Table 2. Probit model of probability of...

Table 2.

Probit model of probability of choosing TN made hard apple cider (4 pack of 16-ounce cans), mean premium and marginal effects (ME)^a

	Estimated coefficient (β)	ME on	ME on ${\hat{P R E M}}_{T N}$ d
Variable	Estimated coefficient (β)	Pr[TN_i = 1]^b	Mean	LCL	UCL
Intercept	−0.744
P_TNi	−0.311**	−0.095**
AGE	−0.004	−0.001	−$0.01	−$0.03	$0.01
FEMALE	0.030	0.010	$0.11	−$0.51	$0.73
COLLGRAD	−0.184	−0.056	−$0.59	−$1.28	$0.10
HHINCDOL	0.000	0.000	$0.00	−$0.01	$0.01
LOCFOOD	0.429**	0.131**	$1.38	$0.84	$1.92
METRO	−0.295**	−0.090**	−$0.95	−$1.60	−$0.30
WEST	−0.087	−0.027	−$0.28	−$1.10	$0.54
MIDDLE	−0.077	−0.024	−$0.25	−$0.94	$0.44
FARMBACK	−0.165	−0.050	−$0.53	−$1.23	$0.17
FAMILIAR	0.009	0.003	$0.03	−$0.32	$0.38
CIDERFREQ	0.023	0.007	−$0.13	$0.08	−$0.28
WKLYBEER	−0.124	−0.038	−$0.40	−$1.16	$0.37
WKLYWINE	0.056	0.017	$0.18	−$0.66	$1.02
WKLYWINECOOL	0.448*	0.137*	$1.44	$0.22	$2.65
WKLKYLIQ	−0.079	−0.024	−$0.25	−$1.22	$0.71
WKLYMALT	−0.559*	−0.169*	−$1.77	−$3.10	−$0.44
WKLYALCEXP	0.008	0.002	$0.02	$0.00	$0.05
CONSEQ	0.101	0.031	$0.32	−$0.02	$0.66
GROCERY	−0.211*	−0.064*	−$0.68	−$1.33	−$0.03
$\bar{P R E M}$ ^c			$3.22	$2.91	$3.52
N = 875
Percent correct = 72.00%
Pseudo R² = 0.2156
LLR χ² (19 df) = 260.96**

	Estimated coefficient (β)	ME on	ME on ${\hat{P R E M}}_{T N}$ d
Variable	Estimated coefficient (β)	Pr[TN_i = 1]^b	Mean	LCL	UCL
Intercept	−0.744
P_TNi	−0.311**	−0.095**
AGE	−0.004	−0.001	−$0.01	−$0.03	$0.01
FEMALE	0.030	0.010	$0.11	−$0.51	$0.73
COLLGRAD	−0.184	−0.056	−$0.59	−$1.28	$0.10
HHINCDOL	0.000	0.000	$0.00	−$0.01	$0.01
LOCFOOD	0.429**	0.131**	$1.38	$0.84	$1.92
METRO	−0.295**	−0.090**	−$0.95	−$1.60	−$0.30
WEST	−0.087	−0.027	−$0.28	−$1.10	$0.54
MIDDLE	−0.077	−0.024	−$0.25	−$0.94	$0.44
FARMBACK	−0.165	−0.050	−$0.53	−$1.23	$0.17
FAMILIAR	0.009	0.003	$0.03	−$0.32	$0.38
CIDERFREQ	0.023	0.007	−$0.13	$0.08	−$0.28
WKLYBEER	−0.124	−0.038	−$0.40	−$1.16	$0.37
WKLYWINE	0.056	0.017	$0.18	−$0.66	$1.02
WKLYWINECOOL	0.448*	0.137*	$1.44	$0.22	$2.65
WKLKYLIQ	−0.079	−0.024	−$0.25	−$1.22	$0.71
WKLYMALT	−0.559*	−0.169*	−$1.77	−$3.10	−$0.44
WKLYALCEXP	0.008	0.002	$0.02	$0.00	$0.05
CONSEQ	0.101	0.031	$0.32	−$0.02	$0.66
GROCERY	−0.211*	−0.064*	−$0.68	−$1.33	−$0.03
$\bar{P R E M}$ ^c			$3.22	$2.91	$3.52
N = 875
Percent correct = 72.00%
Pseudo R² = 0.2156
LLR χ² (19 df) = 260.96**

Notes:

^a**denotes significance at α = 0.01 and *denotes significance at α = 0.05. ^bThe ME of the jth variable X_ij on the probability of the ith respondent selecting the TN hard apple cider over that not labeled is ∂Pr [TN_i = 1]/∂X_ij = ϕ_i ·β_j, where β_j is the parameter on X_ij and ϕ_i is the normal density function. The mean MEs and their associated standard errors are calculated using the Krinsky and Robb method with 5,000 replications (Krinsky and Robb, 1986). ^cThe mean premium (⁠ $\bar{P R E M}$ ⁠) and associated standard errors are calculated for the 95% confidence level using the Krinsky and Robb method with 5,000 replications. ^dThe effects of each variable on WTP, the premium and the associated standard errors are calculated using the Krinsky and Robb method. The ME of the jth non-premium explanatory variable on estimated ${\hat{P R E M}}_{T N}$ is calculated as ${\begin{matrix} \hat{\partial P R E M} \end{matrix}}_{T N} / \partial X_{j}$ = − $β_{j} / β_{P_{T N}}$

[ViewLarge]

Sharing Unavailable