Sociologists usually reserve the term experiment for studies in which the researcher manipulates one or more key independent variables; that is, for studies in which the experimenter controls the decision as to which subjects are exposed to what level of the independent variable. Although experiments comprise a minority of sociological research studies, they are still quite common. Laboratory versions are traditional in much of sociological social psychology (examples are Bonacich 1990; Molm 1990). Field experiments are prominent in program evaluations, particularly in applied areas such as education (e.g., Slavin and Karweit, 1985), criminal justice (e.g. Rossi et al. 1980; see also Zeisel, 1982), and even large-scale social policy efforts such as studies of a proposed guaranteed minimum income (Hannon et al. 1977). Questionnaire experiments (such as Schuman and Presser 1981) test various methodological questions, while "vignette" experiments (such as Rossi and Nock 1982) have become popular approaches to the study of attitudes.

The primary...