Blind testing seems pretty clear to me
I'd only call it delphic if the result lends itself to misinterpretation.
Space/Science » in reply to A delphic approach to global cooling stats?
I'd only call it delphic if the result lends itself to misinterpretation.