Respondent-driven sampling in online survey: dynamics and quality

Abstract

The authors discuss the problems of selection of respondents for participation in online surveys. The results and the experience of the use of the respondent-driven sampling (RDS) used in online surveys among students of the Higher School of Economics were analyzed in the article. The authors checked whether the web-based RDS corresponds to the basic assumptions such as reciprocity of communications between respondents, random selection and respondent` capability to adequately assess the size of the ego network. An additional question referred to the influence of the “seeds” (first RDS respondents) on the dynamics and the quality of the final sampling was also studied. To answer the questions, two online RDS-based surveys were carried out by the author (online RDS-1 and online RDS-2). The key difference between these two studies is that the “seeds” in the second one were selected out of the researchers` own personal contacts. The authors conclude that the dynamics of the online RDS is more effective if the “seeds” know the researcher in person. At the same time, online RDS-1 and online RDS-2 did not provide the exact assessment of the control parameters of the population. Based on the results of the study the authors define problems for future research.