Are Implicit Association Tests a reliable method of market research?

How can we eliminate this possibility and get more accurate, clear, and objective results?