NSUM

NSUM, or the Network Scale-Up Method, is a statistical approach used to estimate the size of hidden or hard-to-reach populations by asking respondents about how many people they know who belong to the target group and about their own network size. It relies on the idea that the proportion of the population in a hidden group can be inferred from individuals’ social networks.

In practice, survey respondents report two numbers: their personal network size, that is, how many people they know personally, and how many people they know who belong to the target population. If the total population size is N and the hidden population size is M, the expected number of visible members in a respondent’s network is ti × (M/N), where ti is the respondent’s network size. Aggregating across respondents yields an estimate for M, such as M ≈ N × sum(ri)/sum(ti), with ri representing the number of visible hidden-group members reported by each respondent.

The method faces several biases. Visibility bias occurs when hidden-group members are not equally recognizable; transmission bias occurs if respondents are unaware of some individuals in the target group; barrier effects reflect nonrandom mixing between groups; and errors in reporting network size can occur. Researchers address these issues with calibration against external data, adjustments for estimated visibility, and model-based techniques that incorporate uncertainty.

NSUM has been applied to estimate the sizes of populations such as sex workers, people who inject drugs, and men who have sex with men in various settings. Its results aid public health planning, surveillance, and the allocation of resources for outreach and services.