Reducing item non-response, unit non-response and social desirability in a survey on sexual health in ethnic minorities

Studying sexual health typically implies some methodological complexities. In SEXPERT I, we studied sexual health characteristics and bio-medical, psychological, demographic and socio-cultural correlates in a population-based probability sample in Flanders (the Dutch speaking community in Belgium). Several methodological strategies were introduced in order to optimize data collection. A follow-up study focuses on a specific population at risk, i.e. the two largest ethnic minorities in Flanders: people with a Turkish or Moroccan ethnic background, between 14 and 60 years old. This study poses even more methodological challenges. First of all, language barriers have to be dealt with and second, we expect sexuality to be an even greater taboo in these ethnic and religious minorities. Among various methodological strategies, we examine the impact of interviewer characteristics on participation rates, item non-response and social desirability. Hence, the study offers an unique opportunity to explore methodological strategies and their impact on participation rates and response validity when studying delicate matters such as sexual health in hard to reach minority populations