Customer preferences heterogeneity toward avocado: a latent class approach based on the best–worst scaling choice modeling

Agricultural and Food Economics

Table 5 Latent class- MLR model: individuals variables predictivity on cluster membership

Clusters	Hedonistic		Avocado fruit lovers		Sustainability sensitive		Value for money
Group Membership	Β	St. Error	β	St. Error	β	St. Error	β	St. Error
Constant	− 18.915***	0.946	− 0.737	0.938	− 0.346	0.979	− 2.790**	1.041
Age = 1	19.162***	0.646	1.573**	0.690	0.716	0.672	2.160**	0.737
Age = 2	19.113***	0.608	1.086	0.651	0.251	0.624	1.095	0.710
Age = 3	19.017***	0.614	1.167	0.657	0.272	0.634	1.353*	0.714
Age = 4	18.657***	0.693	1.123	0.700	1.147	0.656	1.773*	0.742
Age = 5	18.100	0.000	1.018	0.711	0.696	0.681	0.773	0.793
Family size = 1	− 0.779	0.561	− 0.472	0.540	− 0.950	0.566	0.372	0.574
Family size = 2	− 1.261**	0.502	− 0.873	0.478	− 0.990*	0.490	0.083	0.522
Family size = 3	− 2.088***	0.538	− 1.262*	0.479	− 1.720***	0.507	0.090	0.516
Family size = 4	− 1.073**	0.496	− 1.085*	0.478	− 1.255**	0.495	− 0.196	0.523
monthly average income = 0	0.608	0.547	-0.319	0.469	0.984	0.563	1.385*	0.517
monthly average income = 1	0.983*	0.523	0.682	0.421	1.263**	0.543	1.837***	0.500
monthly average income = 2	0.500	0.522	0.605	0.407	1.576**	0.520	1.482**	0.491
monthly average income = 3	0.308	0.596	0.850*	0.440	1.131**	0.573	1.139*	0.539

The “Health sensitive” cluster was considered as the reference for comparison with the other consumer groups.
Model Fit Statistics: Nagelkerke Pseudo-R²= 0.168. Full model χ2 (df = 68) = 142.456, p-value< 0.001. Classification accuracy (67%)
* p-value < 0.05; ** p-value < 0.01; *** p-value < 0.001