A two stage stratified sample by regions (marzes) has been designed to conduct the survey. The method of systematic probability sampling was used to frame the sample. Yerevan and all marzes with rural and urban settlements were covered by sample population with proportion of all HH available in these strata. Based on the administrative and territorial division of RA at the first stage of sampling, the preliminary sampling units, i.e. enumeration areas were selected. 2001 Armenian Population Census results were used to calculate the mentioned proportions.
At the second stage of sampling the primary sampling units, that are HHs to be surveyed, were randomly selected. As a result, during the year have been surveyed 7 680 HHs (monthly 640 HHs), of which 5 160 urban and 2 520 rural areas.
At the same time a reserve sample was selected by the same steps, because it was anticipated that in most cases the questionnaire may not be completed due to household’s refusal to participate in the survey, absence from the republic or other reasons. As a result, the HHs not participated in the survey were replaced from the reserve sample. The selection of households (HH) has been done monthly by the rotation principle.
