Sampling in WVS-7
WVS-7 sample model applied in a country must correspond the following criteria:
- WVS surveys are required to cover all residents (not just citizens) in a country in the age of 18 years older and older;
- PI’s can lower the minimum age limit as long as the minimum required sample size for the 18+ population is achieved;
- Obtained sample should be representative, i.e. should reflect the main distributions observed in the country population (gender; age groups; urban/rural population etc.).
The sample type preferable for using in the World Values Survey is a full probability sample of the population aged 18 years and older.
Probability sampling is a sampling technique wherein the samples are gathered in a process that gives all the individuals in the population equal chances of being selected.
In most cases, a list or registry of all households or voters in the country is required to build a full probability sample. Recognizing that the very high cost of full probability samples may prove prohibitive in some cases, WVSA allows application of a national representative random sample based on multi-stage territorial stratified selection.
Stratified sampling is a probability sampling technique wherein the entire population is divided into different subgroups or strata, then randomly selects the final subjects proportionally to its population from the different strata.
If the multi-stage sample designs are applied, selection of primary sampling units (PSUs) forms the first stage and the selection of, for example, households within selected PSUs follows at the second stage of sampling. For the second stage, simple probability sampling is used.
Depending on specific conditions in the country, other sample design models are possible and have to be discussed with the WVSA Scientific Committee.
Every national team develops their sampling model according to the country peculiarities (geographical and administrative division, size of urban and rural population, available statistica data etc.) and existing WVSA methodological requirements.
There are three main groups of countries in the WVS-7 depending on their sample size:
- the minimum acceptable sample size - i.e. the number of completed interviews which are included into the national data-set in vast majority of countries is 1200;
- for countries with the population of less than 2 million a sample size of 1000 respondents is acceptable;
- countries with bigger population and greater population distribution (Russia, China, Brazil, USA etc.) should aim at bigger sample size of minimum 1500 respondents.
The WVSA relies on national team’s experience with regard to the optimal national sample size in a particular country.
The main method of data collection in the WVS survey is face-to-face interview at respondent’s home / place of residence. Respondent’s answers could be recorded in a paper questionnaire (PAPI) or via CAPI (Computer Assisted Personal Interview).
Application of other methods of data collections (Internet panel, postal interview, telephone interview) is possible though requires justification by the survey team and assurance of the 95%+ population coveerage. Possiblity of application of any alternative survey methods rather than PAPI and CAPI is considered on the country by country basis and requires approval of the WVSA Scientific Committee. Same requirements and procedure are valid for application of mixed mode surveys.