24 Jan 2015

Beware the viral ‘experiment’!

Never before has the internet been so alive with scientific discovery! From Philae’s successful landing on Comet 67P (and a hilarious reaction that followed) to guerrilla social experiments, people have been sharing experiments and discoveries like a banker shares debt.

“‘cept you don’t get all this!”

Here I focus on a kind of experiment that has gone viral, time and again: the social experiment. These profess to reveal, via an empirical method, a deep truth about humans that “EVERYONE SHOULD KNOW”. The results are presented in such a surprising way that people do share them across the world. As viral marketers we’ve a lot to learn from the success of this kind of video, but we also need to assess their appropriateness, given their tendancy to be exposed as hoaxes or as subject to deceptive bias and editing.

Consider a story like the Kiwi lads who released a video of the human-sized eel they ‘discovered’ in the Manawatu River. For its revelatory novelty, this video circled the globe, though it wasn’t long before its makers bemusedly declared it was an obvious hoax. They’d filmed it using a red screen and a regular eel in a bathtub. Given a backdrop of fraudulent sightings of mythical beasts like the Loch Ness Monster, this story doesn’t seem too unusual. Most viewers are quick to realise that it was a hoax and move on.

Like sightings of beasts, social experiments capture the world’s imagination. Dove’s ‘Real Beauty’ and Always’ ‘#LikeAGirl’ campaigns are among the most successful viral advertising campaigns of all time. They both professed to reveal, via a scientific method, an aspect of human behaviour that we’d never before seen demonstrated. Unlike with the Kiwis’ bestial hoax, however, few viewers of these videos are ever made aware of the madness in their method.

The main issue with these two videos is the careful selection of participants, all of whom just happen to be wonderfully confident, healthy and eloquent—and they all agree with one another! Sensation! The problem with this approach is that although it makes the video more flowing and engaging, the facts that it implies are far from robust. If the statistics implied by these experiments were published verbally (e.g. “100% OF WOMEN DISLIKE THEIR APPEARANCE.”) then they’d be shown up for the strange distorted figures that they are. Clearly by showing, not telling, Dove and Always have gotten away with a lot.

Fortunately for the sake of holding rogue experiments to account, a strange, guerrilla form of peer review is now developing on social media. Around the time Hollaback! published a woman enduring everyday harassment during a walk through New York, numerous replication attempts were posted from other parts of the world. New Zealand is apparently free of harassment, while Egypt has lechers of the quiet, staring variety. On the one hand it’s great that communities are building on one another’s discoveries, but the problem of creators carefully selecting their participants (or in this case, precise locations) still looms.

But here for the dramatic climax. The most flagrant use of deception to stage a compelling discovery could be seen in a video published on YouTube last week by a pair of viral marketers. In it a young woman, pretending to be drunk, would ask for directions, only for every sober guy she approached to try to lure her home to bed. This was a shocking revelation to most people: that in plain light of day average blokes could be so shamelessly exploitative. As a result the video was rapidly shared across the world. But here’s the catch—every man featured in this video was confronted before filming and was explicitly asked to try to convince a female actress, when in character, to come home with them. They were assured footage would be used only for a light-hearted comedy sketch. This was sample selection at its most perverse. You can see the blokes’ reactions to the whole saga here.

The video of the drunk girl has now been removed from YouTube, presumably not just for bad science, but for the entrapment and defamation of presumably innocent passers by. The greatest shame out of all of this, though, is the implicit suggestion that the result had to be fabricated. There are probably parts of the world where men are this openly exploitative, but without a fair test many will resume thinking there are not.

We’ve many lessons to learn from this spate of experimentation being published, both as marketers and as viewers. As marketers we can see how popular an experiment can be, especially if its participants are engaging and its results are unexpected. At the same time, we should feel responsible not to distort facts through participant selection and editing. Distorting data through participant selection or deception can be damaging, not only to viewers’ general factual awareness, but also to the cause you set out to promote, should you ever be found out.

As casual viewers we also need to be critical of what we watch. In light of the events above, perhaps we should view all factual content, especially if branded, with one big question on our lips: was the creator of that content both willing and able to tell the truth?