Detailed description of the sample through phases:
- Phase 1 = Population of Serbia (excluding Kosovo, UNSCR 1244), stratified into 4 administrative regions – Belgrade, Vojvodina, Western Serbia and Eastern Serbia – proportionally to their share in the population.
- Phase 2 = The population is further stratified into sampling units in each region.
- Phase 3 = Distribution of sampling units (15 respondents per sampling unit).
- Phase 4 = Sampling units are selected from the list of municipalities based on the size and number of polling stations.
- Phase 5 = Uniform distribution of sampling units based on voter lists with polling station addresses. Each sampling unit is marked as a starting point with the given addresses and additional directions, such as rising and falling house numbers in the street.
- Phase 6 = Random selection of households, starting with the given sampling unit and selecting every second or third household (depending on the type of settlement). The selected home address is on the left side of the interviewer’s route. In residential buildings with up to four floors, the interviewer chooses every fifth apartment. In residential buildings with five or more floors, the interviewer chooses every tenth apartment.
- Phase 7 = Selection of respondents by applying the next-birthday method.
- Phase 8 = Replacement of respondents in case of rejection after three attempts (first visit and two calls).
The sample was created based on the nationally representative statistics by region and type of settlement.