Table 1.
Sample size | No. in the sample non-poor | Probability of detecting health facilities with 80% poor as adequate | Probability of detecting health facilities with 50% poor as inadequate | Provider Risk | Consumer Risk | Total classification error |
(n) | (d) | (a) | (b) | (1-a) | (1-b) | (1-a)+(1-b) |
8 | 0 | 0.17 | 1 | 0.83 | 0 | 0.83 |
1 | 0.50 | 0.96 | 0.50 | 0.04 | 0.54 | |
2 | 0.79 | 0.83 | 0.21 | 0.17 | 0.38* | |
3 | 0.94 | 0.64 | 0.06 | 0.36 | 0.42 | |
12 | 0 | 0.07 | 1.00 | 0.93 | 0.00 | 0.93 |
1 | 0.28 | 1.00 | 0.73 | 0.00 | 0.73 | |
2 | 0.56 | 0.98 | 0.46 | 0.02 | 0.48 | |
3 | 0.80 | 0.93 | 0.21 | 0.07 | 0.28 | |
4 | 0.93 | 0.81 | 0.07 | 0.19 | 0.27* | |
5 | 0.98 | 0.61 | 0.02 | 0.39 | 0.41 | |
14 | 0 | 0.04 | 1 | 0.96 | 0 | 0.96 |
1 | 0.20 | 1 | 0.80 | 0 | 0.80 | |
2 | 0.45 | 0.99 | 0.55 | 0.01 | 0.56 | |
3 | 0.70 | 0.97 | 0.30 | 0.03 | 0.33 | |
4 | 0.87 | 0.91 | 0.13 | 0.09 | 0.22* | |
5 | 0.96 | 0.79 | 0.04 | 0.21 | 0.25 | |
19 | 0 | 0.01 | 1 | 0.99 | 0 | 0.99 |
1 | 0.08 | 1 | 0.92 | 0 | 0.92 | |
2 | 0.24 | 1 | 0.76 | 0 | 0.76 | |
3 | 0.46 | 1 | 0.54 | 0 | 0.55 | |
4 | 0.67 | 0.99 | 0.33 | 0.01 | 0.34 | |
5 | 0.84 | 0.97 | 0.17 | 0.03 | 0.20 | |
6 | 0.93 | 0.92 | 0.07 | 0.08 | 0.15* | |
7 | 0.98 | 0.82 | 0.02 | 0.18 | 0.20 | |
28 | 5 | 0.50 | 1 | 0.50 | 0 | 0.50 |
6 | 0.68 | 1 | 0.32 | 0 | 0.32 | |
7 | 0.81 | 0.99 | 0.19 | 0.01 | 0.20 | |
8 | 0.91 | 0.98 | 0.09 | 0.02 | 0.11 | |
9 | 0.96 | 0.96 | 0.04 | 0.04 | 0.08* | |
10 | 0.99 | 0.90 | 0.01 | 0.10 | 0.11 |
* - Optimal decision rule for a sample size.
Source: Adopted from Valadez 1991, p:73.