What a political poll can and cannot tell you

Imagine two polls released on the same morning with different results. That difference is a starting point for questions, not proof that one poll must be wrong. The surveys may have reached different groups of people or asked about different choices.

Begin with the fieldwork dates and the population surveyed. Adults, registered voters, and likely voters are not interchangeable groups. Read the exact question, including the order of the options, before comparing a result with another survey.

A reported margin of error describes only part of the uncertainty. Who responded, how participants were recruited, and how answers were weighted also affect interpretation. Small changes deserve caution, especially when the underlying methods differ.

This illustrative reading exercise shows why our coverage would explain the method alongside the number. A poll captures responses under particular conditions; it does not settle what people will do later. No polling results or election predictions are being reported here.