Statistics Colloquium: The History and Evolution of the Network Scale Up Model with Application to Estimating HIV Populations by Payton Ahola ’24 & Spencer Huang ’24, Wednesday February 14th, 1:00 – 1:50pm, North Science Building 015, Wachenheim
A very recent method to estimating the size of an unknown population is to use the Network Scale Up Model (NSUM), in which the size of a subpopulation can be estimated from survey responses to “how many X’s do you know?” This presentation will examine the history and use cases of the network scale up model before concluding with an application to modeling the HIV population in Ukraine. The basics of networks and population estimation will be introduced. This will be followed by an introduction (and derivation) of the first network scale up model and subsequent, Bayesian iterations of the model that generate more precise estimates. Building on this, the presentation will shift to an in-depth analysis of the correlated network scale-up model, upon which the crux of the presentation is based. Finally, the presentation will put the correlated NSUM in practice and examine its efficacy to modeling the HIV population in Ukraine.