Table 1 Estimates of the marginal mean of the normally distributed outcome variable and the regression coefficient under four analysis methods for a single simulated dataset (n = 500, m = 300, δ = 0.2)
From: Evaluation of a weighting approach for performing sensitivity analysis after multiple imputation
|  | μ | β | ||
|---|---|---|---|---|
| Â | Parameter estimate | SE | Parameter estimate | SE |
Full dataset (before deletion) | −0.007 | 0.044 | 0.501 | 0.039 |
Complete Case Analysis | 0.282 | 0.057 | 0.426 | 0.058 |
Multiple Imputation under MAR | 0.122 | 0.058 | 0.427 | 0.058 |
Sensitivity analysis under MNAR | 0.022 | 0.045 | 0.518 | 0.046 |