Survey element Survey type Sampling frame Sample selection Stratification / comparison group Representation Sample size
Household
4-wave Panel
Zarqa and Amman (from Jordan Dept. of Statistics (DoS))
Survey geocodes selected based on ex ante matching of treated and control zones, using Census data Random sampling within sample geocodes Replacements selected from sample geocodes WNP only – WNP only control WWNP only – WWNP only control WNP+WWNP – WNP+WWNP control Distinct control groups from: Zarqa Amman
Representative of sample geocodes at baseline, based on comparisons to Census and other sources
3359 3416 3596 3662
Enterprise
2-wave Panel
Zarqa and Amman (from DoS + household referrals)
Same geocodes as household sample Random selection within sample geocodes Referrals for informal enterprises Replacements selected from closely neighboring enterprises
Same as household (though analysis uses all controls for each group to maximize statistical power)
Representative of sample geocodes at baseline Informal enterprises likely under-represented (due to low referral rates) 345 418
Farm
3-wave Panel
Jordan Valley and highlands (from DoS)
Survey zones selected based on expected differences in exposure to treated wastewater Random selection in sample zones Replacements selected within zones Five locations: Highlands u/s KTD (↑ river flow) JV1 North (↑ Non-Compact WW) JV2 Mid-North (↑ Compact WW) JV3 North-Central (↑ Compact WW) JV4 South-Central (little change in WW)
Representative of sample zones
551 539 539
Refugee
Single cross-section
UNHCR registration list for Zarqa and Amman
Priority survey geocodes selected according to household sample, with augmenting based on treatment status outside hh geocodes Random sampling by treatment status Referrals for unregistered refugees Treatment status: WNP only WWNP only Both WNP and WWNP Controls in Zarqa Control Amman Representative of registered population in sample areas Unregistered population likely under-represented (due to low referral rates)
1617
Water vendor
Single cross-section
Shops: Ministry of Health list + canvassing Tankers: Canvassing Full sampling from canvassed locations
None
Representative of water vendors in Zarqa and East Amman in 2018
320
Meter testing
Repeat cross-section
Meter listing in selected zones
Zones selected for variation in JC status, elevation, pressure, throughput (for survey 1), and JC status and meter replacement (for survey 2) Random sample of meters within selected zones
Compact and non-Compact zones
Not representative
37 223
Water loss testing
Single cross-section
Canvassing of land plots in selected areas
“Well isolated” zones selected (as suggested by utility) Comparison of meter registered data to bulk meter inflow Random sub-sample of meters evaluated to adjust for meter error
Meter error testing sub-sample stratified by meter replacement status
Not representative; only relevant to “well isolated” zones
1797
Key informant interviews
Single cross-section
Listing of key JC stakeholders
Contact to all listed stakeholders Replacements included as suggested by stakeholders
None
Representative of institutions, but likely not all perspectives
22