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

The total completed interviews or overall sample size would be 106 000.

How many households would have to be visited?

It would be required to visit 636 000 households in total. Number of households to visit in each stratum: 6360. Number of clusters per stratum: 106. The total households to visit per cluster: 60.

What do you think of this estimated sample size?

The calculated sample size of 106 000 would be representative sample of target population and coverage estimate would be accurate in terms of precision and validity if such a survey would be implemented. However sample size is too far too big and would be hard to achieve it and conduct such a big survey.

How feasible will it be to conduct this survey?

Such survey would require a lot of efforts in terms of human and financial resources. The management and the logistic of the survey also would be complicated. It would require employing many field workers (data collection teams, their supervisors and etc.); their preparation for survey and training would be also expensive. The survey timeline also would be extended and require more time to complete field work, data analysis, drafting preliminary and final reports and etc. In conclusion, such survey is not very feasible.

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

The steering committee has to revise and discuss again the level which survey can be conducted in a more meaningful way in terms of shorter time, less expensive and better quality of implementation. One of the options, which can be discussed, is to conduct survey at province level in ten regions of Harmonia. Another option would be to select randomly some (not all as in current example) districts within provinces and conduct survey in selected districts. The steering committee can discuss the possibility of decreasing the level of precision (e.g. instead of 5% choose 7% or 10% precision); in this case the point estimate will be less precise, but still valid. One more option can be, not estimate, but classify vaccination coverage (when threshold is chosen and calculated coverage is classified as passing, failing or intermediate).