Strategies for surveying platform workers: lessons from a Belgian case study

Abstract

Job quality among platform workers has been high on labour market researchers’ agendas. Nevertheless, many report difficulties in collecting data for this group of workers. Platform workers meet many of the criteria of hard-to-survey populations. The aim of this paper is to evaluate whether purposive, non-probability sampling can provide a good strategy for collecting information on the job quality of platform workers through an internet survey (SEAD Platform Survey). The study on which this paper is based, employed different strategies by dividing platform workers into categories based on type of activity. Sampling techniques were adapted to each category and included referral, social media advertisements, (virtual) convenience sampling and a web panel. Despite the cost and labor intensity of these non-probability sampling techniques, a sizeable sample (N = 490) of platform workers was collected. Moreover, the SEAD Platform Survey showed very similar characteristics to previous probability samples within the study population (LFS Module platform work, COLLEEM II, and the ETUI IPWS). Researchers focusing on (other) hard-to-survey (worker) populations can learn from this endeavor, showing that when there is little to no opportunity for probability sampling, purposive, non-probability sampling techniques can offer a good alternative to reach a rich, statistical resource.