bootmi: 'bootmi': BOOTstrap Multiple Imputed (survey) data

In svolpers/bootmi: Calculates ordinary bootstrap for deterministic imputation methods

Description

The bootmi package first and foremost implements bootstrapping imputed (survey) data as proposed by Shao and Sitter (1996). Furthermore, residual centering as proposed by Little, Bovaird and Widaman (2006), MICE as provided by van Buuren and Groothuis-Oudshoorn (2011), and the "transform, then impute procedure" as proposed by von Hippel (2009) are implemented.

The bootmi method generates bootraps samples and imputes each as proposed by Shao and Sitter (1996).

Usage

bootmi(formula, data, R = 5000, imputationMethod = c("none",
  "norm.predict", "pmm", "mean"), cilvl = 0.95, citype = c("norm",
  "basic", "perc", "bca"), residualinteractions = FALSE,
  centerinteractions = FALSE, simslopinfo = NULL, seed = FALSE,
  parallel = FALSE)

## Default S3 method:
bootmi(formula, data, R = 5000,
  imputationMethod = c("none", "norm.predict", "pmm", "mean"),
  cilvl = 0.95, citype = c("norm", "basic", "perc", "bca"),
  residualinteractions = FALSE, centerinteractions = FALSE,
  simslopinfo = NULL, seed = FALSE, parallel = FALSE)

Arguments

formula	A regression formula
data	A data.set
R	Number of bootstraps.
seed	Value for set.seed, default = FALSE
parallel	TRUE or FALSE
impute	Deterministic imputation method of type "none", "norm.predict", "pmm", "mean" ; More Details? See mice
center_mods	TRUE or FALSE
resint	TRUE or FALSE

Details

In addition to analysing bootstrapped imputed (survey) data, analysis of the simple slopes as proposed by Aiken and West (1991), regions of significance as proposed by Bauer and Curran (2005) and (moderated) mediation analysis as proposed by Preacher, Rucker, and Hayes (2007) are also implemented for linear models (lm), multilevel models (lmer) and mice models (mice).

For faster computation, parallel is implemented.

Value

object of class "bootmi" including

• $formula, formula used

• $data, original data

• $bootstraps, bootstap data

• $replics, bootstrap replics

• $imputation, imputation method

• $seed, seed value

• $parallel, use parallel TRUE or FALSE

• $center_mods, center moderators? TRUE or FALSE

Functions

The main functions are:

bootmi()	Bootstrap and impute the missing data *R* times
lm()	Analyze bootmi data
simslop()	Conducts analysis of the simple slopes
regosi()	Calculates regions of significance
mediate()	Tests (moderated) mediation. (under construction)

Author(s)

Stephan Volpers [e-mail]

References

Bauer, Daniel J.; Curran, Patrick J. (2005): Probing Interactions in Fixed and Multilevel Regression: Inferential and Graphical Techniques. In: Multivariate Behavioral Research 40 (3), S. 373-400.

Buuren, Stef van; Groothuis-Oudshoorn, Karin (2011): mice. Multivariate Imputation by Chained Equations in R. In: Journal of Statistical Software 45 (3).

Hippel, Paul T. von (2009): How to Impute Interactions, Squares, and other Transformed Variables. In: Sociological Methodology 39 (1), S. 265-291.

Little, Todd D.; Bovaird, James A.; Widaman, Keith F. (2006): On the merits of orthogonalizing powered and product terms: Implications for modeling interactions among latent variables. In: Structural Equation Modeling 13 (4), S. 497-519.

Preacher, Kristopher J.; Rucker, Derek D.; Hayes, Andrew F. (2007): Addressing Moderated Mediation Hypotheses: Theory, Methods, and Prescriptions. In: Multivariate Behavioral Research 42 (1), S. 185-227.

Shao, Jun; Sitter, Randy R. (1996): Bootstrap for Imputed Survey Data. In: Journal of the American Statistical Association 91 (435), S. 1278-1288.

svolpers/bootmi documentation built on June 23, 2021, 6:36 a.m.