Ian Laga

Title:   The Network Scale-up Method and the Degree Ratio

Date: Friday, December 1st, 2023
Time: 1:30PM (PDT)
Location: ASB 10900

Abstract:  Members of different populations are known to have different average social network sizes. This is especially true for marginalized populations like sex workers and drug users, where members tend to have smaller social networks than the average person. Statistics describing the differences in social network sizes between groups have been labelled the popularity factor or the degree ratio. These popularity factors play a critical role in the accuracy of the Network Scale-up Method (NSUM). In this paper, we first show how the degree ratio affects NSUM size estimates. We then provide a method to estimate the degree ratio of all probe and target subpopulations considered in the aggregated relational data (ARD) survey. Our novel method relies only on the original ARD responses and is the first method to estimate the degree ratio without sampling members of the target population. Simultaneously, we show that rescaling the size estimates by the average biases of the known populations improves size estimates for smaller subpopulations. Our degree ratio correction and scaling procedure could further improve the utility of the NSUM in population size studies.