Figure 5
From: Estimation and correction of bias in network simulations based on respondent-driven sampling data
Estimation and correction of bias in homophily (in gender) from RDS. Sub figures a, b, and c are the comparison of homophily between original networks (x-axis) and sample networks (y-axis) in random samples (a), RDS samples without preferential recruitmen t (b), and RDS samples with preferential recruitment (c). Sub figure d is the comparison of homophily between original network (x-axis) and model correction from RDS (y-axis). The colors represent different sample sizes. The red dotted line is the diagnoal line to assist comparison.
