Table 3.

Results of the model that evaluates the probability of collisions in relation to different variables at the national scale.

Coefficients	Estimates	Std. Error	z-value	95% Conf. Int.		p-value [Pr (>|z|)]
				Lower	Upper
α0 (Intercept)	1.19	0.42	2.83	0.33	1.99	0.005
α1 (Reflection: Yes)	0.95	0.31	3.11	0.35	1.56	0.002
α2 (Natural_Attractor: Yes)	0.72	0.42	1.70	−0.09	1.59	0.089
α3 (Artificial_Attractor: Yes)	0.08	0.29	0.28	−0.49	0.65	0.780
α4 (Windows)	2.88	0.48	5.84	1.98	3.85	<0.001

We present the estimated value (Logit scale), standard error, z-value and their associated p-value, and the lower and upper limits of the 95% confidence interval. Significant p-values are highlighted in bold. α₀ is the baseline probability representing the situation of buildings with no reflection and no natural or artificial bird attractors.

α₁ is the regression coefficient that represents the effect of window reflection on the probability of collision; α₂ is the regression coefficient that represents the effect of the presence of natural attractors surrounding the building; α₃ is the regression coefficient that represents the effect of the presence of artificial attractors, and α₄ is the regression coefficient that represents the effect of the number of windows on the probability of collision.