Table 8.
Pairwise corrected resampled t test of sensitivity differences using three sets of features as predictors.
Pair | Description | Mean difference (%) | 95% CI of the mean difference | P value (two-tailed) |
1 | 19 features versus 115 features | −7.4336 | −0.1699 to 0.0681 | .13 |
2 | 20 features versus 115 features | 12.9204 | −0.016 to 0.1929 | .011a |
3 | 20 features versus 19 features | 21.9885 | 0.1048 to 0.174 | <.001a |
aP value significant at .0167 (Bonferroni correction).