Skip to contents

Test comparing stationary and non-stationary doubly robust DiD estimators

Source: R/stationarity-test.R
drdid_stationarity_test.Rd

This function compares the ATT estimates from drdid_nonstationary() and drdid_stationary() using the empirical second moment of the difference in their estimated influence functions. Parametric inputs are reported as a Wald equality test; other inputs retain the existing Hausman-type label.

Usage

drdid_stationarity_test(
  nonstationary_fit,
  stationary_fit,
  alpha = c(0.1, 0.05, 0.01),
  check_stabilized = TRUE
)

Arguments

nonstationary_fit

Output from drdid_nonstationary() with inffunc = TRUE.

stationary_fit

Output from drdid_stationary() with inffunc = TRUE.

alpha

Numeric vector of significance levels.

check_stabilized

Logical; if TRUE, require the two inputs to use the same stabilized setting.

Value

A list with the Wald statistic, p-value, critical values, and test decisions at each requested significance level.