Methodology

Unless otherwise noted, all opinion polls of “registered voters” are conducted among a random sample of voters in the specified geography. All surveys of “likely voters” are conducted among a random sample of voters, stratified by vote history, age, political party, gender, and geography to match the expected turnout in the target election based on past election results.

Both sampling methods are based on lists of registered voters as maintained by the Secretary of State and individual counties’ Supervisors of Elections. Phone numbers are scrubbed and enhanced with commercial sources to increase coverage.

Surveys are administered via live telephone interview on multiple nights with callbacks. Interviews are conducted with the listed voters on both landlines and cell phones by trained operators monitored for quality control. Survey results are then weighted to adjust for non-response based on age, political party, vote history, geography, and gender.