Table 3.
mSRI |
||||||
---|---|---|---|---|---|---|
Sample sizes | CMI | B = 20 | B = 40 | B = 60 | B = 80 | B = 100 |
n1 = 175; n2 = 450 | 0.0018 | 0.0418 | 0.0860 | 0.1258 | 0.1637 | 0.2153 |
n1 = 350; n2 = 900 | 0.0023 | 0.0467 | 0.0908 | 0.1390 | 0.1833 | 0.2249 |
mSRI: multiple stochastic regression imputation; CMI: conditional mean imputation.
Times are reported in seconds as the mean across 1000 evaluations using the microbenchmark() function in R. Results are based on application with a 2.9 GHz Intel Core i7 processor and the conditions described in the simulation scenarios of Figure 2.