Water, Sanitation, and Hygiene Project 2011
Republic of Cabo Verde, 2011
Get Microdata
Reference ID
CPV_2011_MCC-WSHP_v01_M
Producer(s)
National Institute of Statistics, Cabo Verde
Collection(s)
Metadata
Created on
May 28, 2015
Last modified
May 28, 2015
Page views
15560
Downloads
326
Identification
Survey ID Number
CPV_2011_MCC-WSHP_v01_M
Title
Water, Sanitation, and Hygiene Project 2011
Country/Economy
| Name | Country code |
|---|---|
| Republic of Cabo Verde | CPV |
Study type
Independent Evaluation
Abstract
The overall objective of this survey is to collect relevant information on water, sanitation and hygiene sectors in Praia, hinterlands of Santiago and Sal Islands. Specifically, the survey collects information on: The socio-demographic characteristics of the household; determine the levels of expenditure and revenues of households ; information on the characteristics of water supply systems ; information on the behaviors regarding water storage ; data on the treatment of drinking water and disposal of feces ; water borne illnesses ; exposure to information on sanitation conveyed by the media ; community participation in health-related activities ; property rights ; real estate registration ; land-ownership conflicts.
Kind of Data
Sample survey data [ssd]
Unit of Analysis
Households
Version
Version Description
Anonymized dataset for public distribution
Scope
Keywords
| Keyword |
|---|
| Cabo Verde |
| WASH |
| Water |
Coverage
Geographic Coverage
Praia, Hinterland of Santiago Island and Sal Island. Specific Municipalities Include: Tarrafal, Santa Catarina, Sao Salvador Mundo, Santa Cruz, Sao Miguel, Sao Lourenco dos Orgaos, Sao Domingos, Ribeira Grande Santiago, Praia, Sal.
Producers and sponsors
Primary investigators
| Name |
|---|
| National Institute of Statistics, Cabo Verde |
Funding Agency/Sponsor
| Name | Abbreviation |
|---|---|
| Millennium Challenge Corporation | MCC |
Sampling
Sampling Procedure
The sampling method is randomly probabilistic, and is set in two stages: In stage one, the primary sampling units are selected , in stage two, households are selected using the 2010 Census Dataset.
The number of households to be drawn per EA took into account that there is no new enumeration of households before the survey. The sampling has involved two selection stages: the first stage focused on the EAs and the second degree is the draw of households. 19 enumeration areas, were selected and within which 20 households have to be selected. Thus, the number of households to be drawn is increased by 9% to outdo the potential loss of households due to the selection of vacant households and of households that have not been located.Interviewers are requested to solely interview pre-selected households. Exchanges of households are not allowed to prevent bias, even for non-respondent households.
In conclusion, the sample selection process has not incorporated a plan for addressing non-response,but took into account this eventuality increasing the sample size.
The number of households to be drawn per EA took into account that there is no new enumeration of households before the survey. The sampling has involved two selection stages: the first stage focused on the EAs and the second degree is the draw of households. 19 enumeration areas, were selected and within which 20 households have to be selected. Thus, the number of households to be drawn is increased by 9% to outdo the potential loss of households due to the selection of vacant households and of households that have not been located.Interviewers are requested to solely interview pre-selected households. Exchanges of households are not allowed to prevent bias, even for non-respondent households.
In conclusion, the sample selection process has not incorporated a plan for addressing non-response,but took into account this eventuality increasing the sample size.
Response Rate
Based on the number of households surveyed (1311) and the planned (1520) number of households, the overall rate of response is 94.3%. It is 93.3% in urban area and 97.4% in rural area. This rate was obtained as a ratio between the total number of completed interviews and the number of households to be surveyed. Data collection showed that among the 1.520 households that were planned for the survey, 1.390 households should be interviewed, the other households were vacant, uninhabitable buildings, etc. Meanwhile, 1.311 households were surveyed.
Weighting
While the sample is approximately self-weighting within each domain, it is not so across the different domains. It is therefore necessary to calculate appropriate weights, which can be applied to each household during the data processing, before tables are created. These weights need to take account of the selection probabilities at the two stages. The weight also take account of the effects of two other things that happen in the field: any differences in the size of the EA, as revealed through the listing of the households, and any non-response (e.g. refusals or non-contacts) that occurs on the survey.
The calculation of this weighting factor can be illustrated for the EA 411005 selected in urban Sal. The sample design allows for the selection of 19 x 20 = 380 households in that domain . The census 2010 has estimated the total urban households in Sal as 6296. The EA 411005, had 157 households in 2010 (no new listing of household was done before the survey).
In the EA 411005 of Sal, the selection probabilities were calculated as follows: First selection probability at the first stage is therefore :P1= 19x20/6296.
Second selection probability at the second stage: P2=157/157 . Third selection probability of households responding successfully :P3=18/20.
The weighting factor is then calculated by multiplying together the inverses of these ratios. This is: Hh_weight = WT 1 x WT 2 x WT 3
The calculation of this weighting factor can be illustrated for the EA 411005 selected in urban Sal. The sample design allows for the selection of 19 x 20 = 380 households in that domain . The census 2010 has estimated the total urban households in Sal as 6296. The EA 411005, had 157 households in 2010 (no new listing of household was done before the survey).
In the EA 411005 of Sal, the selection probabilities were calculated as follows: First selection probability at the first stage is therefore :P1= 19x20/6296.
Second selection probability at the second stage: P2=157/157 . Third selection probability of households responding successfully :P3=18/20.
The weighting factor is then calculated by multiplying together the inverses of these ratios. This is: Hh_weight = WT 1 x WT 2 x WT 3
Data Collection
Dates of Data Collection
| Start | End |
|---|---|
| 2011 | 2011 |
Questionnaires
Questionnaires
Questionnaires were developed by MCC. INECV amended and adapted it to the country reality.
Access policy
Contacts
| Name | Affiliation | |
|---|---|---|
| Monitoring & Evaluation Division | Millennium Challenge Corporation | impact-eval@mcc.gov |
Location of Data Collection
Millennium Challenge Corporation
Archive where study is originally stored
Millennium Challenge Corporation
http://data.mcc.gov/evaluations/index.php/catalog/80
Cost: None
http://data.mcc.gov/evaluations/index.php/catalog/80
Cost: None
Metadata production
DDI Document ID
DDI_CPV_2011_MCC-WSHP_v01_M
Producers
| Name | Abbreviation | Role |
|---|---|---|
| Millennium Challenge Corporation | MCC | Review of Metadata |
| National Institute of Statistics - Cape Verde | Produce Metadata |
Date of Metadata Production
2014-02
DDI Document version
Version 1.1 (March 2014)
Version 2.0 (May 2015). Edited version based on Version 01 (DDI-MCC-CPV-NIS-WATSAN-2011-v01.1) that was done by Millennium Challenge Corporation.
Version 2.0 (May 2015). Edited version based on Version 01 (DDI-MCC-CPV-NIS-WATSAN-2011-v01.1) that was done by Millennium Challenge Corporation.