Abstract

Reporting of sexual behaviour is subject to bias, perhaps most importantly social desirability bias. This is a particular problem in communities where discussion of sex is considered taboo and when disclosure of sexual activity can have serious consequences for the individuals concerned (eg, for young people or those in same-sex relationships). Mis/under-reporting of sexual behaviours can result in the design of interventions being poorly informed as well as in intervention effectiveness being unreliably measured. There is increasing evidence to suggest that questionnaire delivery method (in addition to a host of other factors) can impact the validity of reported data and that validity can be improved by careful consideration of questionnaire delivery mode. Computer-assisted questionnaire delivery has been shown to increase reporting of socially sanctioned behaviours in many settings and even in rural, resource poor settings, where people traditionally have limited experience of using computers, have been shown to be highly acceptable and feasible to research participants.