Table 1.

Conditioning: (a) results of the four-way repeated-measures ANOVA on the change in lines crossed, testing for the effects of risk (low versus high), the conditioning cue (PO versus NO), the fish role (model versus observer, as repeated measures) and the species used for the cue (pike versus sturgeon). (b) Results of a three-way repeated measures ANOVA for social conditionings involving low-risk models and for (c) those involving high-risk models. Significant differences are in italics.

	F	d.f.	p-value
(a) overall RM ANOVA
within subjects
role	22.38	1, 187	<0.001
role × risk	4.58	1, 187	0.034
role × cue	0.28	1, 187	0.60
role × species	0.01	1, 187	0.93
role × risk × cue	2.85	1, 187	0.09
between subjects
risk	<0.01	1, 187	0.96
cue	12.62	1, 187	<0.001
species	0.24	1, 187	0.62
risk × cue	26.67	1, 187	<0.001
(b) low-risk RM ANOVA
within subjects
role	3.40	1, 90	0.07
role × cue	2.23	1, 90	0.14
role × species	0.59	1, 90	0.44
between subjects
cue	37.78	1, 90	<0.001
species	0.69	1, 90	0.41
(c) high-risk RM ANOVA
within subjects
role	25.92	1, 96	<0.001
role × cue	0.78	1, 96	0.38
role × species	0.46	1, 96	0.50
between subjects
cue	1.29	1, 96	0.26
species	0.01	1, 96	0.92