Rise of the zombie facts

Researchers found that striking a power pose boosts confidence and hormomes: fact. Until, of course, other scientists debunked that fact with their own research. Robert Matthews explores the myths and realities of the zombie fact.

A woman prepares for a talk by striking the stance both ‘proven’ and ‘refuted’ to boost confidence. Corbis

The power pose, proven by science to improve confidence before a big speech, was then refuted by further science to have that effect. The original finding will probably live on, because some theories are harder than zombies to kill.

Want to be more confident and relaxed in front of huge audiences? Then just before you go on, strike the pose: get up, put your hands on your hips, feet slightly apart – and wait.

Within two minutes, you’ll not just feel more confident, you’ll act like it. Your hormone levels will change and so will your behaviour.

When scientists at Columbia and Harvard universities published this amazing finding in 2010, it created a sensation.

It went on to become a famous Ted talk (a global set of conferences focusing on Technology, Entertainment and Design) by team member Prof Amy Cuddy, that has so far notched up 24 million views.

The validity of scientific claims is not, however, decided by viewing figures – at least, not yet. It is determined by a time-honoured process that has served science well down the centuries – replication.

Now the results of the first attempt to replicate the power-pose effect are in, and they are not exactly confidence-boosting.

In a study involving 200 people – almost five times the size of the original – an international team of researchers tried to confirm the impact power poses have on hormone levels and behaviour, and came up empty. At best, the poses may just make people feel a bit more confident.

Reporting their findings in Psychological Science, the same journal that published the original results, the researchers have been pretty forthright in their public statements.

“Our study is much more meaningful than the original study, as we have much more data,” said Prof Roberto Weber, of the University of Zurich, one of the co-authors. “The greater number of subjects in our study makes it much less probable that our results are due to coincidence.”

So the power-pose effect looks set to become yet another headline-grabbing discovery that falls apart, attempted replication revealing it to be just another of those fluke findings.

Don’t expect it to die away any time soon, though. It has all the makings of a “zombie fact”, a myth that refuses to die because it has some basis in real science.

Psychology has brought many such zombies to life over the years. One renowned example declares that “science shows people are either left-brain logical types, or right-brain creatives”. Then there is the timeless “scientists have proved we only use 10 per cent of our brains”.

In fact, scientists have shown both claims are complete nonsense, yet it makes no difference. If a zombie fact serves a purpose for enough people, no amount of evidence will kill it off.

Go to enough management training courses and you’ll learn that “research has shown” that when we communicate, 7 per cent of our message comes across in the words we use, 38 per cent in the tone of our voice and a whopping 55 per cent in what we look like.

It’s true that these figures are based on research by psychologist Albert Mehrabian, of the University of California, Los Angeles, almost 50 years ago. But his experiments focused on a very specific issue: the different modes of communicating emotions using single words.

That’s a world away from speeches or presentations, yet such niceties tend to get overlooked when the “7-38-55 rule” is wheeled out during training courses. The rule is intriguing, sounds plausible and is based on real science – and that’s all a zombie fact needs to live forever.

The power-pose effect is another case in point. The idea that poses can affect hormone levels is certainly intriguing and the idea that it changes behaviour seems plausible: we’ve all seen cocky people strike these poses around the office. That the original study was published by serious researchers in a serious scientific journal also adds credibility.

Put all that together and it’s likely all the now-questionable stuff about hormone changes and impact on behaviour will be trotted out in training courses for years to come. The fact that all this has now been debunked will make no difference.

But has it really been proved wrong? That’s certainly the impression the team behind the new research seems keen to create. The University of Zurich claims the findings refute the original study.

Yet while one small study was never going to be able to put the reality of the effect beyond all doubt, is one larger study really enough to prove it’s nonsense?

The idea that science can “prove” anything is arguably the biggest zombie fact of them all. Not even mathematics can serve up unimpeachable facts. Every statement is based on assumptions and axioms that just have to be taken on trust. The same is true of science, especially the experimental variety.

From wonder drugs that prove useless to cosmic insights that turn out to be anything but, there’s no shortage of examples of how hard it is to demonstrate anything using data alone.

In the case of the power-pose effect, it’s certainly true that a conclusion based on 200 people should be more reliable than one based on about 40. The theory of statistics shows that doubling sample size cuts the level of uncertainty by about 30 per cent.

So if effects seen in a small study aren’t seen in a larger one, it’s tempting to think that the original results were just a fluke.

And yet while the larger sample is less prone to the vagaries of chance, it’s not completely free of them. It’s still possible that it’s the larger study that’s wrong.

In any case, chance is not the only possible source of misleading findings.

Failures to replicate have many other causes, such as poor study design or faulty execution. So we could be about to witness two potential zombie facts go head to head.

In the blue corner: the idea that a small study proved power poses boost hormone levels and change behaviour. In the red corner: the claim it’s all nonsense, because a larger study failed to replicate the first.

In a rational world the contest would never take place, because everyone would just wait patiently for more evidence. But sadly we live in a world where, as Mark Twain famously put it, “a lie will go round the world while truth is pulling its boots on”.

Except, of course, he didn’t. It was someone called C H Spurgeon.