Table 3.
Model number | Association of interest | Accuracy differential by laboratory | General number of labs (R) | R = 1 | R = 2 | R = 3 |
1 | Observed vs. true exposure status: EX | Yes | 2R + 2 | 4 | 6 | 8 |
2 | Yes* | 4R + 2 | 6 | 10 | 14 | |
3 | No | 4 | 4 | 4 | 4 | |
4 | No* | 6 | 6 | 6 | 6 | |
5 | Disease vs. true exposure status: DX | Yes | 2R + 3 | 5 | 7 | 9 |
6 | No | 5 | 5 | 5 | 5 | |
7 | Disease with observed exposure status: DE | Yes | 2R + 1 | 3 | 5 | 7 |
8 | No | 3 | 3 | 3 | 3 |
* error rates are (additionally) differential by disease status
D- disease; X-true genotype status; E-genotype status measured by laboratory test.