1. Given this information, what would be the overall sample size?

A = 100; B = 265 (using table B-1 for assumed current of coverage of 85% and precision of +/-5%)

C = 4 (using table C for m=10 and ICC=0.333)

D = 5; E = 1.18

Overall sample size will be 105,600

2. How many households would have to be visited?

Total number of households to visit = 621,177

3. What do you think of this estimated sample size?

This sample is big but is a result of requiring coverage results for each of the 100 districts and at a high level of precision. The result is a large number of clusters per strata, with an even larger number of households per cluster

4. How feasible will it be to conduct this survey?

The time and thus cost required to conduct this survey will be enormous.

5. What are the trade-offs in terms of time, money, and quality of survey implementation?

To conduct the survey in a lesser time, a big survey team has to be deployed. This, though, will not reduce on the overall cost.

If cost is an area of concern to the Steering Committee, I would propose relaxing of the precision level to +/-10%. This would result in a substantial reduction in the sample and subsequently the overall time and costs to conduct the survey.