[H]ow many people will accept the shot?
The answer might depend on how you ask the question.
Four different groups surveyed Americans near the end of last year to find out how many planned to get immunized against Covid-19. The positive responses ranged from 20% to 63%—but the surveys differed in methodology, phrasing and number of possible responses.
Some of the polls contacted participants by telephone; others used online panels. Half mentioned the Food and Drug Administration’s role in approving a vaccine, and half suggested it would be free—both decisions that might have influenced responses. The number of possible answers to each question ranged from two to five, and as the options expanded, the portion of participants who said they would get vaccinated diminished.
“Basically, every decision you make when you conduct a survey is going to shape your results in some way,” said Joy Wilke, lead survey scientist at Civis Analytics, a software and consulting firm that compared the results of the polls.
“The number and type of responses has a strong effect on results,” said Franklin Marsh, the lead data scientist at Civis Analytics. “Giving people a ‘not sure’ option decreases the amount that say yes.”