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

105,759 interviews to be completed if the goal is district level coverage details

(Note that this is many fold greater than what would be required if we settled for national level coverage…if national level we would need 1058 interviews.

2. How many households would have to be visited?

The number of households to be visited would be 622,112.

The reason this is so much larger is because 1) not every house will have a child in the target age range, and 2) some households with an eligible child will not be present or not consent to interview.

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

This is a very large sample size and would require a tremendous amount of resources to be completed.

I would suggest considering a methodology that would be less money, fewer teams, and less time to implement.

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

Not feasible, as to visit over 500K homes would be incredibly resource intensive.

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

A larger number of interviews nad number of households to be visited would mean that much greater quantity of money, technical support / interview teams, and time for implementation (and analysis) would be required. The strain of a very elaborate and high number of interview design could lead to quality being lower, as there may be rushed teams and analysis or less accurate recording of data due to the performance requirements. On the otherhand, if the survey is too small, for example if only giving national level data (1 strata), then there is also a quality cost…for there will be no way to parce out the granularity of data. This means we wouldn't be able to determine areas of higher risk / lower coverage.