Kevin Collins: There's this: split-ticket.org/2024/03/27/w... Look at the weighting methods for most high quality surveys (including, for example, the NYT) and you'll see that they're not nesting weights / weighting on interactions with age. Which can make the toplines right and crosstabs wrong.

Deleted.

@brendannyhan.bsky.social who was posting recently about youth polls not being representative now? Thought it was you but maybe not

um, @johnsides.bsky.social? maybe @kwcollins.bsky.social would remember? not sure

There's this: split-ticket.org/2024/03/27/w... Look at the weighting methods for most high quality surveys (including, for example, the NYT) and you'll see that they're not nesting weights / weighting on interactions with age. Which can make the toplines right and crosstabs wrong.

We Polled Young Voters. Here’s What We Found.split-ticket.org If you’ve followed any of the polling from this presidential election cycle, you’re probably aware of massive purported swings in the voting behavior of the youngest and oldest voters in the electo…

Weirdly there is not a canonical cite I can find for why you should do this (recs welcome!), though the Caughey et al Cambridge element makes an analagous point about weighting for southern respondents devincaughey.github.io/files/caughe...

The way you get a problem in the crosstabs is if the correlation between attitudes and non-response differs across categories (South, here, or age buckets). But I think that it's possible that's what has happened since Oct 7

Of nice, links to the debate over what’s going on and they did two polls to test competing theories? Even better than I expected thanks @kwcollins.bsky.social split-ticket.org/2024/03/27/w...

Post