Performs a simple bootstrap of a fitted DirectEffects model by re-estimating the model with bootstrap samples.
Arguments
- seqg
A fitted sequential_g estimate, computed by
sequential_g
.- boots
The number of bootstrap replicates. Defaults to 1000.
Value
An object of type seqgboots
which is a matrix with boots
rows and columns for each coefficient in the seqg
model. Use summary
to provide summary statistics, such as mean and quantiles.
Examples
# \donttest{
data(ploughs)
form <- women_politics ~ plow +
agricultural_suitability + tropical_climate + large_animals + rugged |
years_civil_conflict + years_interstate_conflict + oil_pc +
european_descent + communist_dummy + polity2_2000 |
centered_ln_inc + centered_ln_incsq
s1 <- sequential_g(form, ploughs)
out.boots <- boots_g(s1)
summary(out.boots)
#>
#> Summary of bootstrapped coefficients:
#>
#> Estimate Std. Err. t value 2.5 % 97.5 %
#> (Intercept) 13.49540 4.59812 2.93498 4.50606 22.5845
#> plow -4.44703 1.81704 -2.44741 -8.17948 -1.0408
#> agricultural_suitability 4.27938 2.50275 1.70987 -0.71519 9.0480
#> tropical_climate -2.94928 2.37061 -1.24410 -7.54445 1.8279
#> large_animals -0.87980 4.08572 -0.21534 -8.90898 7.0103
#> rugged -0.36974 0.50283 -0.73532 -1.30592 0.6396
#>
# }