Table 4 Evaluation criteria for parameter estimations across replicate data sets. \(\Theta \in \mathbb {R}\) denotes the true value of the parameter of interest, \(\hat {\Theta } \in \mathbb {R}^{m}\) the parameter estimations for the m replicate data sets
From: Analysis of contamination in cluster randomized trials of malaria interventions
Evaluation criteria | Abbreviation | Formula |
|---|---|---|
Relative bias | relBias | \((\mathrm {E}[\hat {\Theta }] - \Theta)/\Theta \) |
Empirical standard error | EmpSE | \(\text {Var}[\hat {\Theta }]^{1/2}\) |
Width of 95%CI | Width | Average width of 95%CI |
Coverage probability | CP | Proportion of 95%CI that contained Θ |