Table 5. Results from full and minimum adequate model of a generalized linear mixed model Results from full and minimum adequate model of a generalized linear mixed model fitted on the effects of inferred social rank, cohort, the number of females housed with and trial number on binary spatial discrimination task performances for adult male pheasants tested while in the perceived dominance (PD) social condition. Random intercepts and fixed slopes model.

Trial and mean Elo-rating were standardised (z-scores). The table shows model estimates and standard (SE) for each variable with odds ratio (OR) with low (Lo CI) and high (Hi CI) confidence intervals.

Predictor variable	Estimate	SE	OR	Lo CI	Hi CI
Full model
Cohort	−0.219	0.294	0.803	−0.914	0.476
Correct on first trial	0.208	0.187	1.231	−0.234	0.650
Minimum Adequate model
Intercept	0.881	0.131
Trial	0.844	0.101	2.326	0.605	1.083
Mean Elo-rating	−0.144	0.108	0.866	−0.399	0.111
Female (4 females)	1.002	0.216	2.723	0.491	1.513
Trial * R Elo-rating	0.318	0.094	1.374	0.096	0.540