Table 3. A 2×2 table for the effect of dichotomous exposure E on dichotomous outcome D, restricting to a level of a variable C, given causal relationships shown in Figure 3.

Because C is affected by E only, this is simple random sampling within levels of E; we observe a fixed proportion of individuals within each value of E (in this case, some fraction f when E=1, and g when E=0).

	D=1	D=0	Total
E=1	fA	fB	fA+fB
E=0	gC	gD	gC+gD
Total	fA+gC	fB+gD	fA+fB+gC+gD