RSE: Compute the relative standard error ratio

In SimDesign: Structure for Organizing Monte Carlo Simulation Designs

Compute the relative standard error ratio

Description

Computes the relative standard error ratio given the set of estimated standard errors (SE) and the deviation across the R simulation replications (SD). The ratio is formed by finding the expectation of the SE terms, and compares this expectation to the general variability of their respective parameter estimates across the R replications (ratio should equal 1). This is used to roughly evaluate whether the SEs being advertised by a given estimation method matches the sampling variability of the respective estimates across samples.

Usage

RSE(SE, ests, unname = FALSE)

Arguments

SE	a numeric matrix of SE estimates across the replications (extracted from the results object in the Summarise step). Alternatively, can be a vector containing the mean of the SE estimates across the R simulation replications
ests	a numeric matrix object containing the parameter estimates under investigation found within the Summarise function. This input is used to compute the standard deviation/variance estimates for each column to evaluate how well the expected SE matches the standard deviation
unname	logical; apply unname to the results to remove any variable names?

Value

returns vector of variance ratios, (RSV = SE^2/SD^2)

Author(s)

Phil Chalmers [e-mail]

References

Chalmers, R. P., & Adkins, M. C. (2020). Writing Effective and Reliable Monte Carlo Simulations with the SimDesign Package. The Quantitative Methods for Psychology, 16(4), 248-280. \Sexpr[results=rd]{tools:::Rd_expr_doi("10.20982/tqmp.16.4.p248")}

Sigal, M. J., & Chalmers, R. P. (2016). Play it again: Teaching statistics with Monte Carlo simulation. Journal of Statistics Education, 24(3), 136-156. \Sexpr[results=rd]{tools:::Rd_expr_doi("10.1080/10691898.2016.1246953")}

Examples

R <- 10000
par_ests <- cbind(rnorm(R), rnorm(R, sd=1/10),
                  rnorm(R, sd=1/15))
colnames(par_ests) <- paste0("par", 1:3)
(SDs <- colSDs(par_ests))

SEs <- cbind(1 + rnorm(R, sd=.01),
             1/10 + + rnorm(R, sd=.01),
             1/15 + rnorm(R, sd=.01))
(E_SEs <- colMeans(SEs))
RSE(SEs, par_ests)

# equivalent to the form
colMeans(SEs) / SDs

SimDesign documentation built on Sept. 11, 2024, 8 p.m.