I haven’t looked at the methodologies for recent polls, but any surveys conducted via phone are inherently flawed.
How many people are going to answer a call from ‘unknown’ or ‘rando research center’? How many of these people are willing to take the time to provide answers to political questions to a stranger? Now, who do you know that fit this criteria and what kind of person are they?
Fivethirtyeight are pretty good at recognising flaws in polling and adjusting for them. You can go deep and look at their methodologies, as well as the margin of error they give to their results.
Sometimes they give a massive margin of error when things are uncertain, sometimes they’re more confident in the data they’ve got.
I haven’t looked at the methodologies for recent polls, but any surveys conducted via phone are inherently flawed.
How many people are going to answer a call from ‘unknown’ or ‘rando research center’? How many of these people are willing to take the time to provide answers to political questions to a stranger? Now, who do you know that fit this criteria and what kind of person are they?
Fivethirtyeight are pretty good at recognising flaws in polling and adjusting for them. You can go deep and look at their methodologies, as well as the margin of error they give to their results.
Sometimes they give a massive margin of error when things are uncertain, sometimes they’re more confident in the data they’ve got.