Weighting Venue-Based Sample Data for the National HIV Behavioral Surveillance System among Men Who Have Sex With Men
*Lillian Shengjung Lin, CDC
Keywords: VBS, location sampling, center sampling, intercept sampling, indirect sampling, NHBS/MSM2, MSM, weight-share method
A survey that uses venue-based sampling (VBS)—also known as location sampling, center sampling, or intercept sampling—samples a population by collecting data from individuals contacted at a sample of venues (i.e., locations) and time periods. This paper discusses the calculation of survey weights for the VBS design used in the National HIV Behavioral Surveillance System among men who have sex with men, conducted by the Centers for Disease Control and Prevention in 2008. We start by calculating visit weights based on the probabilities of selecting venues, day/time-periods, and venue attendees and adjusting these weights to compensate for nonresponse at various points in the data collection process. Then, since persons can make multiple visits during the survey’s data collection period, the visit weights are adjusted to create person weights using the weight-share method. We assess the resulting person weights by comparing weighted and unweighted estimates and by examining the effects of the variability of weights on the precision of key estimates.