Skip to contents

Apply Selective Correction to Flagged Parameters

Source: R/der_correct.R
der_correct.Rd

Rescales the posterior draws of parameters flagged by der_classify() so that their covariance matches the sandwich variance target, leaving unflagged parameters untouched. The correction preserves posterior means.

Usage

der_correct(x, method = c("block_cholesky", "marginal"))

Arguments

x

A svyder object with classification (from der_classify()).

method

"block_cholesky" (default) or "marginal".

Value

A svyder object with corrected_draws, scale_factors (marginal SD ratios of flagged parameters), and correction_method populated.

Details

With method = "block_cholesky" (the paper's Algorithm 1), the flagged block F is transformed jointly: writing \(\Sigma_F = L_1 L_1^\top\) and \(V_F = L_2 L_2^\top\) for the MCMC and sandwich covariance of the flagged block, each draw is mapped through \(\phi^* = \hat\phi + L_2 L_1^{-1} (\phi - \hat\phi)\), so the corrected block has covariance exactly \(V_F\). With a single flagged parameter this reduces to scaling by \(\sqrt{V_{ii}/\Sigma_{ii}}\).

With method = "marginal", each flagged parameter is rescaled independently by \(\sqrt{V_{ii}/\Sigma_{ii}}\). This matches marginal variances only and is provided for transparency and for reproducing results computed this way; it is not a Cholesky correction when more than one parameter is flagged.

References

Lee, J., Williams, M. R., & Savitsky, T. D. (2026). Design Effect Ratios for Bayesian Survey Models: A Diagnostic Framework for Identifying Survey-Sensitive Parameters. Journal of Survey Statistics and Methodology. Submitted.

Savitsky, T. D., & Williams, M. R. (2022). Pseudo Bayesian mixed models under informative sampling. Journal of Official Statistics, 38(4), 901–928. doi:10.2478/jos-2022-0039

Williams, M. R., & Savitsky, T. D. (2021). Uncertainty estimation for pseudo-Bayesian inference under complex sampling. International Statistical Review, 89(1), 72–107. doi:10.1111/insr.12376

Higham, N. J. (2002). Computing the nearest correlation matrix—a problem from finance. IMA Journal of Numerical Analysis, 22(3), 329–343. doi:10.1093/imanum/22.3.329

See also

der_classify() for flagging parameters, der_compute() for computing DER, as.matrix.svyder() for extracting corrected draws.

Other core-pipeline: der_classify(), der_compute(), der_diagnose()

Examples

data(nsece_demo)
result <- der_compute(
  nsece_demo$draws,
  y = nsece_demo$y, X = nsece_demo$X,
  group = nsece_demo$group, weights = nsece_demo$weights,
  cluster = nsece_demo$psu, family = "binomial",
  sigma_theta = nsece_demo$sigma_theta,
  param_types = nsece_demo$param_types
)
result <- der_classify(result, tau = 1.2, verbose = FALSE)
result <- der_correct(result)