Table 1.

Influence of 2-month competitive ability and 1-month winning/losing experience on 1-day winner/loser effect

		Initiating gill displays (n = 447)			Initiating attacks (n = 432)			Winning non-escalated (n = 272)
Variable	df	b±SE	χ2	P	b±SE	χ2	P	b±SE	χ2	P
2 month-CA(worse)	1	-0.10±0.20	0.23	0.631	-0.02±0.21	0.01	0.905	0.23±0.31	0.55	0.459
1 month-W/L(losing)	1	-0.29±0.20	2.16	0.142	-0.38±0.21	3.47	0.063	0.01±0.29	0.00	0.972
1 day-W/L(losing)	1	-0.70±0.20	12.91	<0.001*	-0.87±0.21	18.51	<0.001*	-1.98±0.29	51.84	<0.001*
2 month-CA × 1 d-W/L	1	0.11±0.39	0.08	0.772	-0.03±0.41	0.01	0.935	0.88±0.61	2.10	0.147
1 month-W/L × 1 d-W/L	1	-1.25±0.39	10.15	0.001*	-2.06±0.42	25.69	<0.001*	-1.74±0.58	9.15	0.003*
Size	1	0.06±0.04	1.82	0.177	0.01±0.04	0.02	0.902	-0.01±0.07	0.05	0.826
Strain	4		6.50	0.165		0.31	0.989		3.65	0.456

Multiple logistic regression (likelihood ratio χ² statistic) modeling the influence of competitive ability measured two months previously (2 month-CA; the pairs comprising better competitors were the baseline group), contest experience received one month previously (1 month-W/L; the pairs where both contestants received a winning experience were the baseline group) and contest experience received one day previously (1 day-W/L; the W-N pairs where the focal individuals received a winning experience were the baseline group) on the contest behaviors of the focal individuals when fighting against the control opponents, controlling for the body size and the strain type of the contest pairs.