Skip to main content

Advertisement

Table 6 Results (p values of tests) of application to two real datasets.

From: Testing for differences in distribution tails to test for differences in 'maximum' lifespan

Test Data from Vasselli et al. [10]1 Data from Redmann & Argyropoulos [14]2
Tests of H 0,A (Wang-Allison tests)   
QT3 with τ set to 110/100#. 0.002 0.027
QT3 with τ set to sample 90th percentile. 0.038 0.186
QT4 with τ set to 110/100. 0.002 0.022
QT4 with τ set to sample 90th percentile. 0.033 0.146
Tests of H 0, B   
Wilcoxon-Mann-Whitney with τ set to 110/100. 0.289 0.868
Wilcoxon-Mann-Whitney with τ set to sample 90th percentile. 0.750 N/A*
Permutation test with τ set to 110/100. 0.281 0.738
Permutation test with τ set to sample 90th percentile. 0.634 N/A*
Tests of H 0, C (new tests)   
Wilcoxon-Mann-Whitney with τ set to 110/100. 0.001 0.022
Wilcoxon-Mann-Whitney with τ set to sample 90th percentile. 0.026 0.243
Permutation test with τ set to 110/100. 0.001 0.014
Permutation test with τ set to sample 90th percentile. 0.024 0.072
  1. Notes: In each dataset, males and females have been combined. 1For the data from Vasselli et al. [10] two groups of rats (EO-HF and WL-HF) are compared; each group has 49 observations.
  2. 2The data from Redmann & Argyropoulos [14] consists of 16 mice with genotype '+/+' and 21 mice with genotype '-/-'.
  3. # For Data from Vasselli et al. [10]τ is set to 110; for data from Redmann & Argyropoulos [14]τ is set to 100.
  4. *Only one group has observations above the threshold τ.