The basic procedure in survey is that people are asked a number of questions on that aspect of behavior which the sociologist is interested in. A number of people carefully selected so that their representation of their population being studied are asked to answer exactly the same question so that the replies to different categories of respondents may be examined for differences.