Skip to main content
. 2013 Nov 19;8(11):e78329. doi: 10.1371/journal.pone.0078329

Table 1. Results of log-linear analysis to detect interactions between 1) Population (5: Butihondo, La Oliva, Caleta de Famara, Nazaret-Teguise and Yaiza), 2) Tail state (Tail, 2: original or regenerated), 3) Trial (5 trials per individual), and 4) Refuge Side (3: Left, Neutral, Right).

Partial association Marginal Association
df Chi-square p Chi-square P
Population 4 <0.0001 1.000 <0.0001 1.000
Tail 1 11.141 0.001 11.141 0.001
Race 4 <0.0001 1.000 <0.0001 1.000
Side 2 35.000 0.000 35.000 0.000
Population*Tail 4 18.843 0.001 17.630 0.001
Population*Trial 16 0.238 1.000 <0.0001 1.000
Population*Side 8 20.310 0.009 18.872 0.016
Tail*Trial 4 0.024 1.000 <0.0001 1.000
Tail*Side 2 1.836 0.399 0.613 0.736
Trial*Side 8 7.414 0.493 7.166 0.519
Population*Tail*Trial 16 0.863 1.000 <0.0001 1.000
Population*Tail*Side 8 12.498 0.130 12.432 0.133
Population*Trial*Side 32 16.248 0.991 18.116 0.977
Tail*Trial*Side 8 7.235 0.512 8.819 0.358

Model: Population*Side; Chi-square135 = 89.097, p = 0.999. (p values lower than 0.05 are marked in bold).