Impact of large-scale agricultural investments on the food security status of local community in Gambella region, Ethiopia

Table 12 Covariates balance summary

Variable	Standardized differences		Variance ratio
Variable	Raw	Weighted	Raw	Weighted
hhhead_sex	0.0086092	− 0.0002477	1.008092	0.9998906
Marital_HHH	− 0.1092941	0.0383494	0.8199685	1.099758
age_hhhead	− 0.201809	0.0766109	0.7789605	0.8793335
edu_hh	0.3104894	0.0315416	0.9861348	1.021953
HHH_OCC	0.0450241	0.0006894	1.164896	1.002301
total_famsize	0.2004223	0.0120328	1.033727	0.9646712
dep_ratio	0.1230064	0.0292944	0.9124085	0.8594963
land_size	− 0.4637001	− 0.0609299	0.6501092	1.223166
livestock_tlu	0.2063226	− 0.149653	1.338235	0.4849166
distance_health	0.2681744	− 0.1253866	2.395219	0.7003591
distance_market	− 0.5013956	− 0.0364788	0.9154446	1.435774
acess_credit	0.0250797	− 0.0782048	1.007776	0.9872627
radio_infosou ~ e	0.0035633	0.1039974	1.009693	1.326099
family_infoso ~ e	0.145002	0.1440818	1.138736	1.136617
keblead_infos ~ e	0.3808343	0.0645619	1.420597	1.036628
DA_infosource	− 0.0227411	0.0940016	1.011387	.9692412
imprseed_use	0.1476102	0.1476096	1.131153	1.130154
Over identification for covariate balance test
HO: covariates are balanced
Chi² = 10.784
Prob > chi² = 0.9047

Source: own survey, 2018
We checked covariate balance by carrying out over identification test and checking the module adjusted difference in means and ratio of variance between the treated and comparison for each covariate. The result shows that the differences in weighted means are insignificant, and the variance ratio for most of the variables is close to one (Table 12). The over identification test shows that we accept the null hypothesis that says the covariates are balanced (Table 12). This test and diagnostics confirm that covariates are balanced

ISSN: 2048-7010