Social desirability bias in sexual behavior reporting: evidence from an interview mode experiment in rural Malawi.

CONTEXT Social desirability bias is problematic in studies that rely on self-reported sexual behavior data. Where gender norms create different expectations about socially acceptable behavior, males and females face distinct pressures in reporting certain outcomes, which can distort assessments of risk for HIV and STIs. METHODS In 2009, relationship and sexual behavior data were collected from 1,750 never-married males and females aged 16-18 via audio computer-assisted self-interviewing (audio-CASI) during the third round of the Malawi Schooling and Adolescent Study. A comparison group of 311 youth completed an identical questionnaire in face-to-face interviews. To assess whether interview mode may have influenced participants' reporting of sensitive behavior, reports of sexual experience in the two groups were compared. Multiple logistic regression analysis was used to identify associations between interview mode and reports of these behaviors, by gender. RESULTS In adjusted regression models, males were less likely to report ever having had a girlfriend in audio-CASI than in face-to-face interviews (odds ratio, 0.4), but they were more likely to report having had sex with a relative or teacher (3.5). For females, reports of ever having had a boyfriend or having had sex did not differ between modes. A small proportion of females reported ever having had sex with a relative or teacher in audio-CASI, while none did so in face-to-face interviews. CONCLUSIONS The method used for collecting relationship and sexual behavior data may influence the reported prevalence of some key behaviors, particularly among males. Further research is needed to improve methods of collecting sensitive data.