There are advantages and disadvantages of using different sampling methods in social research. This article describes them.
1. The advantage of the household survey method is that it provides a sample of an entire population of a country or region. Such a survey might be good to use if one wants to know what people throughout the area are thinking, such as obtaining opinions on an issue to help a politician take a position on a bill for a new park or park improvements. This type of survey might also be good for the government to use in learning about people’s preferences about a recreational facility for the area, such as whether they would prefer a zoo or nature park and who would be interested in going.
A disadvantage of the household survey is that it is very time-consuming and laborious to try to obtain a sampling of everyone in the whole population. Also, while most organizations doing national surveys seek to use multi-stage and clustered sampling to make the process of gathering the sample more manageable, there can be problems in doing this. For example, if the researchers try to reduce costs by interviewing a certain number of people on the street and choosing every nth house, they could make the number of clusters too small to include a complete variety of people in the area. Then, too, the researchers might find that people in certain types of households might not want to participate in the survey, most notably those from lower-class homes, which would skew the sample to have a greater percentage of middle and income people represented than their percentage in the population.
2. The advantage of sampling for a site, user, or visitor survey is that one can obtain feedback from a population that is actually using a facility or visiting a site. Another advantage is that one can control for randomness by having a trained interviewer select the nth person passing the interviewer or the nth person the interviewer passes after concluding an interview with the previous selected person. This approach can also be adapted for a quota sample, where the interviewer is asked to get the nth person for a number of categories, and as each category is filled, the interviewer stops looking for someone for that category.
A disadvantage of this approach is that an interviewer may fail to follow procedures about selecting interviews and might select someone who seems convenient or avoid other types of interviewees who are more difficult to find or interview. Another disadvantage is that the types of individuals who come at different times of the day may vary, such as having more families with kids coming during the day and teenagers coming at night to an amusement park. As a result, the sampling could become unrepresentative unless it is weighted or a quota sampling approach is used to balance out the different types of people. Another disadvantage could be the problem of non-response, if people entering or leaving the site don’t want to be bothered – those on the way in because they are eager to enjoy the site and those on the way out because they are eager to get home. Still another disadvantage if a questionnaire is used is that many respondents may not complete the questionnaire and those who do may be unrepresentative of the whole group, particularly if there is a high return rate. Also, personnel at the site may not carefully hand out the questionnaires, and it may be difficult to supervise this personnel, since their priority is on helping visitors and guests, rather than helping with a survey.
3. The advantage of sampling for a street survey is that it can be good to find out about who is using a particular street, shopping, or tourist area, much like a site, user, or visitor survey. Such a street survey can be combined with a quota sampling approach to obtain information on different types of users or visitors who come to the site at different times, if the background of the population in the area is known.
However, the disadvantage of a street survey, much like a site, user, or visitor survey, is that people on the street may be busy and might not want to take the time to participate in an interview or may not be interested in filling out a questionnaire. As a result, the sample will become biased as a result of differences between those who answer questionnaires or fill out a questionnaire and those who don’t. Another disadvantage in some areas is the possible danger of stopping and interviewing people on the street. For instance, in an inner city area, people may be suspicious of outsiders, thinking they might be government workers or undercover cops, so they don’t want to answer questions, and there could be a danger to interviewers or those distributing questions, since they might be threatened or attacked. Another disadvantage is that a quota method can’t be used if one doesn’t already know the background of the people in the area to draw up quotas. Another disadvantage is if people don’t want to take the time to respond to a survey on the street, very few will respond to a take-home and return the survey, so it will have a low response rate. An attempt to do a quota survey by e-mail has the disadvantage in that not only may there be a low-response rate, but certain types of people are less likely to use email, such as the lower-income individuals. However, weighting is often used with such surveys to get a sample that corresponds more closely to the percentage of the larger population.
4. The advantage of sampling for a mail survey using the regular mails is it can be a large completely random survey, since the mail generally goes to everyone in the population. However, a disadvantage is that it can be expensive to send out a mail survey to a large group, especially since postal rates have been rising. Another disadvantage is that there is normally a high non-response rate to mailed questionnaires, and those who respond may be different from those who do respond. While the cost factor can be reduced substantially with an email survey, the disadvantage of doing an e-mail survey is that those with e-mails or those who are most likely to respond to emails may be different from the population that doesn’t respond as well as to the small population that doesn’t have emails, usually because of having a lower income.