A survey is a poll in which the researcher gathers facts or attempts to determine the relationships among facts. Surveys are often done when the researcher wants to describe, compare, and predict knowledge, attitudes, and behavior. For example, a community survey might describe and compare such things as income. educational level, and type of employment in regard to people's attitudes about a juvenile curfew ordinance that prohibits adolescents from being out on the streets at certain nighttime hours. Researchers frequently select a representative sample (a small group of respondents) from a larger population (the total group of people) to answer questions about their attitudes, opinions, or behavior.
For example, if the larger population consists of 10,000 people, 51 percent of whom are female, with 30 percent over age 35, a representative sample will have fewer people (perhaps 1.000) but must still consist of 51 percent females, with 30 percent over age 35 (Fink, 1995). Respondents are persons who provide data for analysis through interviews or questionnaires.
The Gallup and Harris polls are among the most widely known large-scale surveys; however, government agencies such as the u.s. Census Bureau conduct a variety of surveys as well. Unlike many polls that use various methods of gaining a representative sample of the larger population, the Census Bureau attempts to gain information from all persons in the United States. The decennial census occurs every 10 years, in the years ending in ·0:' The purpose of this census is to count the population and housing units of the entire United States.
The population count determines how seats in the U.S. House of Representatives are apportioned; however, census figures are also used in formulating public policy and in planning and decision making in the private sector. The Census Bureau attempts to survey the entire U.S.
population by using two forms-a "short form" of questions asked of all respondents, and a "long form" that contains additional questions asked of a representative sample of about one in six respondents. Statistics from the Census Bureau provide information that sociologists use in their research. An example is shown in the Census Profiles feature: "How People in the United States Self-Identify Regarding Race." Note that because of recent changes in the methods used to collect data by the Census Bureau, information on race from the 2000 census is not directly comparable with data from earlier censuses. Surveys are the most widely used research method in the social sciences because they make it possible to study things that are not directly observable-such as people's attitudes and beliefs-and to describe' a population too large to observe directly (Babble, 2004). Let's take a brief look at the most frequently used types of surveys.