How the poll was conducted

This SurveyMonkey Election Tracking Survey for Sept. 1-8, 2016, was conducted online in English and Spanish among a national sample of 4,212 adults who said they are registered to vote in California.

Two forms of the survey were fielded in both English and Spanish. The forms were identical except for the questions on ballot propositions in California, which were split to four per form. Respondents were randomly assigned to receive one form, with about half of respondents assigned to each.

Questions that were asked of the full sample have margin of error estimates of 2 percentage points in either direction. Each of the ballot proposition questions has an error estimate of 3 percentage points.

Data have been weighted for age, race, sex, education, region and voter registration status using the Census Bureau and Bureau of Labor Statistics’ Current Population Survey to reflect the demographic composition of California.

Respondents for this survey were selected from the nearly 3 million people who take user-generated surveys on the SurveyMonkey platform each day. After completing a user-generated survey on the SurveyMonkey platform, a subset of respondents in the United States was asked to complete an additional survey on current events.

Individuals who received the request to participate in the surveys are restricted from taking the survey multiple times. Respondents are restricted from being asked to participate in the SurveyMonkey Election Tracking Survey for 30 days. If a respondent completes another user-generated SurveyMonkey survey, they are ineligible to participate if that survey is completed within the 30-day period.