Difference in difference
We call 'Uncofoundedness' a scenario where a treatment is not randomly assigned to participants, so confounders effect on treatment assignment and outcome. We have client - level data. Confounders were measured before treatment and outcome after
Result
| unit_id | calendar_time | treated_time | y | region | market_traffic | avg_order_value | market_competition | macro_index | seasonality_index | y_cf | mu_cf | mu_treated | tau_mean_true | tau_realized_true | tau_rate_true | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 | control_market_001 | 2021-01 | 0 | 3302.404671 | central | 2.270322 | 43.653459 | 0.443270 | 0.918298 | 1.080000 | 3302.404671 | 3312.519731 | 3312.519731 | 0.0 | 0.0 | 0.0 |
| 1 | control_market_001 | 2021-02 | 0 | 3464.267194 | central | 2.246363 | 47.089670 | 0.453695 | 0.951317 | 1.120000 | 3464.267194 | 3317.006425 | 3317.006425 | 0.0 | 0.0 | 0.0 |
| 2 | control_market_001 | 2021-03 | 0 | 4912.564041 | central | 2.440752 | 45.544394 | 0.431990 | 0.901793 | 1.098564 | 4912.564041 | 3929.754343 | 3929.754343 | 0.0 | 0.0 | 0.0 |
| 3 | control_market_001 | 2021-04 | 0 | 4145.708011 | central | 2.694084 | 45.436522 | 0.402491 | 1.078746 | 1.080000 | 4145.708011 | 4043.013237 | 4043.013237 | 0.0 | 0.0 | 0.0 |
| 4 | control_market_001 | 2021-05 | 0 | 3723.949154 | central | 2.496220 | 44.289741 | 0.443995 | 0.955392 | 1.098564 | 3723.949154 | 3438.519375 | 3438.519375 | 0.0 | 0.0 | 0.0 |
| 5 | control_market_001 | 2021-06 | 0 | 3405.257515 | central | 2.504419 | 45.535514 | 0.447076 | 0.929668 | 1.120000 | 3405.257515 | 3581.296689 | 3581.296689 | 0.0 | 0.0 | 0.0 |
| 6 | control_market_001 | 2021-07 | 0 | 3663.077982 | central | 2.416596 | 45.304231 | 0.432838 | 0.918653 | 1.080000 | 3663.077982 | 3304.610262 | 3304.610262 | 0.0 | 0.0 | 0.0 |
| 7 | control_market_001 | 2021-08 | 0 | 2636.664688 | central | 2.012741 | 46.667927 | 0.441644 | 0.893138 | 0.960000 | 2636.664688 | 2554.059095 | 2554.059095 | 0.0 | 0.0 | 0.0 |
| 8 | control_market_001 | 2021-09 | 0 | 2364.397235 | central | 1.856487 | 45.459895 | 0.414938 | 1.016915 | 0.821436 | 2364.397235 | 2102.840509 | 2102.840509 | 0.0 | 0.0 | 0.0 |
| 9 | control_market_001 | 2021-10 | 0 | 2537.998976 | central | 1.988973 | 43.533768 | 0.401445 | 1.137315 | 0.760000 | 2537.998976 | 2148.905142 | 2148.905142 | 0.0 | 0.0 | 0.0 |
Inference
Result
| value | |
|---|---|
| field | |
| estimand | average_post_effect |
| model | CallawaySantAnnaStaggeredDID |
| estimator | dr |
| control_group | not_yet_or_never |
| anticipation | 0 |
| base_period | universal |
| include_pre_periods | False |
| value | 194.0136 (ci_abs: 112.9152, 275.1120) |
| std_error | 41.3775 |
| alpha | 0.0500 |
| p_value | 0.0000 |
| is_significant | True |
| n_att_gt_cells | 18 |
| n_skipped_cells | 0 |
| inference | clustered_influence |
| time | 2026-05-17 |
Oracle effect
Result
np.float64(127.67201792553853)