Table A6.

Both the alternative models fit the data better than the EG model for the growth only subset of the data in the Dataset 4. We selected the data on B16 tumor growth with OT1 = 0 resulting in $n=30$ data points and fit the EG, Alt 1, and Alt 2 models (Equations (3), (8) and (9), respectively) to these data (see Figure 4B for model fits). We show the results of the Shapiro-Wilk (SW) normality test of the residuals. Other details are similar to those in Table 1.

Dataset 4    OT1 = 0    $\mathit{n}=30$
Models	$\alpha$	r	t’	d	$f_d$	SSR	AIC	$\Delta$	w	SW p
EG	2.22	0.5				2.24	13.4	12.8	0	0.46
Alt 1	2.48	1.13	8			1.37	0.6	0	0.59	0.6
Alt 2	1.79	3.12		1.03	0.95	1.3	1.3	0.7	0.41	0.43