P2-S3.08 Selling sex on the doorstep: development work for the third British National Survey of sexual attitudes and lifestyles (Natsal 3)

Abstract

Background The Third British National Survey of Sexual Attitudes & Lifestyles (Natsal 3) builds upon groundbreaking research undertaken for Natsal 1 and Natsal 2 in 1990 and 2000, respectively. Natsal 3 aims to interview 15 000 men and women aged 16–74 years during 2010–2012. However, the first pilot (Pilot 1) for Natsal 3 achieved a response rate of only 38.4%, leading us to explore ways to maximise survey participation. Methods Following Pilot 1, we consulted widely with interviewers, survey methodologists and communications experts to improve documents sent to households in advance of the interviewer calling, and the survey branding. This included simplifying the advance letter language and removing specific references to “sex”; developing a more detailed information leaflet to send with the letter; developing a study logo and respondent website. In Pilot 2, households were randomised to be sent the leaflet with the advance letter or to be given it when the interviewer called. We also examined the effect of the token of appreciation on participation rates by randomising households to receive either a £15 or a £30 voucher for participation. Results The overall response rate increased to 50.3% (218/434 eligible addresses) in Pilot 2. This did not vary by whether or not households received the information leaflet in advance (50.8% and 49.6%, respectively), but interviewers felt that households sent the leaflet were more informed about the study and they were more confident approaching these households. Interviewers preferred the improved advance letter and some respondents reported that the website was a key factor encouraging them to participate. Pilot 2 response rates differed by token of appreciation (47.1% vs 53.4% for £15 vs £30) but the sample size was too small to conclude that the difference was significant, so randomisation continued during the first wave of fieldwork. Response rates in wave 1 before reissuing were 48.4% vs 50.6% for £15 and £30, respectively (p=0.2). Conclusions Despite survey response rates declining generally, Natsal 3 development work has shown that it is possible to attain an adequate response in a population-based survey of sexual behaviour. Careful wording of participant documents, attractive survey branding and a participant website all contributed to increasing response rates. Response rates were slightly higher for the larger token of appreciation but increasing the value was not considered cost-effective.