J. D. @jdetc.bsky.social

A 1% response rate. What if the method of communication is a confounding variable? In other words, can you show the 1% of people who answer their phone to an unsolicited political survey in 2024 are representative of the population? Particularly problematic for weighted lower n subsamples.

1 replies 0 reposts 0 likes

Travis Dawry @tdawry.bsky.social [ View ] Jul 04, 2024 at 23:21 UTC

How could it not be a confounding variable? They interviewed an enormous number of people. Maybe it's not great, but it's not worse than most polling.

0 replies 0 reposts 0 likes