Egypt - Labor Force Survey, LFS 2017
---> Sample Design and Selection
The sample of the LFS 2017 survey is a self-weighted two-stage stratified cluster sample.
The main elements of the sampling design are described as follows:
* Sample Size
The sample size in each quarter is 22,896 households with a total number of 91,584 households annually. These households are distributed on the governorate level (urban/rural), according to the estimated number of households in each governorate in accordance with the percentage of urban and rural population in each governorate.
* Cluster size
The cluster size is 18 households.
* Sampling stages:
(1) Primary Sampling Unit (PSU):
The 2006 Population Census provided sufficient data at the level of the Enumeration Area (EA). Hence, the electronic list of EA's represented the frame of the first stage sample; in which the corresponding number of households per EA was taken as a measure of size.
The size of an EA is almost 200 households on average, with some variability expected. The size of first stage national sample was estimated to be 5,024 EA.
(2) Sample Distribution by Governorate:
The primary stratifying variable is the governorate of residence, which in turn is divided into urban and rural sub-strata, whenever applicable.
(3) First Stage Sample frame:
The census lists of EAs for each substratum, associated with the corresponding number of households, constitute the frame of the first stage sample.
The identification information appears on the EA's list includes the District code, Shiakha/Village code, Census Supervisor number, and Enumerator number.
Prior to the selection of the first stage sample, the frame was arranged to provide implicit stratification with regard to the geographic location.
The urban frame of each governorate was ordered in a serpentine fashion according to the geographic location of kism/ district capitals. The same sort of ordering was made on the rural frame, but according to the district location. The systematic selection of EA's sample from such a sorted frame will ensure a balanced spread of the sample over the area of respective governorates.
The sample was selected with Probability Proportional to Size (PPS), with the number of census households taken as a Measure of Size (MOS).
(4) Core Sample allocation
The core sample EAs (5,024) were divided among the survey 4 rounds, each round included 1,272 EAs (585 in urban areas and 687 in rural areas).
A more detailed description of the different sampling stages and allocation of sample across governorates is provided in the Methodology document available among external resources in Arabic.
On the national level (90.5 %)
Urban level (82.6%)
Rural level (97.3%)
Response rates on the governorate level are presented in the methodology document attached to the documentation materials published in Arabic.