In Thursday's Star-Exponent article, "Pollsters survey Culpeper citizens about solar project," Allison Brophy Champion reported an inaccurate and misleading statement about a survey conducted in Culpeper County by our firm, McLaughlin & Associates. She wrote, "Initial results of the survey of less than 1 percent of the county’s 31,646 active voters ..."
The inaccuracy is that the number refers to registered voters, and not all registered voters "actively" vote. In the 2016 general election for president, 22,218 adults voted in Culpeper. In 2015, about 8,600 adults voted in county and Virginia General Assembly races.
The survey we conducted was designed to reflect a likely turnout for the November 2019 general elections for the Culpeper Board of Supervisors and Virginia General Assembly. The 2019 turnout is more likely to resemble the 2015 turnout of around 9,000 voters.
By saying "the survey of less than 1 percent of the county's 31,646 active voters," Champion is suggesting the survey isn't scientific, and therefore it isn't credible.
The survey we conducted among 300 Culpeper County voters who are likely to vote in the 2019 general election has a margin of error of +/- 5.7% at a 95% confidence interval. The survey is scientific and meets all academic polling standards.
We used election turnout from 2015 and 2017 to determine appropriate geographic and demographic models that would best reflect the turnout in November 2019, to ensure a representative sample of likely voters.
The standard scientific sample size for a region the size of Culpeper County is 300 interviews. A standard sample size for the state of Virginia would be 600 to 800 interviews, and the standard sample size for a national survey would be 1,000 interviews.
The statement in the article is concerning because it tries to undermine a scientific survey. Whether it was intentional or not, the statement misleads its readers.
McLaughlin and Associates