IPW (MLE)
Empirical coverage of the inverse probability weighting estimator with the propensity score fit by maximum likelihood. Spike scope: one DGP (default), logit link, analytical variance.
nonprobsvy version: 0.3.0 | R: 4.6.0 | run: 2026-05-24 07:58:26 | commit: 4b2ba9a
| var_method | num_boot | alpha | n_reps | mean_se | coverage | ci_width |
|---|---|---|---|---|---|---|
| analytic | NA | 0.05 | 500 | 0.108 | 0.932 | 0.425 |
| analytic | NA | 0.10 | 500 | 0.108 | 0.890 | 0.356 |
| bootstrap | 200 | 0.05 | 500 | 0.112 | 0.938 | 0.439 |
| bootstrap | 200 | 0.10 | 500 | 0.112 | 0.898 | 0.369 |
Notes
- DGP: linear outcome
y = 1 + x1 + 0.5 x2 - 0.3 x3 + N(0, 1)with logistic selection biased byx1andx2. True target is the population mean ofy. - The MC SE of the point estimate is reported next to each cell — it bounds how precisely we can pin down the empirical coverage.
- Bootstrap variants and the full estimator × DGP matrix will be added in subsequent passes.