double_bootstrap: Double bootstrap procedure

In adzemski/netprobitFE: Network probit with fixed effects

Description

Double bootstrap procedure

Usage

double_bootstrap(wide, rho_bs0, model, bs_iterations = 200,
  alpha = 0.1, trimming_bound = 0, bias_estimator = "theorem2",
  force_robust = FALSE)

Arguments

wide	link data in wide format
rho_bs0	within-dyad correlation of bootstrap distribution
model	model description in special list format
bs_iterations	number of bootstrap replications
alpha	significance level
trimming_bound	bound for trimming small probabilities
bias_estimator	estimator of bias (see triangle test)
force_robust	logical, force using numerically stable (but possibly slow) algorithm to fit probit

Value

list with bootstrapped SE and quantiles

adzemski/netprobitFE documentation built on May 17, 2019, 11:40 a.m.