Statistical inference for causal effects in clinical psychology: Fundamental concepts and analytical approaches

A central problem in clinical psychology is how to draw statistical inferences about the causal effects of treatments (i.e., interventions) from randomized and nonrandomized data. For example, does exposure to childhood trauma really impair cognitive functioning in adults, or does a new medication for treating a psychological condition reduce the occurrence of negative physical health outcomes relative to existing treatments? This chapter describes a general approach to the estimation of such causal effects based on the Rubin Causal Model (RCM). Under this framework, causal effects are defined in terms of potential outcomes and inferences are based on a probabilistic assignment mechanism, which mathematically describes how treatments are given to units. Frequentist and model-based forms of statistical inference for causal effects in randomized experiments and observational studies are presented, and the application of this approach to a number of data collection designs and associated problems commonly encountered in clinical research is discussed.