Table 2

Residua and statistical comparison of the residua of regression models

Regression modelRegression residuum (ms)Optimum cases (all/women/men)
The top part of the table shows the mean regression residua obtained with the QT/RR regression models, each individually optimised for each subject. For each regression model, the top part also shows the number of subjects for whom the particular model was the optimum among all regression models considered (that is, leading the lowest residuum in the given subject). The bottom part of the table shows p values of the paired comparisons (Wilcoxon test) of the regression residua summarised in the top part of the table.
(A) Linear11.08 (1.98)20 / 8 / 12
(B) Hyperbolic11.96 (2.15)0 / 0 / 0
(C) Parabolic11.14 (1.99)5 / 4 / 1
(D) Logarithmic11.27 (2.01)6 / 3 / 3
(E) Shifted logarithmic11.07 (1.98)12 / 5 / 7
(F) Exponential11.18 (2.00)7 / 5 / 2
ModelABCDE
B8.58 × 10−9
C0.08481.48 × 10−9
D0.0003127.65 × 10−101.82 × 10−7
E0.3825.11 × 10−90.01043.12 × 10−5
F0.01608.632 × 10−100.0001571.51 × 10−80.00232