mceBoot {MCE}R Documentation

Bootstrap-After-Bootstrap estimate of MCE

Description

Just like using the bootstrap to mimic multiple samples, the bootstrap can be used to mimic generating multiple sets of stochastic outcomes to estimate the MCE.

Usage

mceBoot(data, B, type = "", truth = NULL)

Arguments

data vector. replication reults
B numeric (integer). number of bootstrap replications to use.
type character. type of simulation result of interest including "mean", "SE", "PB" and "RE" for mean, standard error, percent bias and relative efficiency
truth numeric. if type="PB" the true value.

Details

1.
Draw a sample with replacement of length R from the replication results.
2.
Calculate and record the bootstrap result as a function of the sampled replication results to generate a simulation-based estimate.
3.
Repeat steps 1 and 2, B times.
4.
A standard deviation of these B simulation-based estimates can be used to obtain an estimate of the Monte Carlo error associated with R replications.

Value

bootstrap-after-boostrap estimate of MCE

Note

Unlike CLT methods, this method is not limited to simulation-based estimates calculated as a mean

Author(s)

Elizabeth Koehler and Sebastien Haneuse

See Also

mceCLT

Examples

test <- matrix(rnorm(10000, mean=1))
mceBoot(test, type="PB", B=500, truth=1)


[Package MCE version 1.0 Index]