This page contains information about how we conduct our polls, general issues in election polling and survey methodology, answers to basic questions, and links to Center reports and commentaries that examine methodological issues in more detail. Most of our national surveys are conducted by randomly dialing telephone numbers in the continental United States. About Our Survey Methodology in Detail
Pre-election polling is a rare chance for pollsters to "check their work" by measuring how well their polls matched election outcomes.
Most of our national surveys of the general public are conducted by telephone using a random digit sample of listed and unlisted telephone numbers in the continental United States.
Survey researchers employ a variety of techniques in the collection of survey data. In this section, we have chosen the reports and commentaries that best highlight these methods.
Question design is often thought to represent the art of survey research but there is also plenty of science involved. What are the pitfalls and best practices of survey design?
- Questionnaire development
- Measuring change over time
- Question wording
- Question order
- Order of answer categories
- Pretesting
- Pilot tests and focus groups


