Skip to main content
. 2015 Aug 7;282(1812):20150934. doi: 10.1098/rspb.2015.0934

Table 3.

Testing: (a) results of the four-way MANOVA on the change in lines crossed and the change in shelter use, testing for the effects of risk (low versus high), the conditioning cue (PO versus NO), the testing cue (conditioning odour versus NO), and the species used for the conditioning cue (pike versus sturgeon). (b) Results of a three-way MANOVA for observers conditioned with low-risk models, and for (c) observers conditioned with high-risk models. Significant differences are in italics.

F d.f. p-value
(a) overall MANOVA
 risk 1.37 2, 189 0.26
 conditioning cue 1.63 2, 189 0.20
 testing cue 2.36 2, 189 0.10
 species 0.26 2, 189 0.77
 risk × conditioning cue 5.56 2, 189 0.005
 risk × testing cue 3.01 2, 189 0.052
 conditioning cue × testing cue 3.87 2, 189 0.023
 risk × conditioning cue × testing cue 3.11 2, 189 0.047
(b) low-risk MANOVA
 conditioning cue 7.6 2, 93 0.001
 testing cue 6.11 2, 93 0.003
 species 0.85 2, 93 0.43
 conditioning cue × testing cue 8.7 2, 93 <0.001
(c) high-risk MANOVA
 conditioning cue 0.51 2, 94 0.60
 testing cue 1.1 2, 94 0.34
 species 0.06 2, 94 0.94
 conditioning cue × testing cue 0.13 2, 94 0.88