Skip to contents

Non-parametric Inference for SI (Agarwal et al. 2025)

Source: R/si.R
si_inference.Rd

Estimates SE and confidence intervals for the ATT by re-estimating the fit on resampled units: a bootstrap of the treated and the control units, drawn separately, or a jackknife that drops one unit at a time. Works for sharp, staggered and multi-arm SI fits. Each refit keeps the rank k the fit used.

Usage

si_inference(
  fit,
  method = c("bootstrap", "jackknife", "jackknife_global"),
  n_boot = 499L,
  level = 0.95,
  alternative = c("two.sided", "greater", "less"),
  seed = NULL
)

Arguments

fit

A coresynth object from scm_fit() with method = "si".

method

"bootstrap" (default), "jackknife", or "jackknife_global" (staggered fits only).

n_boot

Number of bootstrap replications (default 499L; ignored for jackknife).

level

Confidence level (default 0.95).

alternative

"two.sided" (default), "greater", or "less".

seed

RNG seed for reproducibility (default NULL).

Value

A list of class coresynth_inference.

Details

Holding the treated units fixed would leave out their own noise, which is most of the variance when few units are treated, so they are resampled too; with a single treated unit (in any arm) there is nothing to resample and the function stops. For a sharp fit with one treated unit, conformal_inference() gives a valid interval.

The controls are drawn once per replicate and shared by every arm and every cohort, as the SVD basis is a property of the shared donor pool. Treated units are drawn within their arm. On a staggered fit they are pooled across cohorts, so cohort sizes vary across replicates, and the cohorts are rebuilt from the draw and combined with their N_treated * T_post weights. "jackknife" drops each cohort's own donors and treated units and combines the per-cohort variances by the delta method, which ignores the covariance that shared donors induce between cohorts; "jackknife_global" (staggered fits only) drops every unit from the whole panel in turn and keeps it. Both jackknives drop treated units, so every cohort needs at least two (Clarke et al. 2024, Algorithm 3).