Post Distribution Monitoring Survey in Kalobeyei Settlement, March 2020
Kenya, 2020
UNHCR
Created on
April 07, 2021
Last modified
April 07, 2021
Page views
723
Documentation in PDF
Metadata
DDI/XML
JSON
Identification
Survey ID Number
KEN_2020_PDMS-KAL_v01_M
Title
Post Distribution Monitoring Survey in Kalobeyei Settlement, March 2020
Country/Economy
| Name | Country code |
|---|---|
| Kenya | KEN |
Abstract
The present PDM was conducted under time and resource constraints related to COVID-19 emergency. Due to the restrictions on public gathering and partial restrictions on movements, the survey findings could not have been triangulated with the FGDs or market assessment, which will be an integral part of all subsequent PDMs.
The PDM household survey data collection took place over three days on 25 - 27 March 2020. ProGres V4 data of Kalobeyei persons of concern was used as a sampling frame, with a sample drawn using stratified random sampling based on random numbers generation. The original sample included over 400 households (adjusted for a non-response rate) aiming at a confidence level of 95% with a confidence interval of 5. However, the enumerators managed to conduct 457 interviews with respondents added through convenience sampling. Due to poor quality of some of the records, however, only 388 data entries were validated. This nevertheless allows us to remain within the same degree of precision in the inference, although affected by a bias linked to a non-probability sampling.
The PDM household survey data collection took place over three days on 25 - 27 March 2020. ProGres V4 data of Kalobeyei persons of concern was used as a sampling frame, with a sample drawn using stratified random sampling based on random numbers generation. The original sample included over 400 households (adjusted for a non-response rate) aiming at a confidence level of 95% with a confidence interval of 5. However, the enumerators managed to conduct 457 interviews with respondents added through convenience sampling. Due to poor quality of some of the records, however, only 388 data entries were validated. This nevertheless allows us to remain within the same degree of precision in the inference, although affected by a bias linked to a non-probability sampling.
Kind of Data
Sample survey data [ssd]
Unit of Analysis
Household
Version
Version Description
v2.1: Edited, anonymous dataset for public distribution
Version Date
2020-04-01
Scope
Notes
Household
Topics
| Topic |
|---|
| Cash Assistance |
| Basic Needs |
| Protection |
Coverage
Geographic Coverage
Kalobeyei Refugee Camp
Universe
Mixed - heads of households & the registered receipient of cash assistance
Producers and sponsors
Primary investigators
| Name |
|---|
| UNHCR |
Sampling
Sampling Procedure
UNHCR Nairobi CBI and Information Management Officers developed and coded the data collection tools and facilitated, together with Data Management Kakuma, the sampling. ProGres V4 data of Kalobeyei persons of concern was used as a sampling frame, with a sample drawn using random numbers generation. Unfortunately, because Assistance Record has not been updated in the V4, some of the respondents that appeared on the lists turned out to be non-recipients (newly relocated refugees, recently split households, etc.). As a result of this sampling frame error and a high non-response rate (including recipients not present at their places of residence), despite the sample size having been initially adjusted for the non-response rate, the teams opted to replace them with available respondents found on the ground (convenience sampling). The decision was taken as replacing the respondents with the ones from the original sample was logistically unfeasible as enumerators were allocated to particular locations and would have risked exhausting the sample without reaching saturation, while further delay was unacceptable in the face of evolving COVID-19 pandemic and the risks that the disease might reach the camp thus halting the exercise.
The original sample included over 400 households (adjusted for a non-response rate) aiming at a confidence level of 95% with a confidence interval of 5. However, the enumerators managed to conduct 457 interviews with respondents added through convenience sampling. Due to poor quality of some of the records, however, only 388 data entries were validated. This nevertheless allows us to remain within the same degree of precision in the inference, although affected by a bias linked to a non-probability sampling.
The original sample included over 400 households (adjusted for a non-response rate) aiming at a confidence level of 95% with a confidence interval of 5. However, the enumerators managed to conduct 457 interviews with respondents added through convenience sampling. Due to poor quality of some of the records, however, only 388 data entries were validated. This nevertheless allows us to remain within the same degree of precision in the inference, although affected by a bias linked to a non-probability sampling.
Weighting
Computed post hoc
Data Collection
Dates of Data Collection
| Start | End |
|---|---|
| 2020-03-16 | 2020-05-23 |
Data Collection Mode
Computer Assisted Personal Interview [capi]
Data Collectors
| Name |
|---|
| UNHCR |
Access policy
Archive where study is originally stored
United Nations Refugee Agency Microdata Library
https://microdata.unhcr.org/index.php/catalog/242
Original Archive Study ID:DDI_KEN_UNHCR_PDM_CBI_Kalobeyei_2020_v1.0
Cost: None
https://microdata.unhcr.org/index.php/catalog/242
Original Archive Study ID:DDI_KEN_UNHCR_PDM_CBI_Kalobeyei_2020_v1.0
Cost: None
Metadata production
DDI Document ID
DDI_KEN_2020_PDMS-KAL_v01_M
Producers
| Name | Affiliation | Role |
|---|---|---|
| David WO Oremo | UNHCR Bureau for East, Horn & The Great Lakes Region | Regional Microdata Curator |
Date of Metadata Production
2021-04-01
DDI Document version
Version 1 (April 2021). This version is identical to UNHCR DDI ID: DDI_KEN_UNHCR_PDM_CBI_Kalobeyei_2020_v1.0, except the following edit was made:
- Specified the archive where study is originally stored
- Specified the archive where study is originally stored