Start here · Essential guide
How Opinion Polls Work and Why They Differ
A plain overview of how opinion polls are designed, from choosing a sample to weighting answers, and why two careful polls can still report different pictures.
Evidence and Method
How opinion is measured and why polls disagree.
Start here · Essential guide
A plain overview of how opinion polls are designed, from choosing a sample to weighting answers, and why two careful polls can still report different pictures.
What a poll's margin of error does and does not tell you: sampling uncertainty explained without formulas, plus how to read gaps, subgroups and close races.
Why the people who decline or never reach a survey matter, how non-response bias can tilt results in a particular direction, and what pollsters try to do about it.
How self-selected online polls differ from randomly drawn probability samples, what each kind of survey can support, and how to judge a result by its source.
Why survey respondents sometimes shade answers toward what seems acceptable, which topics are most affected, and how researchers try to reduce the effect.
Why asking the same question the same way over time is the most trusted way to see opinion shift, and what can break a trend line such as wording or method changes.
How raw survey responses are adjusted to match a population, what weighting can and cannot fix, and the ways adjustments can go wrong or add uncertainty.
How opinions on deep values often change through generations replacing each other more than through individual conversion, and what that means for reading survey trends.
Related topics