Any scientific process begins with description, based on observation, of an event or events, from which theories may later be developed to explain the observations. In psychology, techniques used to describe behavior include case studies, surveys, naturalistic observation, interviews, and psychological tests.
Case studies. A case study is a method of obtaining information from the detailed observation of an individual or individuals. Much information about behavior and mental processes has been obtained through such studies of individual clinical cases. (Sigmund Freud, for example, formulated psychoanalytic theory after many years of treating and studying patients with emotional problems.) Although valuable information about certain types of problems may be obtained by this method, the procedure is time consuming, and it is difficult to obtain data from a broad sampling of people.
Surveys. In a survey, people from a wide sample are asked questions about the topic of concern. The Kinsey survey on sexual behavior is a well‐known example. Surveys can supply useful information, but they have their problems and limitations. For example, the people who respond may not be representative of the population in general, or those polled may be reluctant to respond to questionnaires or to answer them accurately.
Naturalistic observation. In another approach to gathering information, naturalistic observation, people or animals are observed in their everyday behaviors, and their behaviors of interest are documented. For example, valuable information on wild animals, such as lions, has come from studying them in their natural habitats as opposed to observing them in a zoo because their zoo behavior may be quite different from their natural behavior. Similarly, the behavior of a human in a home environment may differ considerably from that in a laboratory.
Psychological testing. Many standardized procedures ( tests) have been developed to measure specific behaviors or characteristics of organisms. Most of us have been subjected to such tests—for example, the intelligence, aptitude, and achievement tests used to predict behaviors. To be useful, tests must be both reliable and valid.
Correlation. Correlation, a statistical measure of a relationship between two or more variables, gives an indication of how one variable may predict another. The descriptive techniques discussed above permit a statement, in the form of correlations, about that relationship. However, correlation does not imply causation; that is, simply because two events are in some way correlated (related) does not mean that one necessarily causes the other. For example, some test data indicate that boys receive higher math‐aptitude scores on college entrance exams than girls, indicating a correlation of gender with mathematical ability. But before concluding that gender determines mathematics aptitude, one must demonstrate that both the boys and the girls in the study have had the same mathematics background. Some studies have shown that girls are discouraged from taking or at least not encouraged to take more than the minimum mathematics requirements. Such discrepancies in mathematical accomplishment may also arise in the home—for example, from a parental belief that a girl does not need much mathematical training to be a good wife and mother.