Table 2.

Example simulation results based on Kim et al. (2009) showing that in models conditioning on the exposure estimate, health effect estimate properties vary by the predictability of the underlying exposure surface as well as the approach to exposure prediction

True Exposure	Fitted Exposure (R 2)	Bias2	Variance	Mean Square Error	Coverage probability of 95% confidence intervals
Least predictable (shortest range)	True	0	9	9	0.95
	Nearest Kriging (0)	327	23	350	0.03
		342	778	1,120	0.58
	True	0	31	31	0.95
	Nearest Kriging (0.20)	33	58	91	0.76
		1	734	735	0.74
	True	0	69	69	0.95
	Nearest Kriging (0.40)	30	125	155	0.87
		1	426	427	0.89
Most predictable (longest range)	True	0	56	56	0.96
	Nearest Kriging (0.47)	34	105	139	0.85
		0	153	153	0.92