How to Sample Drug Users and Other ‘Hidden’ Populations: Methodological Implications for Qualitative Marketing Researchers Abstract

Social marketing has been employed to change the behaviour of consumers for noncommercial reasons connected to both the individual and the public good, for example in protecting women against domestic violence, and in facilitating government rehabilitation programmes for drug users and others. Marketing researchers are often tasked with studying the effectiveness of government rehabilitation programmes among drug users by using survey questionnaires and interviews. However, there are accessibility and methodological problems in finding respondents among this hidden population. This paper therefore proposes the use of stratification techniques for both probabilistic and non-probabilistic snowball sampling to investigate hidden populations. Stratification is defined as the division of the target population into strata or groups that are linked by some demographic characteristic, such as age or income, in order to enhance the research findings. In reality, since most respondents (or research topics) are heterogeneous and knowledge of a particular topic is randomly distributed in the population, the use of stratification leads positively to sampling diversity and efficiency, i.e. higher precision, leading in turn to enhanced quality of information and more robust findings, which can then be used to design strategic marketing policy applicable to the population under investigation as a whole.