Table 5.

Multinomial regression of consumer knowledge group

	Overestimation group				Underestimation group				Knowledge group
	B	SE	Exp(B)	p	B	SE	Exp(B)	p	B	SE	Exp(B)	p
Gender	.035	.200	1.035	.861	.252	.190	1.286	.185	−.083	.185	.921	.656
Age	−.007	.009	.993	.452	.001	.009	1.001	.945	−.006	.009	.994	.475
Education	.356	.179	1.427	.047	−.254	.143	.776	.076	.109	.155	1.116	.481
Monthly income	.320	.079	1.377	.000	.175	.078	1.192	.025	.226	.075	1.253	.003
Food involvement	.416	.156	1.516	.007	.208	.140	1.231	.137	.681	.153	1.976	.000
Health involvement	.160	.161	1.173	.321	.180	.151	1.198	.232	.220	.150	1.246	.144
Media dependency	.289	.222	1.334	.194	−.085	.236	.919	.720	.356	.209	1.427	.089
GM foods-related information tendency	−.490	.124	.613	.000	.034	.114	1.035	.765	−.015	.111	.985	.891

Notes: Reference group = ignorance group; −2Log likelihood = 2.336E3; Chi-square = 136.703 (df = 24, p <.000); Gender (0 = female, 1 = male); Education (0 = No education, 1 = Elementary school, 2 = Middle school, 3 = High school, 4 = College, 5 = Graduate school); Monthly income (1 = under 1 million, 2 = 1–1.99 million, 3 = 2–2.99 million, 4 = 3 − 3.99 million, 5 = 4–4.99 million, 6 = 5–5.99 million, 7 = 6–6.99 million, 8 = 7 million +); Media dependency (0 = off-line, 1 = on-line); GM foods-related information tendency are scored on a 5-point Likert scale, with responses ranging from 1 (very negative) to 5 (very positive), with a higher score corresponding to a greater positive orientation. GM = genetically modified.