They say this stuff came up "never" or "about once or twice" in past year, but there was substantial difference across topics. Most common topics were related to religion, immigration, and general sentiment toward politicians (we fielded the survey in summer 2024).

Some notable difference across groups: Republicans say they talk more about religion. Democrats say they talk more about extramarital affairs (!). Older respondents say they talk more about social security. Younger respondents say they talk more about homosexuality and gay marriage.

Average reported discussion frequency of each issue highly skewed by a few people who talk a lot about most things -- your friends who are always talking about politics. People who say they talk a lot about one thing likely to say they talk more about everything else, too.

For example, people with graduate degrees cluster in a small corner of Cultural Blau Space, while people with only a high school diploma are spread across the whole space.

When we look at homogeneity of categorical labels (how close any two members of a group are in Cultural Blau Space) in space defined by 150 GSS opinion items, we find ascribed categories have low homogeneity, but voluntary/achieved categories are quite homogeneous (esp. high levels of education).

Preprint! Since @vaiseys.bsky.social & I said adults rarely change, there's been pushback saying ppl change all the time. So we (plus @edelmann.bsky.social, @tkeskinturk.bsky.social & @isabellabouklas.bsky.social) tried to quantify how much cultural difference change explains: osf.io/preprints/so...