The experimental method

This is the main method used by natural scientists. Experimentation normally involves the testing of a hypothesis about the relationship between an independent variable (cause) and a dependent variable (effect). Experiments are usually set up so that the scientist controls the introduction of possible independent variables. Used more by psychologists than sociologists, e.g. see Milgram, Zimbardo. In the natural sciences, such control is enhanced by use of a laboratory. Any change in the participant’s behaviour should be the result of the change introduced by the experimenter. Interpretivist sociologists note that the experimental method is rarely used in sociological research for both practical and ethical reasons:

• Practical reasons: Sociologists can never be sure that behaviour is caused by the social phenomena they are interested in. For example, people usually know they are taking part in an experiment. Their performance may be distorted by their desire to impress the experimenter. Moreover, only a limited number of social conditions can be re-created in the laboratory.
• Ethical reasons: Some sociologists argue that it is morally wrong not to tell people they are part of an experiment or to expose them to adverse social conditions. E.g. media research on children, see Bandura.

These difficulties have led to some sociologists adopting variations on the experimental method. The comparative method is an experimental method which uses the social world as the laboratory. The researcher, usually using official statistics, compares a before and after situation in a social group where a change has taken place with one where it has not. I.e. Durkheim’s suicide.

Interpretivist sociologists have used field experiments – a qualitative examination of particular social contexts in order to explore the interpretations which underpin everyday interaction. However, these too involve some degree of manipulation.

Positivists note that this type of experiment may suffer lower levels of reliability. I.e. Rosenthal and
Jacobson