Survey research and HIV-related behaviours: a case for caution

Abstract

THIS paper discusses some of the findings from a recent survey of HIV and Aids-related issues amongst a random household sample of 16-30 year- olds. The focus of the discussion is on the method ological problem of respondent refusal, particularly internal non-response when significant numbers of respondents refused to answer the more sensitive questions asked by the interviewer. Differences be tween respondents who refused to answer such questions, and those who were willing to answer, are presented. Refusers were younger than those who responded, and less likely to be in a stable partner ship. It is argued that this loss of information has important implications, and a general need for cau tion in interpreting the results of such surveys is emphasised.