Cross-cultural cases are often not independent, because related or neighbouring societies share traits through descent, diffusion and contact.
Galton raised the objection at a meeting of the Royal Anthropological Institute in 1889, after Tylor presented correlations across hundreds of societies: if the societies inherited or borrowed the trait from one another, they are not independent observations and the sample size is a fiction. The problem has never gone away, and every modern cross-cultural study has to say how it handles it.
In practice
- A survey of forty firms in one industry that all copied the same consultancy has one observation with thirty-nine echoes.
- A finding replicated across twelve European countries has been replicated across one civilisation.