Unmatching data aggregation: Need insights on applied weights

I am writing a question about data aggregation.

I am currently analyzing ANC4 coverage.
I assume that if province A includes districts a, b, c, and d,
the expected equation should be A = (a+b+c+d)/4.
However, in the actual data, the values on both sides don’t match.
It seems like weights were applied during aggregation process.

I am wondering how these weights were determined and what the reasoning behing their application is.
How can I verify the rationale for the applied weights?

Your assistance with this matter would be greatly appreciated.
Thank you!

Hi Tammy Kim
I am not 100% sure if I understand your question.
Normally the denominator for ANC Coverage is population expectant pregnant. Each district would have its own population figure. The coverage for District a = 85%, for district b = 63%, district c = 72%, district d = 81%. When calculating the provincial coverage, it will sum up all the population figures and divide by the ANC 4th visit. View the raw data for population pregnant (denominator) and ANC 4th visit (numerator) at provincial level and calculate manually. Do the manual values match the displayed values.

1 Like