Telephone (random-digit dialing) polls are conducted with a sample size of 800-1000 adults (age 18+) and a maximum theoretical sampling error of +/- 3.5 percent. Data are demographically representative of adult U.S. residents (state or national).
Online polls are conducted with a sample size of 1000-2000 adults and sampling error of +/-3.1 percent. The data is weighted in two stages to ensure accurate representation of the U.S. adult population. Figures for age, sex, race, education, and income were weighted where necessary to bring them into line with population demographics. "Propensity score" weighting was also used to adjust for respondents' propensity to be online.