The Role of the Experiment in the Social Sciences

Abstract

Laboratory research is often criticized for its artificiality and its lack of random sampling. These criticisms are examined within the context of two approaches to social scientific knowledge: the descriptive and the formal. The goal of a descriptive theory is the explication of social events and processes as they exist at one point or specific points in time, while the goat of a formal theory is the development of law-like statements. We argue that if the concern is the development of descriptive theory the experiment can legitimately be criticized on the basis of artificiality and lack of random sampling. We argue further, however, that if the concern is the development of formal theory, artificiality is an advantage and random sampling is not a requirement.