Table 6.

Regression analysis

Variable name	Estimated coefficient	Standard error	T ratio (36 df)	P
Logit 1 [proportion left branch play | down at x1] Raw moment r2 = 0.6858
Constant	−0.25887	0.1360	−1.904	0.065
Trials, t	0.03673	0.01136	3.232	0.003
Game Form	−0.08076	0.2061	−0.3919	0.697
(Game Form) t	0.04291	0.0180	2.384	0.023
Logit 2 [proportion (50, 50) outcome | left at x2] Raw moment r2 = 0.7482
Constant	−0.027058	0.2257	−0.1199	0.905
Trials, t	0.055026	0.01726	3.188	0.003
Game Form	0.92697	0.2049	4.524	<0.001
Logit 3 [proportion (40, 40) outcome | right at x2] Raw moment r2 = 0.4610
Constant	0.32615	0.2292	1.423	0.163
Trials, t	0.02805	0.01850	1.125	0.268
Game Form	0.1438	0.2171	0.6596	0.514

Observations have been adjusted for heteroskedasticity.