The social sciences are in need of an alternative to the variance approach to causal inference, which—because it requires a counterfactual—restricts the claim of valid inference to conclusions drawn from experimental and quasi-experimental methods. Process theory proposes an alternative by way of demonstrating the presence of observable characteristics of the causal mechanism, a method that, while accepted in principle, has proven elusive in practice.  It is suggested that system dynamics can improve the process approach, and in so doing open a path for itself to wider application in the social sciences.   Educational research is suggested as a place to start, and results from two models drawn on topics from that field are offered as examples.