The Limitations of Before and After Designs

Abstract

Quasi-experimental research methods widely used in health and social science research are often used to make causal inferences. Yet (with certain exceptions, for example, the regression discontinuity design, Cook and Campbell, 1979) their designs are often not sufficiently reliable to do so. Many quasi-experiments cannot ‘design-out’ potential bias, unlike randomised controlled trials. In this chapter we discuss the particular problems of the before and after study.