World Values Survey 2009, Wave 5

The sampling universe included the adult population of Georgia residing in both rural and urban areas, excluding the conflict zones of Abkhazia and Ossetia. Military bases and prisons were also not included. In addition, some villages near the regional city of Gori and Zugdidi that are still under occupation by Russian troops were not included in the sampling. The sample design involved a fivestage random cluster sampling. The sampling frame design is based on the 2002 census information.

In this sampling design the sampling units were:

1. Regions and individual cities (Tbilisi and other principal cities)
2. Towns and villages (primary sampling units, PSUs)
3. Districts in cities, towns, and villages in rural areas (sampling points, SPs)
4. Household (by household we mean a group of individuals who live under the same roof and use the same kitchen for cooking)
5. Randomly selected adult members of households At the first stage, the number of respondents was allocated by probability-proportional-to-size (PPS) method. Likewise, at the second and third stages PSUs and SPs were selected by the probability proportional to the unit size (PPS) method. Households were selected via a random route technique and respondents at the household level were selected via the next-birthday technique:

Stage 1 - Primary sampling units At the first stage of the sampling design Georgia was divided into 11 regions; the division being based on the official administrative and geographic regions of the country.

 1 Tbilisi 
 2 Kakheti 
 3 Shida Kartli 
 4 Kvemo Kartli 
 5 Samtskhe Javakheti 
 6 Ajara 
 7 Guria 
 8 Samegrelo 
 9 Imereti & Svaneti 
10 Mtskheta Mtianeti 
11 Racha 

Each region was stratified according to three criteria:

 a) Large cities over 45,000 individuals - There are seven large cities in Georgia including the capital. All of them will be included in the sampling frame and are regarded as having been selfrepresentative cities or PSUs. 
 b) Other cities and towns with populations of less than 45,000 
 c) Rural settlements The number of interviews in all 10 regions was allocated proportional to the size of the adult population in each region. 

Stage 2 - Selection of PSUs In this stage the PSUs are equivalent to rayons- there are a total of 59 rayons (PSUs) in Georgia (excluding Abkhazia and Ossetia). The final sample covered 24 PSUs; this included seven self-representative PSUs were also included in this number. Due to the security reasons, areas close to Ossetian (town of Akhalgori, which was and continues to be under by Russian troops and the buffer zone areas), as well as the town Zugdidi (villages and small towns surrounding town of Zugdidi) were excluded from the sampling framework.
Stage 3 - Selection of sampling points (SPs) In urban areas the SPs were census districts whereas in rural areas an entire village was considered as an SP. There are total of 16,582 registered census districts in Georgia and for each one, information existed as to its location/address and the size of the adult population. In the pre-selected PSUs (according to PPS), the number of SPs were determined and per each selected SP around 10 interviews were completed. Rural areas villages are considered as a separate SP and from the list of villages, (this list contains information on the number of adult population per village), and the SPs was selected by PPS. The achieved sampling framework consisted of 188 randomly selected (via PPS) SPs
Stage 4 - Selection of households Selection of households was conducted by the application of a random route technique. For each one, SP starting points were identified and given to supervisors who then instructed interviewers as to how sampling mechanism was to be completed. Interviewers were then instructed to make up to two call backs if the original respondent was not available at the time of the initial contact.

Remarks about sampling:

The interviewer was given a route map in which a starting point for each sample point was accurately indicated. Every interviewer was then expected to have conducted no less than 10 interviews for urban SP and 5 among rural sampling points. The choice of starting points for all SP was made by the project manager or supervisor and was not left to the interviewers discretion. The STARTING POINT may be any point along the route, including day care establishments, schools, hospitals, administration buildings, or the beginning or end of a street (the starting point was indicated on the route map beforehand). If the starting point was the beginning of a street, it is necessary to keep to one side of the street (right or left). If a crossroad is met during the route, it is necessary to turn at this juncture and stay to the same side of the route until an appropriate respondent was chosen (i.e. if the left side is chosen, it was necessary to keep to the left side of the crossroad). If the required number of appropriate respondents was not found and the street ended, the interviewer should than have turned to the other side of the street and continued to the left handed side of the street. If the starting point had been a multi-storied building, the interviewer should have proceeded to the top floor and knocked at the door of the apartment on the side of which he followed during the route. It was not possible skipped any apartment until the appropriate respondent was found. After the interview with the appropriate respondent was completed the interviewer was to have followed the route and selected every fifth apartment. The interviewer followed this method after a successful interview was completed. In other cases s/he should have visited the next apartment until an interview was completed. If the interviewer were meeting private houses/plots on the sampling route, he should follow the instructions as indicated above and to have visited every fifth household. Interviews were held only in buildings that contained residences. Exceptions were those buildings (private hospitals, shops, restaurants, etc.), where one or more families permanently resided. The interviewer must allowed the supervisor to have been informed of and coordinated with him any changes that were concerned with the route that occurred during the fieldwork.

The sample size for Georgia is N=1500 and includes the national population aged 18 years and over for both sexes.