Jennifer Wu presents her paper 'Respondent-driven sampling as a recruitment method' co-authored by Rojan Ezzati, in Parallel session II(B) of the conference Examining Migration Dynamics: Networks and Beyond, 24-26 Sept 2013
Respondent-Driven Sampling (RDS) is a peer-to-peer sampling method similar to snowball sampling, but with a mathematical model that weights the sample to compensate for biases in such recruitment. In this paper we focus on the recruitment side of the method, describing how we used RDS in our data collection among Brazilian and Ukrainian migrants residing in Oslo, Norway. The peer-to-peer recruitment approach of RDS helped us succeed in reaching our target sample size in the Ukrainian case, but not the Brazilian. In this paper we explore possible reasons for this. First, our two cases demonstrate that the target population size is not determinative of the failure or success of RDS, as the Brazilian and Ukrainian populations in Norway are roughly the same size. Nor does it appear that the social network size of the initial individuals selected to get recruitment started played a role. In our data collection, we detected considerable concerns regarding stereotypes within the Brazilian community. Hence we question whether alternative incentives to the monetary ones we offered for participation and recruitment (as part of the RDS procedure), would have provided us with better results. Furthermore, we find that our questionnaire-based interviews were longer with Brazilians than with the Ukrainians. Given that RDS relies so heavily on the recruiters' accounts of their experiences to potential recruits, it is highly vulnerable to any negative experiences. Finally, how different populations respond to RDS design varies from one case to another. In the Brazilian case, we found that the need for respondents to recruit others following the required RDS procedures was perceived as a burden, which impeded further recruitment.