Table 1.

Randomisation balance: no significant differences in observed means between households assigned different latrine use questions

	Individual	Household	Difference
	(1)	(2)	(3)
No of households	1215	1216	−1
No of households by project
World Vision (in Bihar)	316	313	3
Coastal Salinity Prevention Cell (in Gujarat)	309	319	−10
Wateraid (in Karnataka)	297	296	1
Rural Welfare Institute (in Odisha)	293	288	5
Household members	5.685 (0.0950)	5.604 (0.101)	0.0803 (−0.117 to 0.278)
Female	0.491 (0.00485)	0.489 (0.00490)	0.00190 (−0.0115 to 0.0153)
Hindu	0.967 (0.00882)	0.960 (0.0102)	0.00737 (−0.00383 to 0.0186)
Household head completed at least 8 years of schooling	0.288 (0.0192)	0.319 (0.0199)	−0.0310 (−0.0675 to 0.00552)
Count of assets (max 13)	8.202 (0.111)	8.234 (0.117)	−0.0327 (−0.196 to 0.130)
Has latrine	0.943 (0.00795)	0.946 (0.00781)	−0.00251 (−0.0215 to 0.0165)
Got help from government or NGO to build toilet (given has toilet)	0.625 (0.0304)	0.656 (0.0274)	−0.0309 (−0.0692 to 0.00747)
Pit size (cubic feet, given has toilet)	179.0 (14.33)	180.7 (14.27)	1.666 (−25.94 to 22.61)
Toilet looks used (given has toilet)	0.805 (0.0234)	0.794 (0.0233)	0.0103 (−0.0243 to 0.0450)

Cluster robust SEs, clustered by village, next to means in columns 1 and 2. 95% CI next to differences in column 3. *P<0.05, **p<0.01.

None of the differences are significant at the 0.05 or 0.01 level, which is why there are no * or ** in the table.