Summary
The episode works through four questions that students commonly find difficult in the lesson on surveillance and sampling. It examines why a rise in reported cases from passive surveillance may reflect changes in testing, using a hypothetical test-positivity example and the restriction of publicly funded testing in British Columbia during the Omicron wave of early 2022. It then distinguishes sensitivity from predictive value positive in syndromic surveillance and shows how sampling weights correct the estimate from a survey that oversamples a rural stratum. A multi-step calculation combines the sample size for a proportion with the design effect, non-response and whole-class sampling, and Sarah corrects a common error in the non-response adjustment. The episode closes with a discussion between the hosts on whether non-probability samples can estimate prevalence, drawing on the 2011 National Household Survey and Pew Research Center studies of opt-in panels.