This “potentially causal variable” is called, traditionally, the “independent variable.” With the current example, that would be the coffee.

This paper makes explicit and takes issue with the bizarre view, which is unfortunately prevalent among social scientists, that causal relations are features of models only. There are some good ...