Table 3.

Summary statistics for the Generalization to novel stimuli across animals and the four conditions (Fig. 4).

Animal	Task	Sound	Trials	Hit Rate	Binomial test on performance (Fig. 4c)
					N	Degrees of freedom	Adjusted p-value
a	Twitter vs Tone	Tone	530	0.93	529	1	3.75E−100
	Phee vs Tone	Tone	186	0.83	183	1	1.78E−20
	Juvenile vs Twitter	Juvenile	333	0.6	333	1	8.72E−03
	Phee vs Tone	Phee	185	0.84	184	1	1.94E−21
	Twitter vs Tone	Twitter	535	0.84	528	1	5.05E−61
	Juvenile vs Twitter	Twitter	335	0.61	334	1	4.65E–04
	Noise vs Twitter	Twitter	54	0.57	54	1	1
	Noise vs Twitter	Noise	54	0.67	54	1	3.97E–01
b	Twitter vs Tone	Tone	200	0.81	198	1	5.52E–19
	Phee vs Tone	Tone	320	0.66	315	1	1.59E–08
	Juvenile vs Twitter	Juvenile	218	0.44	217	1	1
	Phee vs Tone	Phee	316	0.67	314	1	1
	Twitter vs Tone	Twitter	191	0.88	190	1	9.98E–28
	Juvenile vs Twitter	Twitter	214	0.7	214	1	7.76E–08
	Noise vs Twitter	Twitter	150	0.85	149	1	7.60E–18
	Noise vs Twitter	Noise	154	0.36	146	1	1
d	Twitter vs Tone	Tone	389	0.93	389	1	7.86E–76
	Phee vs Tone	Tone	328	0.87	324	1	3.57E–45
	Juvenile vs Twitter	Juvenile	40	0.63	33	1	9.10E–02
	Phee vs Tone	Phee	336	0.9	335	1	2.74E–53
	Twitter vs Tone	Twitter	386	0.91	386	1	1.87E–66
	Juvenile vs Twitter	Twitter	33	0.55	33	1	1
	Noise vs Twitter	Twitter	129	0.87	129	1	4.89E–17
	Noise vs Twitter	Noise	130	0.9	129	1	1.71E–21
i	Twitter vs Tone	Tone	129	0.79	128	1	1.56E–10
	Phee vs Tone*	Tone	9	0.89	9	1	7.81E–01
	Juvenile vs Twitter*	Juvenile	1	1	1	1	1
	Phee vs Tone*	Phee	10	0.7	10	1	1
	Twitter vs Tone	Twitter	120	0.8	119	1	1.71E–10
	Juvenile vs Twitter*	Twitter	4	0.5	4	1	1
	Noise vs Twitter	Twitter	142	0.76	141	1	3.37E–09
	Noise vs Twitter	Noise	144	0.88	143	1	2.50E–21
k	Twitter vs Tone	Tone	590	0.79	588	1	7.60E–46
	Phee vs Tone	Tone	597	0.88	596	1	5.35E–87
	Juvenile vs Twitter	Juvenile	751	0.71	749	1	2.69E–29
	Phee vs Tone	Phee	632	0.82	629	1	4.03E–63
	Twitter vs Tone	Twitter	601	0.79	598	1	5.14E–49
	Juvenile vs Twitter	Twitter	787	0.7	785	1	1.25E–28
	Noise vs Twitter	Twitter	541	0.64	532	1	7.54E–11
	Noise vs Twitter	Noise	624	0.47	615	1	1

Columns under “ Binomial test on performance (Fig. 4c)” report information regarding the statistical deviations of performance (across stimuli and task type) from a theoretically expected distribution of observations (one-sided), with p-values adjusted with a post-hoc Bonferroni correction for multiple comparisons. Significant values are indicated in bold font. Stars behind the comparison in the column "Task" indicate low trial numbers performed suggesting a low statistical power.