A sample is as a small group or subject drawn from a larger population and it should be as representative as possible of this larger population. To this end a number of random sampling methods are used; otherwise the sample will be biased.
The main sampling methods which are designed to create a representative sample are the following, which have these principles and characteristics:
1) Sampling using a household survey. This kind of sampling is based on taking a survey of households drawn from the total population of a country or region, since it is unrealistic to list everyone in the country or region. Such a sample is designed so everyone has a equal chance of being drawn from a sampling process, and for this purpose, commonly a multi-stage and clustered sampling approach is used. In the multi-staged approach, a sample is divided into sections, such as the regions to be sampled, and subsamples are selected from electoral registers or streets in the area. Then every nth house on the street is selected, such as every third, fourth, or fifth house, to obtain the number of houses to be included. Then, the person in the household matching the criteria sought, such as head of household or oldest female, is included in the survey.
2) Sampling for a site, user, or visitor survey. This kind of sampling is based on drawing a sample from the users or visitors to a leisure or tourism site, such as a nature park or amusement park. This sampling can be done by the interviewer being stationery and interviewing a sampling of users or visitors as they pass by, such as if the interviewer is at the entrance, exit, or other place where all users or visitors go by. Alternatively, the interviewer can be mobile and interview a sampling of users who are stationery, such as when visitors are in a campsite or barbecue area. In both cases, randomness can be introduced by the interviewer following certain rules, such as a stationery interviewer interviewing every nth person who passes a particular location or a mobile interviewer interviewing every nth person he or she passes on a particular route.
3) Quota sampling, usually in a street survey. This kind of sampling is based on seeking a certain number or quota of people of a certain type, so they will represent a certain percentage of the targeted population. This approach is used to have a broad sampling of different type of people in a population, as long as the breakdown of that population is known. This method might be used when it is not possible to get a representative sampling by other means, such as if certain members of the population are less receptive to surveys or may be more difficult to find. An example might be lower class individuals who are less likely to be home when an interviewer comes to do an interview. A sampling approach is a street survey, when the interviewees are contacted on the street, such as when they go to a park or are shopping in a mall or shopping area on a street.
4) Sampling for a mail survey. This kind of sampling is based on sending a survey to people by mail, or today, by e-mail, and recipients are asked to fill in the survey through self-completion and return it, usually in a pre-addressed envelope which is often pre-stamped, or they are asked to send an e-mailed survey to the return e-mail or to a link which the recipient fills in the survey online.
