LogFAQs > #970525400

COVxy posted...

This whole analysis seems entirely shit if age is a cross-sectional variable. If it's longitudinal, that would be interesting, but I don't think it's likely. They don't say anywhere, so I assume it's cross-sectional, and all they have done is segregated up the usual plot based on generation. Means nothing, really.

To be clear, the data are meaningless since age is confounded with year of birth. Older people are mostly more conservative because they grew up with more conservative culture, not because they became more conservative as they got older. So, hard to understand what's going on exactly with millenials in this plot. But it's almost certainly not "millenials have gotten less conservative as they have gotten older".