Experiment Suggests Evidence for Subliminal Advertising

(MoneyWatch)  Subliminal advertising really does work, The Telegraph (wrongly) concludes after reporting on a University College London experiment in which words were briefly flashed on a computer screen and subjects, who did not have time to read them, were asked to categorize them as positive or negative.

They judged the negative words (agony, despair, murder, etc) correctly 66 percent of the time, but the positive ones were identified correctly only 50 percent of the time -- as bad as pure chance.

As the BBC pointed out, the experiment does not reflect real life. Subliminal advertising has long been banned in the U.S. and Europe even though there is scant evidence that showing people imperceptibly brief messages has any affect on their behavior.

A much more interesting experiment was done by Prof. Gavan Fitzsimons of Duke University, who showed that when people were subliminally exposed to the Apple logo, they performed more "creatively" in tasks than those exposed to the IBM logo. Similarly, the Disney logo made people respond more honestly to a questionnaire than the E! Channel logo.

And finally: Brands really do "live" in different neural circuits in the brain, if this unscientific test -- in which your author was inserted into an fMRI while being exposed to different brands -- is any guide.