The computer-based random sampling that will govern Creighton University’s COVID-19 surveillance program is a common statistical method, says Pierce Greenberg, PhD, assistant professor of sociology in the College of Arts and Sciences and a specialist in survey research methods.
Greenberg says that the University’s approach mirrors random sampling principles that underlie public opinion surveys.
“The overall idea is that the University can apply the same principles that people apply to public opinion surveys, which is my expertise, to obtain an estimate of the prevalence of COVID-19 among the student population,” he says.
Random sampling is important in the fight against COVID-19 due to its ability to identify asymptomatic or presymptomatic individuals who otherwise may not be tested. People without symptoms can still spread the virus, so using random sampling to identify and isolate them can be a helpful tool in the fight against COVID-19.
The University’s surveillance testing plan calls for randomly sampling 500 students on Creighton’s Omaha campus every week. The names of students to be tested will be drawn randomly using a statistical software program. Students will then be notified and must report to the on-campus testing center during operating hours at any time the following week.
There also will be opt-in surveillance testing for on-campus employees. Faculty and staff are encouraged to participate and can do so by completing this form.
The testing will be conducted at the COVID-19 Testing Center inside the Kiewit Fitness Center. The center – which is open Monday, Wednesday and Friday from 9 a.m.-noon and Tuesday and Thursday from 9 a.m.-12:30 p.m. – will also continue to offer walk-in service.
It’s not perfect, Greenberg said, but surveillance testing is the best method available short of testing the entire student population all the time.
“When you draw a sample from a population, it will always be considered an estimate and have error,” he said. “It is certainly not as good as contacting everyone in the population. But when resources – such as availability of testing – are limited, random sampling can be a useful alternative.