search

athowes / multi-agyw

Spatio-temporal estimates of HIV risk group proportions for AGYW across 13 priority countries in sub-Saharan Africa
https://athowes.github.io/multi-agyw
MIT License
2 stars 2 forks source link

Account for survey under-reporting #39

Closed athowes closed 2 years ago

athowes commented 2 years ago

Olli suggested that AGYW tend to under-report sexual activity, and asked if this is something which we have taken account of. I am aware of this issue, but wasn't sure that any bias term we use here could be informed by data, as, unlike in other work, we do follow the same cohort in order to cross-reference their answers. However, I will speak to Kinh about this to see if he thinks that there is anything that we can do in this work to account for under-reporting. At the least we should include a paragraph explaining why we have chosen not to adjust our estimates.

One point of discussion could be to what extent we imagine the under-reporting to vary over each variable. The primary aim of our estimates is to help inform policy. If under-reporting is a constant, then it may be that it wouldn't change resource allocation. It's relatively easy for me to imagine that under-reporting might be constant over space and time. However, I would believe a-priori that there would be differential under-reporting by age, and as such that this would alter the resource allocation if it were taken into account.

athowes commented 2 years ago

Added section about this in limitations.