Causal inference is essential in science studies, yet many publications lack methods to substantiate causal claims. Structural causal models, often represented graphically with directed acyclic graphs, make causal assumptions transparent and improve communication. We illustrate the application with a hypothetical model of Open Science.