Table 3.

Ten best-fitting models showing which practitioners were more likely to change their mind about an intervention after reading the evidence presented in the Bird Conservation Evidence Synopsis.^a

Model	Intercept	Prior	Past exposure to		Effective-	Certainty							Past exposure*
		knowledge	intervention		ness	of evidence	Experience	Organization		Region			experience		AIC	Δ AIC
			heard of	used				NGO	other	NZ	UK	other	heard of	used
1b	0.661						−0.042								1272.0	0.0
2b	−0.313		1.210	0.934			0.030						−0.086	−0.071	1272.4	0.4
3	0.251									−0.475	−1.114	0.047			1274.7	2.7
4 (null)	−0.099														1275.5	3.5
5	0.631		0.043	0.029			−0.042								1276.0	4.0
6	−0.049	−0.317													1276.4	4.4
7	−0.032				−0.002										1277.3	5.3
8	−0.158					0.002									1277.4	5.4
9	−0.069							−0.098	0.068						1279.3	7.3
10	−0.109		0.028	−0.007											1279.4	7.4

^aThe default categories in the model output are as follows: past exposure to an intervention—neither heard of or used; organization type—government organization; region—Australia. The average variances for the random effect variables intervention and practitioner across the 10 models are 0.14 and 2.18, respectively.

^bSelected as the best models based on their low AIC values.