Table 5.

Multilevel, mixed-effects tobit regression models of food insecurity (HFIAS) scores using three measures of household water expenditures and controlling for select household characteristics, including an interaction term for water insecurity score and water expenditures (n = 3655).

	Model 7		Model 8		Model 9
	Β	SE	β	SE	β	SE
Fixed effects
Water expenditures measure
Absolute USD Z-score	−0.42	0.24	–	–	–	–
% income	–	–	0.05*	0.02	–	–
% Income Z-score	–	–	–	–	1.23***	0.27
Household characteristic
Age	0.02*	0.01	0.02*	0.01	0.02*	0.01
Gender	−0.26	0.26	−0.23	0.26	−0.23	0.26
Number of children	0.37***	0.07	0.34***	0.07	0.34***	0.07
Amount of stored drinking water (100 l)	−0.00	0.00	−0.00	0.00	−0.00	0.00
Total water storage (100 l)	−0.00	0.00	−0.00	0.00	−0.00	0.00
Primary water source is purchased/vended	−2.03***	0.37	−2.44***	0.37	−2.53***	0.37
Rural context	1.23	0.64	1.34**	0.64	1.31*	0.64
Water insecurity score	0.51***	0.02	0.49***	0.02	0.49***	0.02
Water insecurity score * expenditures (interaction term)	−0.05*	0.02	0.00	0.00	−0.04	0.02
Random effects
Cluster	5.52	2.05	5.34	2.05	5.34	2.04
Site	16.68	6.49	16.57	6.46	16.56	6.47
Model diagnostics (log likelihood)	−8994.81		−9001.76		−8995.97

Note

^*= P < 0.05.

^**= P < 0.01.

^***= P < 0.001.