Survey research aims for a systematic and comprehensive collection of information about the attitudes, beliefs, and behavior of people. The most common means of data collection are face-to-face interviews, self-administered questionnaires (for example, by mail, in a classroom, or online), telephone interviews. Typically, the questionnaire or the interview schedule is set up so that the same questions are asked of each respondent in the same order with exactly the same wording and the validity of surveys is dependent on the design of the questions asked (Babbie, 2017).
Many surveys put questions to a random sample of the population; this population may be at the national, state, or city levels, or it may be the population of a large university.
The use of random sampling allows researchers to generalize the results of a survey to the population from which respondents to the survey come.
Survey studies tend to be larger than is typically the case in observational or experimental studies. Usually, data are collected at one time, although a survey approach can be used to study trends in opinion and behavior over time. Because of its ability to cover large areas and many respondents and to generalize to the population, the survey method is the dominant method of data collection in sociology.
Survey methods, like other research methods, have their pitfalls. A potential difficulty involves a survey’s response rate. To be able to generalize the results of a random survey to a population, it is essential that the sample maintain its representativeness, which may be affected severely when a large number of respondents fail to participate in the study. In addition to a subject’s refusal to participate, other factors affect the response rate. These factors include the inability of the subject to understand the question, the possibility that the subject may have moved or died, and a possible physical or mental disability of the subject.
An illustration of the use of survey methods can be seen in the efforts of the U.S. Department of Justice to gain a more accurate measure of the extent of crime in the United States. For years, both law enforcement agencies and criminologists have had to rely on official records compiled by such agencies as the Federal Bureau of Investigation (FBI) to measure the amount of crime. However, there have been concerns about the accuracy of these reports, and criminologists generally believe that officially recorded crime statistics are a far better indicator of police activity than they are of criminal activity (Barkan, 2018).
For the past several decades, the U.S. Department of Justice has conducted the sophisticated National Crime Victimization Survey (NCVS) in an attempt to supplement official crime records and to overcome some of the problems of accuracy therein (Truman and Morgan, 2016). This survey interviews tens of thousands of Americans annually to determine how many crimes have been committed against them. In addition to determining the volume of crime, the surveys are also used in developing a variety of information on crime characteristics and the effects of crime on the victims—victim injury and medical care, economic losses, time lost from work, victim self-protection, and reporting of crime to the police.
Two advantages of the NCVS help make it superior to self-report studies and the FBI’s Uniform Crime Reports (Barkan, 2018). First, the NCVS by design seeks out information about crimes rather than waiting for victims to report them as is the case with the Uniform Crime Reports. Second, the NCVS relies on a large representative sample of the U.S. population, so that its results can be generalized to the entire population. The NCVS is very expensive, and its respondents might for various reasons overreport or underreport the crimes they have suffered. Still, it has yielded highly useful information since it was established several decades ago.
Researchers have used survey methods in a variety of cross-cultural studies dealing with knowledge and opinion about law and legal issues. Some of these studies reveal interesting findings. For example, a European study once asked residents of Poland, the Netherlands, and Germany whether they thought people should obey the law. They found significant national variations; more Germans (66%) than Poles (45%) or Netherlander (47%) answered yes to this question (Friedman and Macaulay, 1977:216).