Abstract

From a methodological point of view, experiments are the gold standard for causal inference. In theory, active manipulation of the treatment by the researchers secures causal ordering, random assignment avoids biases due to selection, and the use of control groups in a tightly controlled setting helps to control for other nuisances. However, in practice, many things can go wrong, threatening the validity of experimental results. Careful planning and implementation of an experiment is thus important to address these challenges and pitfalls. Against this background, this entry provides information on planning and conducting experiments in the laboratory and compares laboratory experiments with experiments in other settings. The laboratory experiment is framed as a social situation consisting of interactions between experimenters and subjects before, during, and after the experiment. A particular emphasis is put on challenges caused by these social interactions.