. 2022 Mar 19;55(6):1584–1600. doi: 10.1111/ejn.15646

TABLE 1.

Selection between linear and nonlinear models among different sleep stages for each cortex

Cortex	No. of cats	Model	BF
Right dorsolateral prefrontal	2	Linear
Right dorsolateral prefrontal		Quadratic *	6.88 × 10¹⁸
Right primary motor	2	Linear
Right primary motor		Quadratic *	1.90 × 10¹⁷
Right primary auditory	1	Linear
Right primary auditory		Quadratic *	4.42 × 10⁶
Right primary somatosensory	2	Linear
Right primary somatosensory		Quadratic	0.681
Left primary somatosensory	1	Linear
Left primary somatosensory		Quadratic *	8.53 × 10²⁰
Right posterior parietal	3	Linear
Right posterior parietal		Quadratic *	1.76 × 10⁷
Left posterior parietal	1	Linear
Left posterior parietal		Quadratic *	7.19 × 10⁹
Right primary visual	3	Linear
Right primary visual		Quadratic *	3.70 × 10¹¹
Left primary visual	1	Linear
Left primary visual		Quadratic *	3.77 × 10¹⁷
Right lateral geniculate	1	Linear
Right lateral geniculate		Quadratic *	7.00 × 10²⁰

Note: Mixed effects models were formulated for each cortex including the cat as a random effect when applicable. Bayes factors (BF) were used to decide between U‐shaped and linear fits. With the exception of the right primary somatosensory cortex, all model comparisons showed the supremacy of the quadratic fit. The asterisks indicate substantial evidence for a quadratic fit (BF > 5).