Model | Description | df ^{ a } | SSWR ^{ b } | F Score ^{ c } | Critical F Value ^{ d } | Significance ^{ e } | AIC Value ^{ f } |
---|---|---|---|---|---|---|---|
A | No nonlinear pathway | 189 | 653 | NA ^{ g } | NA | NA | 3297 |
B | Nonlinear only | 188 | 837 | −41.3 | No | 3370 | |
C | Nonlinear only | 188 | 808 | −36.0 | No | 3322 | |
D | Nonlinear only | 188 | 695 | −11.4 | No | 3293 | |
E | Nonlinear and | 187 | 287 | 119 | Yes | 3244 | |
F | Nonlinear , , and | 186 | 250 | 100 | Yes | 3235 | |
G | Nonlinear , , and | 186 | 234 | 111 | Yes | 3225 | |
H | Nonlinear , , , and | 185 | 71.4 | 377 | Yes | 3214 |
↵ ^{a} The df is the number of data points (n = 198) used in the model minus the number of parameters being fitted.
↵ ^{b} The SSWR was calculated using . The total SSWR for verapamil, preformed, and formed norverapamil data in perfusate, bile, and liver compartments was calculated.
↵ ^{c} The F score was calculated using , where df _{j} > df _{i} .
↵ ^{d} The critical F value was obtained from the F table with a numerator of df _{j} –df _{i} and denominator of df _{i} .
↵ ^{e} The significance level was α = 0.05; Models E–H showed significant improvement in the goodness-of-fit compared with Model A.
↵ ^{f} The AIC, a measurement of the goodness-of-fit, is calculated as , where k is the number of parameters fitted in the model, and likelihood is the maximized value of the likelihood function.
↵ ^{g} Not available