Article Text

S09-2 Respondent-driven sampling in sampling hard-to-reach precarious workers
  1. Theo Bodin1,
  2. Gun Johansson1,
  3. Tomas Hemmingsson1,
  4. Charlotta Nylén2,
  5. Katarina Kjellberg1,
  6. Bo Buström2,
  7. Per-Olof Östergren3
  1. 1Institute of Environmental Medicine, Unit of Occupational Medicine, Karolinska Institutet, Stockholm, Sweden
  2. 2Department of Public Health Sciences, Division of Social Medicine, Karolinska Institutet, Stockholm, Sweden
  3. 3Division of Social Medicine and Global Health, Lund University, Malmö, Sweden


Introduction Sweden has a long-standing tradition of high quality labour market and work environment research due to excellent access to population surveys and registers. However, there are several challenges when investigating precarious employment and health. Many precarious workers don’t qualify for social security schemes or unemployment insurance and cannot properly be traced in the registries. Also, the response rates to most surveys have decreased rapidly over time, especially among those that are most likely to be in precarious employment, e.g. young people, immigrants and those with low level of education. Therefore, new sampling methods are needed, in addition to current data sources, in order to study this hard-to-reach population.

Methods Respondent-driven sampling (RDS) is a development of the “snowball sampling” chain referral sampling method that relies on peer-to-peer recruitment of study participants. By limiting number of respondents recruited by the same person and by using coded coupons, it is possible to limit social clustering of respondents and to reduce recruitment bias and high “homophily”, i.e. that all subjects have similar characteristics. It also permits linking respondents with recruiters and perform weighted analysis to account for network properties. In the ongoing study we will identify participants in industries where precarious employment conditions are common and assess the feasibility of the method in this population. Hotels, restaurants, retail, staffing industry, taxi services, private security and/or journalists are industries where we expect to find many of our participants. They will respond to a web-based survey and allow for register data to be linked.

Results The pilot phase of the study will be ready in time for the conference and preliminary results and lessons learned will be discussed.

Statistics from

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.