TABLE 8.

Summary of final logistic regression models for each component of basic knowledge^a

Knowledge component	Variable	Estimate (β)	SE	z value	p value	Odds ratio
Pathogen-likeness	Intercept	0.669	0.114	5.863	<0.001	1.95
	Unmodified pathogen	3.116	0.472	6.598	<0.001	22.6
	Vaccine is treatment	−2.281	0.424	−5.385	<0.001	0.10
	Vaccine directly harms pathogen	−2.340	0.597	−3.920	<0.001	0.10
	Single vaccine for all diseases	−1.911	0.418	−4.568	<0.001	0.15
Immune activation	Intercept	0.048	0.108	0.442	0.66	1.05
	Unmodified pathogen	0.496	0.189	2.626	0.009	1.64
	Vaccine is treatment	−2.625	0.608	−4.320	<0.001	0.07
	Vaccine directly harms pathogen	−3.025	1.029	−2.939	0.003	0.05
	Single vaccine for all diseases	−2.163	0.540	−4.009	<0.001	0.11
Prevention	Intercept	1.169	0.120	9.717	<0.001	3.22
	Unmodified pathogen	0.530	0.235	2.258	0.02	1.70
	Vaccine is treatment	−1.288	0.295	−4.373	<0.001	0.28
	Vaccine directly harms pathogen	−1.465	0.370	−3.956	<0.001	0.23

^an = 635. Bolded values are significantly different from 0 (p < 0.05). These final models are the product of a model selection process that finds the most significant variables. Initially included variables were the presence or absence of the five most common inaccurate ideas (listed in Table 6).