When 1000 Genomes Project enrolled a sample to a population (e.g. CHB or LWK), what criteria did they use to ensure the individual really belongs to the underlying population? Did they require something like, to qualify for a LWK sample, 8 of your great grandparents have to only lived in where whatever LWK is supposed to mean? Any post-enrollment filtering (e.g. PCA) to double check the individual's self-report?

Thanks in advance!