Table 5.
Predicted treatment probabilities (propensity scores) and stabilized survey weights for the illustrative example and data-generating mechanisms with and without practical positivity violations from the scenario with a constant treatment effect. The minimum, maximum, mean and standard deviation (SD) are given.
Propensity Scores | Survey Weights (stabilized) | |||||
---|---|---|---|---|---|---|
Analysis | Mean (SD) | Minimum | Maximum | Mean (SD) | Minimum | Maximum |
No positivity violations, simulation |
0.45 (0.10) | 0.10 | 0.80 | 1.00 (1.58) | 0.14 | 19.94 |
Positivity violations, simulation |
0.41 (0.20) | 0.003 | 0.99 | 1.00 (1.58) | 0.14 | 19.94 |
Illustrative Example | 0.30 (0.24) | 0.03 | 0.92 | 1.00 (1.22) | 0.04 | 13.23 |