Table 1.

Phylogenetic generalised linear mixed models (PGLMM) testing the influence of parity mode and egg attendance on the occurrence of social grouping in squamate reptiles

	Data set	Parameter	Estimate	Test statistic
Parity mode
	Conservative (n = 324)	Intercept (β0)	−2.01 ± 0.96	Z = −2.09, P = 0.04
		Parity mode (β1)	2.34 ± 0.42	Z = 5.59, P < 0.001
		Citation count	<0.001 ± 0.001	Z = 0.27, P = 0.79
		Signal in residuals (s2)	2.84	P < 0.001
		Signal in response (s2)	4.19	P < 0.001
	Relaxed (n = 1210)	Intercept (β0)	−3.25 ± 1.50	Z = −2.16, P = 0.03
		Parity mode (β1)	2.82 ± 0.49	Z = 5.70, P < 0.001
		Signal in residuals (s2)	7.28	P < 0.001
		Signal in response (s2)	8.92	P < 0.001
Egg attendance (analyses restricted to oviparous species)
	Conservative (n = 219)	Intercept (β0)	−1.61 ± 0.99	Z = 1.62, P = 0.11
		Egg attendance (β1)	−0.83 ± 0.54	Z = −1.52, P = 0.13
		Citation count	<0.001 ± 0.001	Z = −0.06, P = 0.95
		Signal in residuals (s2)	2.66	P < 0.001
		Signal in response (s2)	2.35	P < 0.001
	Relaxed (n = 1049)	Intercept (β0)	−3.89 ± 1.51	Z = 2.58, P = 0.01
		Egg attendance (β1)	1.04 ± 0.63	Z = 1.66, P = 0.10
		Signal in residuals (s2)	7.07	P < 0.001
		Signal in response (s2)	7.92	P < 0.001

N equals the number of species included in each analysis Significant terms are shown in bold The ‘signal in response (s2)’ parameter is derived from a model fit with no predictor variables and provides an estimate of phylogenetic structure of social grouping from each data set. Model estimates are reported ± SE. For confidence intervals of parameter estimates, see Supplementary Table 4. For a details on the conservative and relaxed data set, see Methods