Pre-trend test for parallel trends — morie_did_test_parallel

Regresses the outcome on group-by-time interactions in the pre-period and reports both per-period coefficients and a joint Wald (chi-square) test that they are all zero.

Usage

morie_did_test_parallel_trends(
  data,
  outcome,
  treatment,
  time,
  unit = NULL,
  cluster = NULL,
  pre_periods = NULL
)

Arguments

data: A data frame.
outcome: Outcome column name.
treatment: Binary treatment-group indicator.
time: Time column (integer-valued).
unit: Optional unit identifier (currently unused; reserved).
cluster: Cluster variable for robust SE.
pre_periods: Optional explicit list of pre-treatment times.

Value

A list with coefficients, joint_chi2 (and its alias joint_f_stat), joint_df, joint_p_value, parallel_trends_plausible.

Examples

set.seed(1)
df <- expand.grid(unit = 1:30, time = 1:6)
df$treat <- as.integer(df$unit <= 15)
df$d <- as.integer(df$treat == 1L & df$time >= 4)
df$y <- 0.1 * df$time + 0.7 * df$d + rnorm(nrow(df), sd = 0.4)
res <- morie_did_test_parallel_trends(df, "y", "treat", "time",
                                      pre_periods = c(1, 2, 3))
res$parallel_trends_plausible
#> [1] TRUE