons.


mifa documentation built on Jan. 22, 2021, 5:10 p.m.