Dunbar found that the scientists tended to treat these surprising outcomes as the result of flaws in their experimental method: some kind of contamination of the original tissue perhaps, or a mechanical malfunction, or an error at the data-processing phase. They assumed the result was noise, not signal.
