Skip to main content

Table 4 LCM estimates of precooked beans product attributes and segment membership

From: Consumer demand heterogeneity and valuation of value-added pulse products: a case of precooked beans in Uganda

Attribute

Multinomial logit model

Latent class model

Nutrition enhancement lovers

Conservative self-reliant bean consumers

Precooked bean enthusiasts

Utility function estimates: precooked bean attributes

 ASC

− 1.7409***

(0.1363)

− 5.6721***

(0.4747)

− 3.5850***

(1.0841)

0.4933***

(0.1855)

 TIME

− 0.0082***

(0.0017)

− 0.0155***

(0.0056)

− 0.0106

(0.0115)

− 0.0075***

(0.0025)

 FUEL

0.0051***

(0.0010)

0.0134***

(0.0042)

0.0188**

(0.0093)

0.0048***

(0.0015)

 NUTRI

1.6555***

(0.0513)

4.1696***

(0.2125)

1.6852***

(0.3883)

0.6213***

(0.0810)

 WATER

0.1553***

(0.0473)

0.2159

(0.1527)

− 0.4683

(0.4934)

0.3425***

(0.0677)

 PRICE

− 0.0061***

(0.0006)

− 0.0073***

(0.0019)

− 0.0344***

(0.0077)

− 0.0077***

(0.0008)

Segment membership function estimates: consumer characteristics

 Constants

− 0.1316

(0.2692)

− 1.2409***

(0.4307)

 Age of the respondent (Ln years)

− 0.0001

(0.0009)

− 0.0011

(0.0013)

 Sex of the respondent (1 = male, 0 = female)

− 0.4845**

(0.2370)

− 0.3201

(0.4631)

 Education of the respondent (1 > primary level, 0 ≤ primary level)

0.5313**

(0.2359)

0.1253

(0.4137)

 Bean supply sufficiency (1 = self-sufficient, 0 = not self-sufficient)

0.8200***

(0.2446)

0.5915**

(0.2447)

 Processing benefits (PB) index

0.1307

(0.1090)

− 0.5639**

(0.2438)

 Consumer location (1 = urban, 0 = otherwise)

0.3316

(0.2452)

0.0825

(0.4171)

 Quantity of beans consumed per week (kg)

− 0.8211***

(0.2445)

− 0.5944**

(0.2445)

 Household types (1 = dual household, 0 = otherwise)

0.0116

(0.2285)

− 0.7860*

(0.4069)

  1. Model simulation: MNL LL = − 3210.53, AIC = 6433.1, BIC = 3235.35, R2 Adj = 0.1742, LCM: LL = − 2738.23, AIC = 5548.5, BIC = 2887.09, R2 Adj = 0.3589, observations (choices) = 3906, sample size = 558
  2. Dependent variable is the Choice option selected by the respondent
  3. *p < 0.10; **p < 0.05; ***p < 0.01