Six questions, one figure

A district health slide carries one number.

Penta3 coverage, 2024: 78%

Nothing else. No denominator, no reporting rate, no period note. You have the extract it was computed from and forty minutes before the meeting.

The file

vaccination-coverage-2024.v1.csv — 2,736 rows of monthly aggregate reporting from 38 facilities across six antigens, with a submission flag and a target population. Synthetic and DHIS2-shaped.

The task

Answer the six interrogation questions from lesson 8, each with code and a one-sentence conclusion.

1. Is it a data element or an indicator? Say which, and give the numerator and denominator expressions the figure implies.
2. What is the denominator’s year, and how does it aggregate over periods? Show what target_population does when summed across twelve months, and what the figure becomes.
3. What was the reporting rate? Overall and by month. Say what it does to the status of the 78%.
4. Which org unit level? Say which level the extract is at, and whether a district figure computed from it can double-count.
5. Was the period complete? Say what the extract can and cannot tell you here.
6. What is it being compared to? The slide implies a comparison. Name what it would have to be comparable to, and whether it is.

Reproduce the 78%

Before answering, work out how the figure was produced. There are at least three routes to a number near 78% and they are not equivalent — one of them is the defensible one and the others are arithmetic accidents that land close.

Show which route you believe was used and why.

Two questions the extract cannot answer

Two of the six cannot be answered from this file at all. Identify them, say precisely what field or metadata would be needed, and write the request you would send to the HMIS focal point — one sentence each, naming the endpoint or the field.

That request is part of what you hand in. An analyst who says “I cannot tell” and stops has done half the job; one who says what would settle it has done all of it.

Three reference points

Computed on reported facility-months only, penta3 coverage for the year is about 77.5%. Computed against every facility’s full annual target, it is 60.8%. Computed by summing the monthly target population across the year, it is 6.5%.

If your reproduction does not land on one of those three, work out which denominator you built before going further.

The questions to answer in prose

Three sentences each.

1. Two of your three routes to the figure give 77.5% and 60.8%. Say what each one assumes about the 107 facility-months that did not report, and which assumption the slide is implicitly making.

2. The reporting rate is 29% in August. Explain what that does to an annual figure, and whether it makes the 78% too high or too low.

3. The slide will be compared to last year’s. State the two things you would need to know about last year’s figure before allowing that comparison, and what you would write on the slide if you could not get them.

What to hand in

A Python script producing a six-row interrogation table — question, answer, evidence, and whether the extract could answer it — plus the seven-field provenance block from lesson 8 as a text file, with unknown written explicitly in any field the extract cannot fill.

How to know you are done

Every row of your interrogation table cites a number computed from the file, or says explicitly that the file cannot supply one. Your provenance block has no silently blank fields — an honest unknown is a pass, an empty line is not.