Table 2.

The role of social factors in explaining variation in co-representation

Species	Fixed factor	β	SE	95% CI	OR	z	p
(a) Tonkean	Intercept	1.05	0.43
macaques	Compatibility	− 2.75	0.35	− 3.43, − 2.07	0.06	− 7.87	3.53 × 10–15***
	Absolute rank difference	0.07	0.06	− 0.04, 0.19	1.07	1.22	0.22
	Eigenvector centrality	-0.32	0.59	− 1.47, 0.83	0.73	− 0.55	0.58
	DGI	0.09	0.20	− 0.29, 0.48	1.09	0.48	0.63
	DSI	− 0.01	0.10	− 0.20, 0.19	0.99	− 0.08	0.94
				N = 7 individuals, ${\chi }_4^2$  = 2.64, p = 0.62, ∆AIC = 5.36
(b) Brown	Intercept	1.03	0.31
capuchins	Compatibility	− 1.98	0.33	− 2.63, − 1.33	0.14	− 5.95	2.66 × 10–9***
	Absolute rank difference	0.03	0.08	− 0.12, 0.18	1.03	0.36	0.72
				N = 6 individuals, ${\chi }_1^2$ = 0.13, p = 0.72, ∆AIC = 1.87

The effect of stimulus compatibility, absolute rank difference, eigenvector centrality (i.e. integration in the affiliative social network), DGI and DSI (i.e. bond strength) with a given partner on an individual’s strength of co-representation (i.e. joint Simon effect; errors in incompatible vs. compatible trials) in (a) the Tonkean macaques (n = 7) and (b) the brown capuchins (n = 6). Parameter estimates, standard errors, 95% confidence intervals (CI), odds ratios (OR) and statistical significance are obtained from generalized linear mixed effect models. Significant effects are indicated with p-values in italics