Skip to main content
. 2018 Aug 24;15(1):e12656. doi: 10.1111/mcn.12656

Table 4.

Determinants of number of food groups consumed

Poisson (beta coefficient) Ordered logit (odds ratio)
(1) (2) (3) (4) (5) (6)
Wealth quintile 0.0687*** (0.0213) 0.0683*** (0.0203) 0.0705*** (0.0216) 1.409*** (0.158) 1.412*** (0.151) 1.423*** (0.163)
Knowledge (all) 0.0630 (0.193) 1.608 (1.639)
Knowledge (behaviour) 0.0848 (0.178) 1.532 (1.488)
Knowledge (mechanisms) −0.0451 (0.140) 1.007 (0.759)
Livestock −0.00915 (0.0224) −0.00919 (0.0225) −0.00917 (0.0225) 0.951 (0.0935) 0.950 (0.0932) 0.950 (0.0951)
Garden −0.0685 (0.0498) −0.0666 (0.0508) −0.0674 (0.0508) 0.873 (0.218) 0.882 (0.230) 0.871 (0.233)
Education (years) 0.00804 (0.00661) 0.00788 (0.00659) 0.00916 (0.00642) 1.043 (0.0335) 1.044 (0.0337) 1.048 (0.0324)
District fixed effects Yes Yes Yes Yes Yes Yes
Observations 248 248 248 248 248 248
F test 3.22 3.31 2.98 2.29 2.22 2.18
Prob > F 0.0175 0.0156 0.0247 0.0669 0.0747 0.0790

Note. Nutrition knowledge index defined as per cent answered correctly (out of all questions, behaviour questions or mechanisms questions only) of the questions that were answered correctly by 10–90% of respondents, normalized to the 95th percentile. Wealth is quintiles of an asset‐based index. Livestock measured in TLUs. Education is years of schooling. Age measured in years. Constant terms not shown. Heteroskedasticity robust standard errors clustered at the care group level shown in parentheses, with significance levels denoted. TLUs: tropical livestock units.

*

P < 0.1.

**

P < 0.05.

***

P < 0.01.