Table 2.

Model selection for response variables from generalized linear mixed models

Variable	Random factor	Model	AICc	ΔAICc	Relative likelihood	Akaike weight
Scan flights	(1|dyad/session/phase/min)	Familiarity pairing* Experimental phase (Full model)	933.120	1.981	0.371	0.227
		Familiarity pairing+Experimental phase	934.735	3.596	0.166	0.101
		Familiarity pairing	940.415	9.276	0.010	0.006
		Experimental phase	931.139	0	1.000	0.610
		Intercept only	935.894	4.755	0.093	0.057
Number of birds perched in trees near loudspeakers	(1|dyad/session/phase/min)	Familiarity pairing* Experimental phase (Full model)	875.822	4.466	0.107	0.052
		Familiarity pairing+Experimental phase	872.553	1.642	0.440	0.215
		Familiarity pairing	870.807	0	1.000	0.488
		Experimental phase	873.237	2.574	0.276	0.135
		Intercept only	873.584	2.962	0.227	0.111
‘Haa’ calls	(1|dyad/session/phase/min)	Familiarity pairing* Experimental phase (Full model)	838.646	11.086	0.004	0.002
		Familiarity pairing+Experimental phase	832.241	5.126	0.077	0.047
		Familiarity pairing	830.137	3.126	0.210	0.127
		Experimental phase	828.893	2.026	0.363	0.220
		Intercept only	826.826	0	1.000	0.605
Defensive calls	(1|dyad/session/phase/min)	Familiarity pairing* Experimental phase (Full model)	1191.620	10.674	0.005	0.004
		Familiarity pairing+Experimental phase	1187.525	6.024	0.049	0.037
		Familiarity pairing	1185.683	4.286	0.117	0.088
		Experimental phase	1184.991	3.738	0.154	0.116
		Intercept only	1181.212	0	1.000	0.754

Bold type indicates the best models, which were determined based on relative AICc values (ΔAICc) and computed relative likelihood and Akaike weights.