Table 2.

Value of the conditional odds ratio, θ^c_x, (top half) and corresponding estimates of actual power (bottom half), based on a case–control study with n₀ = n₁ = 1250, for an assumed marginal effect size of θ^m_x = 1.3, under various scenarios for the strength of confounding. The (marginal) outcome and exposure prevalences are Pr(Y = 1) = 0.05 and Pr(X = 1) = 0.2, respectively, and the (conditional) confounder prevalence among the unexposed is fixed at Pr(Z = 1|X = 0) = 0.2

	Confounder/exposure odds ratio, ϕxz
Confounder effect, θcz	0.33	0.50	0.67	1.00	1.50	2.00	3.00
Conditional odds ratio, θcx
0.33	1.19	1.22	1.25	1.30	1.38	1.44	1.57
0.50	1.22	1.24	1.26	1.30	1.35	1.40	1.49
0.67	1.25	1.26	1.27	1.30	1.33	1.36	1.42
1.00	1.30	1.30	1.30	1.30	1.30	1.30	1.30
1.50	1.38	1.36	1.34	1.30	1.26	1.23	1.18
2.00	1.45	1.41	1.37	1.30	1.23	1.17	1.10
3.00	1.59	1.49	1.42	1.30	1.18	1.09	0.98
Estimated actual power
0.33	41	52	62	77	90	96	100
0.50	51	59	66	77	88	94	98
0.67	61	66	71	78	84	89	93
1.00	76	77	77	78	78	77	76
1.50	90	88	84	78	66	55	38
2.00	96	94	89	76	55	36	15
3.00	100	98	94	76	38	15	5