In general, there are two methods of random sampling. The first is the “simple random sample method”, and the second is the “quota method”.
- With the simple random sampling method, the research organization uses randomly selected telephone numbers or e. e-mail addresses (for telephone or Internet surveys), or randomly selected addresses and names from databases such as electoral registers (for “face-to-face” surveys) are visited. Then the survey of respondents is carried out in an adequate way.
- The quota method represents a sample that implies fulfilling the quota according to the given characteristics, such as, for example. years and a half, during which the interviewer then looks for different people who meet those characteristics. Sampling quota can also be used in telephone surveys to select a person from a household to be interviewed in order to speed up the survey process.
Are survey companies doing anything else to ensure a representative sample?
Yes. Even if sampling is done well, there are several reasons why surveys may contain too many respondents from one group and too few from another. It is common for research companies to ask respondents some more personal questions, about themselves and not just solely about their attitudes. This information is then used to compare it with official statistics, such as census data. In that case, the proportion of a certain population (young, old…) in the sample can be adjusted to match the profile of the population under investigation.