How does Community-Led Total Sanitation (CLTS) affect latrine ownership? A quantitative case study from Mozambique

Table 3 Predictors of Latrine Ownership in Logistic Regression Analysis

Model	B	SE	Wald X² (1)	OR	95% CI
Model 4: significant context and RANAS factors from model 1 + 2 + 3
Context factors
Relationship status^a	.545	.388	1.969	.58	.27, 1.24
Years at school	.188	.070	7.247**	1.21	1.05, 1.39
Risk of flooding	−.351	.128	7.546**	.70	.55, .90
Social dilemma	.046	.131	.123	1.04	.81, 1.35
Social capital (solidarity)	.110	.093	1.411	1.12	.93, 1.34
Social capital (trust)	−.080	.103	.602	.92	.75, 1.13
Social capital (social cohesion and inclusion)	.377	.119	10.068**	1.46	1.16, 1.84
RANAS factors
Vulnerability (personal general risk for diarrhea)	−.626	.113	30.734***	.54	.43, .67
Feeling more respected	−.381	.141	7.327**	.68	.52, .90
Beliefs about costs and benefits (money, space, time)	−1.143	.267	18.246***	.32	.19, .54
Others’ behavior (community)	1.176	.141	69.105***	3.24	2.46, 4.28
Others’ (dis)approval (personally important others’)	.544	.161	11.479**	1.72	1.26, 2.36
Confidence in recovery of broken latrine	.994	.199	25.029***	2.70	1.83, 3.99
Communication	.155	.136	1.297	1.17	.89, 1.52
Constant	−8.13	1.62	25.381***

Note. N = 598. For the overall model of significant context and psychosocial factors (Model 4) R² = .74 (Nagelkerke). X²(15) = 468.19, p < .0005. Latrine ownership was coded ‘1’ and no latrine ownership was coded ‘0’.
^aNo relationship as reference category;
OR = odds ratio; CI = confidence interval; **P < .005; ***P < .0005

ISSN: 1471-2458