Is there anyone out there who can speak to, and/or at least refer me to a good source for, issues of household respondents in a sample I’ve taken up? The situation is that my unit of analysis is the household, but all of us in the social sciences know how problematic that concept is (especially in Africa, where my research takes place). Nevertheless, in an effort to reduce any age or gender bias, I randomized the adult with whom I spoke for responses concerning the household to get a true range of what HH’s in my sample were, rather than solely from the head of household. Obviously interviewing individuals is preferable when trying to capture the true variability of what exists in a HH, but what’s done is done. Has the strategy I’ve described been written up anywhere? Phrased more formally than what I have here? Thanks in advance!