Abstract
Randomized experimental design is seen by many researchers in the social sciences as the gold standard of causal inference, but as a method, it is blind to the mechanistic processes that lead from cause to outcome. While experimental researchers are highly conscious of the mechanisms underlying their theorization and inference, they are mostly implicit in their research design and analysis. We argue that process tracing used as an adjunct method alongside an experiment can be used to make the mechanistic assumptions explicit in experimental design and analysis. Through this combination, experimental researchers can become more conscious about causal mechanisms in their theorization, leading to better experimental designs, more effective ways to discuss problems with causal heterogeneity, and even strengthening the internal and external validity of findings.