Our Sampling Approach
Define the target population
We begin by clearly defining who the study represents (e.g., households, youth, service users, institutions) and the geographic scope. Inclusion and exclusion criteria are set at this stage to avoid ambiguity during fieldwork.
02
Selecting the appropriate sampling method
Based on study objectives and access conditions, we apply probability-based methods such as stratified random sampling, multistage cluster sampling, or random digit dialing (RDD) for phone surveys. Stratification is typically applied by governorate, locality type (urban, village/camp), gender, and other relevant characteristics.
03
Calculate sample size
Sample sizes are calculated using confidence levels and margins of error, taking into account population size, design effects, and expected non-response. This ensures statistically reliable results without over-sampling.
04
Develop the sampling frame
We construct or validate sampling frames using official statistics, administrative records, mobile number databases (for CATI), or field-based listing where required. Frames are reviewed for coverage gaps and updated before selection.
05
01
We apply multiple, concrete controls at each stage of data collection to prevent errors rather than correct them later.
Select the sample
Units are selected using random or systematic procedures appropriate to the design (e.g., PPS for cluster sampling). Selection protocols are documented and locked before data collection begins.
06
Implement field allocation and controls
Samples are geographically allocated to field teams with clear quotas and replacement rules. Replacement procedures are predefined and tightly controlled to prevent selection bias.
07
Monitor representativeness during data collection
Throughout fieldwork, we monitor sample composition in real time to ensure alignment with planned distributions. Corrective action is taken if deviations emerge.
08
Document assumptions and limitations
All sampling decisions, assumptions, response rates, and limitations are transparently documented and reported with the final results.