Does Methodology Matter? Adolescent Reporting of Sexual Behaviors in a Randomized Trial in Urban India

Jaya Jaya, Johns Hopkins Bloomberg School of Public Health

Adolescents often have difficulty reporting sexual behaviors on standard face-to-face interviews, particularly in socially conservative settings. Audio Computer Assisted Self Interview (ACASI) has been suggested to improve reporting of sensitive behaviors; however, ACASI has not been well tested among less educated adolescents in developing countries. We compare face-to-face interviews, ACASI and culturally appropriate interactive interviews in a randomized crossover trial with a matched pair design. 583 boys and 475 girls, ages 15-19, in four slums in Delhi, India, received two survey methods in random order. Each had a face-to-face interview and either ACASI or interactive interview. On interactive, 28% of boys and 7% of girls reported sexual intercourse while 20% of boys and 2% of girls reported having sex on face-to-face. We find that ACASI does not consistently lead to higher reporting; however the culturally appropriate interactive methodology led to more reports across a range of sensitive behaviors for both boys and girls. (153 words)