Skip to contents

Computes standard errors by nonparametric bootstrap: resample individuals, re-estimate the GAMLSS, re-extract parameters, and re-compute the decomposition B times. Standard errors are the standard deviation across replicates. Percentile confidence intervals are also computed.

Usage

bootstrap_se(
  data,
  formula_mu,
  formula_sigma,
  treat,
  group,
  time = NULL,
  post = NULL,
  ref = NULL,
  ystat = "Var",
  estimand = "marginal",
  order = c("behavioral", "compositional", "pretreatment"),
  B = 100L,
  parallel = FALSE,
  ncores = NULL,
  seed = NULL,
  verbose = TRUE,
  cl_type = NULL,
  blend_params = NULL
)

Arguments

data

Data.frame, the original individual-level data

formula_mu

Two-sided formula for the mean (mu) equation

formula_sigma

One-sided formula for the log-SD (sigma) equation

treat

Character, treatment variable name

group

Character, grouping variable name

time

Character, time variable name. NULL for cross-sectional.

post

Character, pre/post indicator for DiD. NULL for simple diff.

ref

Numeric, reference time period for longitudinal decomposition

ystat

Character, "Var" or "CV2"

order

Character vector of length 3, decomposition ordering

B

Integer, number of bootstrap replicates. Default 100.

parallel

Logical, use parallel computation. Default FALSE.

ncores

Integer, number of cores. Default: all but one.

seed

Integer, random seed. Default NULL.

verbose

Logical, print progress. Default TRUE.

cl_type

Character, parallel cluster type, "fork" or "psock". Default NULL auto-selects (fork off Windows, psock on Windows). See boot_config for details.

Value

An object of class "ineqx_boot" containing:

se

List of SEs matching delta_method_se output structure

replicates

Matrix (B_successful x K) of replicate estimates

ci

List of 95% percentile confidence intervals

B

Total requested replicates

B_successful

Number of successful replicates

B_failed

Number of failed replicates

type

"cross" or "longit"

point_estimates

Named vector from original (non-bootstrap) decomposition

seed

The random seed used

Details

For longitudinal data with repeated cross-sections, resampling is performed within each time period independently, preserving the sample size per period.