Syndicate content

Add new comment

Submitted by Berk Ozler on
I looked at the working paper on your website - interesting work. I am discounting the results on HIV testing, as your explanation was immediately what I thought when I saw the findings (that people may be thinking about other offers of HIV tests than the specific one mentioned in your question). However, on the HIV status: is it possible that the fact that you knew the answers played a role? Meaning that if the respondents knew that you (or your enumerators) knew, that would make them less likely to lie to the enumerator's face. This could have also influenced how they answered the block questions including the HIV one. I wonder verifiable HIV status was the elephant in the room and other information that really only the respondent (and a few other anonymous people) know would see better performance from LR...