May 21, 2021 - 7:00am

In the best film ever, Fight Club, the antihero character Tyler Durden has various jobs which he uses as opportunities for social terrorism. He is, for instance, a waiter at a fancy hotel, and pees in the soup. But he also acts as a projectionist at a cinema, and (this being the days before digital projection) cuts single frames of pornography into children’s movies.

“So when the snooty cat and the courageous dog with the celebrity voices meet for the first time in reel three, that’s when you’ll catch a flash of Tyler’s contribution to the film,” says the narrator. “Nobody knows that they saw it, but they did.” In the next shot, children in the cinema burst into tears.

This is, essentially, “subliminal advertising”: the idea that momentary appearances of images or words, too fast for the conscious mind to detect, are picked up by our unconscious minds and influence our behaviour. It arose out of the work of a man called James Vicary, who claimed that by flashing an image of the words “eat popcorn” and “drink Coca-Cola” for 1/60,000th of a second to audiences in cinemas, it increased the sales of those goods by 57.5% and 18.1% respectively. He said he’d shown it to 50,000 moviegoers over six weeks.

Since then, subliminal advertising has developed an almost spooky hold on the public imagination. It is banned in the UK, but companies still try to use it. 

I had assumed that the original research was garbage, because so many flashy and well-publicised claims about social psychology are. More specifically, subliminal advertising seems to fit the model of “social priming”, a subfield of psychology based on the idea that you can subconsciously put ideas in people’s heads and get dramatic changes in behaviour: making them think of words like “bingo” and “wrinkle” made them walk more slowly because they’re “primed” to feel old, for instance. 

Social priming has suffered badly in the last decade: dozens of its best-known findings have been shown to be false. I’d assumed subliminal advertising would be much the same, given that it made ludicrous claims like a 57.5% increase in sales.

What I hadn’t realised, and learnt recently via the statistically savvy psychologist Daniël Lakens, was that the original Vicary study was a total fraud. Vicary never performed the research at all. There was no published paper, and the owner of the cinema denied that Vicary had ever carried out a test. It was just made up out of whole cloth.

And yet…everyone still believes in it! “Subliminal advertising” is a phrase we all know. It’s a staple of popular culture (to pick a random example, the plot of an episode of Columbo relies on it). But it’s based on nothing.

(There was, I should admit, a 2006 study called “Beyond Vicary’s fantasies” which apparently found a weaker effect in some circumstances, making people buy Lipton ice tea but only when they were thirsty. But all studies before about 2011, especially un-preregistered ones like this, are suspect; this in particular was statistically suspect. I would bet at quite low odds that it wouldn’t replicate; in fact, a BBC documentary which did try to replicate it found nothing.)

Luckily, the plot of Fight Club doesn’t turn on it: I would be very sad if psychology’s statistical failings ruined this film. But it’s yet another victory for my heuristic that, if you are not a psychologist, it’s worth assuming that every exciting psychological finding that you’ve actually heard of is probably false.

Tom Chivers is a science writer. His second book, How to Read Numbers, is out now.