Confidence intervals for a copreg model

Confidence intervals for the augmented-regression coefficients, either from the normal approximation using the bootstrap standard errors, or as bootstrap percentile intervals taken directly from the resampled coefficients.

Usage

# S3 method for class 'copreg'
confint(object, parm, level = 0.95, type = c("normal", "percentile"), ...)

Arguments

object

a fitted model of class "copreg".

parm

which parameters to report; a vector of names, or missing for all of them.

level

the confidence level, default 0.95.

type

"normal" (the default), for coef +/- qnorm(...) * std.error, or "percentile", for quantiles of the bootstrap draws in object$boot.

...

currently unused.

Value

A matrix with one row per requested parameter and two columns giving the lower and upper confidence limits.

References

Qian, Y., A. Koschmann, and H. Xie (2025). A practical guide to endogeneity correction using copulas. Journal of Marketing.

Examples

set.seed(1)
n  <- 150
w  <- rnorm(n)
p  <- 0.4 * w + rt(n, df = 3)
xi <- 0.5 * p + rnorm(n)
y  <- 1 + 2 * p + 1.5 * w + xi
dat <- data.frame(y = y, p = p, w = w)

fit <- endogCopula:::.copreg_fit(
  formula = y ~ p | w, data = dat,
  ctor = endogCopula:::.ctor_twostage(TRUE),
  method = "2sCOPE", cdf = "rank.n", ties = "max",
  nboots = 25, verbose = FALSE)

confint(fit, level = 0.90, type = "normal")
#>                    5 %       95 %
#> (Intercept)  0.7276542 1.02755219
#> p            2.4644183 3.21299074
#> w            1.2388654 1.61980332
#> p_cop       -1.1081964 0.08979654
confint(fit, type = "percentile")
#>                  2.5 %    97.5 %
#> (Intercept)  0.6641320 0.9852160
#> p            2.2179720 2.9440920
#> w            1.3286318 1.7592068
#> p_cop       -0.8223605 0.4500005