Samples are drawn from the entire population of 18 years and older. The minimum sample is 1000. In most countries, no upper age limit is imposed and some form of stratified random sampling is used to obtain representative national samples. In the first stages, a random selection of sampling points is made based on the given society statistical regions, districts, census units, election sections, electoral registers or voting stations and central population registers. In most countries the population size and/or degree of urbanization of these Primary Sampling Units are taken into account. In some countries, individuals are drawn from national registers.
Data collection (Field work)
Following the sampling, each country is left with a representative national sample of its public. These persons are then interviewed during a limited time frame decided by the Executive Committee of the World Values Survey using the uniformly structured questionnaires. The survey is carried out by professional organizations using face-to-face interviews or phone interviews for remote areas. Each country has a Principal Investigator (social scientists working in academic institutions) who is responsible for conducting the survey in accordance with the fixed rules and procedures. During the field work, the agency has to report in writing according to a specific check-list. Internal consistency checks are made between the sampling design and the outcome and rigorous data cleaning procedures are followed at the WVS data archive. No country is included in a wave before full documentation has been delivered. This means a data set with the completed methodological questionnaire and a report of country-specific information (for example important political events during the fieldwork, problems particular to the country). Once all the surveys are completed, the Principal Investigator has access to all surveys and data.
"The mode of data collection for WVS surveys is face-to-face interviewing. Other modes (e.g., telephone, mail, internet) are not acceptable except under very exceptional circumstances and only on an experimental basis. In any case, EC approval in writing is necessary for modes of data collection other than face-to-face interviewing." (Source: http://www.worldvaluessurvey.org/SebTest/wvs/articles/folder_published/survey_2005/files/WVS-rules_for_PIs_wave5.pdf)
A. Method of Sampling
The preferred method of sampling for WVS surveys is the full probability sample. However, recognizing that the very high cost -in terms of finances, manpower and time- of full probability samples may prove to be prohibitive in some cases, WVS allows quota sampling provided that the following principles are strictly adhered to:
- Selection of Primary Sampling Units must be probabilistic (and preferably Probability Proportionate to Size).
- Selection of first stage clusters within PSUs must be probabilistic (and preferably PPS).
- Quota sampling should be used only within reasonably small sized clusters that have been selected probabilistically.
Whether the sampling method is full probability or a combination of probability and quota, the minimum number of PSUs is 30. A design with less than 30 PSUs is not permissible.
B. Sample Size
The minimum sample size (i.e., the number of completed interviews) is 1,000. However, given the fact that in most designs the “effective sample size” (sample size net of design effects) is lower than the actual sample size, larger sample sizes are strongly recommended if at all possible.
C. Non response
Non-response is an issue of increasing concern in sample surveys. Investigators are expected to make every reasonable effort to minimize non-response. More specifically,
- In countries using a full probability design, no replacements are allowed. Pis should plan on as many call-backs as the funding will allow.
- In countries using some form of quota sampling, every effort should be made to interview the first contact.
- In any case, and as indicated below, a full report on non-responses is required.
WVS surveys are required to cover all residents (not only citizens) between the ages of 18 and 85, inclusive. PI’s can lower the minimum age limit as long as the minimum required sample size for the 18+ population is achieved."
WVS requires a complete explanation of proposed sampling procedures BEFORE the start of the fieldwork. The sampling plan must be approved by the EC in writing.
All deposited data has been made anonymous at the PI side and the archive deposited files have no means to trace the respondents.