>when one corroborates this with other sources of data, it is a "valid" technique.

Couldn't you say the same about the technique of having 10,000 monkeys hammer randomly on typewriters?