Online Program

Comparing survey and sampling methods for reaching sexual minorities as a hidden population

*Alexis Dewaele, Ghent University 
Maya Caen, Ghent University 

Keywords: Sexual minorities / hidden population / sample and survey methods

Lesbian women, gay men and bisexuals (i.e. LGBs) are widely considered a hidden population. We use two different approaches to target this population. By lack of a sampling frame, we set up a targeted sampling design followed by an online survey as a first strategy. Web surveys offer a highly accessible and anonymous way to pose delicate questions. On the other hand, these techniques are known for inducing, for instance, self-selection bias. In our national study, using the same questionnaire, we draw on a probability sample (2.000 respondents aged 14-80) of the Flemish population (the Dutch speaking community in Belgium). By identifying an LGB population in the national survey, we can explore how two different sampling procedures and survey designs lead to differences in sample characteristics, comparing sexual health and biological, psychological, demographic and socio-cultural characteristics. Hence, we gain insight in self-selection processes, discuss the possibility of poststratification to reduce self-selection bias, and explore future strategies for improving the quality and representativity of non-random samples based on self-selection.