Question 1. What additional information would you need? Post your information request in the comments below. (Additional information needed will be posted on Wednesday of Week 3.)

It would help to refine the sample size calculation if we had more accurate data on the average HH size (we assumed 6), on the real penta3 vaccination coverage (we assumed 85%), and a realistic estimate based on experience of the average number of HH that can be visited in a day in most parts of this country (we assumed 75 HH).

Question 2. Based on this new request, what would be the overall sample size?

For a survey with 10 strata, using the same figures we used in the week 2 assignment, the sample size calculation would be as follows:

Total interviews:10,560

Total HHs visited: 74,552

Total HH/stratum: 7,455

Total clusters: 1,060

Total clusters/stratum: 106

Total HHs/cluster: 71

Question 3. What decision would you make regarding the survey scope and objectives? What is your rationale for this decision?

This huge sample size would require a level of resource investment and management capacity that may not be practical or feasible in this country under current circumstance. In order to achieve a lower sample size, with the consequent reduction in cost and duration of the survey, but still generate useful information, a classification design could be selected. This would generate results which will identify provinces that demonstrate low performance and require correction action to raise vaccination coverage, but will not generate a province-specific coverage estimate. However, the provincial results can be combined during data analysis to generate a national coverage estimate with relatively high precision.