If you quickly agreed with these statements, you might be drawing too much from the psychological experiments the media and self-help gurus regularly use to explain our behaviour.
In an unprecedented year-long project, psychologists from around the world attempted to replicate 100 major studies — and could only reproduce the original result 39 per cent of the time.
Alex Holcombe, associate professor of psychology at the University of Sydney, toldnews.com.au: “Most of the science and health news you see in the media is based on individual studies. Those studies are rightly sometimes called ‘groundbreaking’ because they provide the first evidence. Unfortunately, as is now obvious, a single study is not enough. Possibly about half the studies reported in the media cannot be easily reproduced, which is very disappointing.”
It’s not just psychology that faces this problem, but many other areas where important research is being carried out.
“Concerns about replication have been raised in cancer biology, economics, political science, ecology, and other sciences,” said Dr Holcombe.
“Psychology was simply the first field to get its act together sufficiently to do a systematic, scientific study of how frequently its science replicates. The limited evidence we do have suggests that in cancer biology, reasonable attempts at replication also fail at least 50 per cent of the time.”
Here are some of the popular ideas that couldn’t be confirmed.
MEN CAN’T TELL THE DIFFERENCE BETWEEN FRIENDLINESS AND FLIRTING
Men are known to see women as having greater sexual intent than other women see. A US study claimed this was because men are not as good at distinguishing between friendliness and flirtiness as women. That result couldn’t be replicated. Scientists said this could be because of cultural differences between the US and the UK (where the latest study was attempted), or because the images from the original 2008 test seemed ‘dated’. These caveats aside, there’s no reason to believe men don’t know when you’re flirting.
MOTIVATIONAL WORDS HELP US LEARN
There’s a huge amount of interest in unconscious, unintentional learning, from hypnotherapy to listening to tapes while asleep. Psychologists showed that if people are primed with motivational, goal-directed words, they learn better without even realising it. They claimed participants learned more after completing a word search containing words like ‘ambition’, ‘progress’ and ‘success’, than after a neutral one. But this year’s experimenters found no such effect.
WE TRY TO BRING DOWN MORAL OUTLIERS
The original study said we denigrate ‘moral rebels’, i.e. those who take an unpopular, morally laudable stand, when that choice is different to our own behaviour and can be seen as threatening to our choice. It said that because we think they will look down on us, we consider them as a threat and perform a “pre-emptive strike” to lessen the anticipated sting. The replicated study found no such reaction.
MULTI-TASKING RUINS OUR CONCENTRATION
The first study said task-switching interferes with ‘simultaneous working memory processing’. Our performance was thought to be affected by the difficulty in task-switching and its frequency.
But the new study found that increased switching between tasks doesn’t necessarily make us perform worse, although it may take us longer to process numbers. It suggests that multi-tasking may be fine, depending on what we’re doing.
WE CLING TO OUR PARENTS WHEN WE ARE REMINDED OF DEATH
A study on ‘terror management’ claimed parental attachment helps young adults manage their anxiety over mortality and the inevitability of death. It suggested that insecure individuals were more likely to rely on relationships with their parents for support, whereas secure individuals were more likely to rely on relationships with romantic partners. But the replication of the study did not find that our terror altered the time we allocated to parents, siblings, friends and partners, or affected that specific attachment.
ANGER IS ABOUT POWER-PLAY
Researchers said our response to communicated anger in a bargaining exchange depends on power and knowledge. They claimed that the party with the advantage in terms of information or power will become more determined not to give in and even increase demands in response to anger. The latest experimenters disagreed.
HANDWASHING MAKES US FEEL CLEANSED OF WRONGDOING
Psychologists said that individuals make less severe judgments when they are primed with the concept of cleanliness, or wash their hands after experiencing disgust. But the second-round researchers did not find that physical cleanliness reduced the severity of moral judgments, or that hand-washers could erase feelings of disgust. They also found no evidence for possible opposing predictions that physical self-cleansing would lead to more severe moral judgments.
WE’RE MORE LIKELY TO CHEAT IF WE DON’T BELIEVE IN FREE WILL
Two original studies found reading an essay that undermined a belief in free will led to more cheating in an experimental task. While in the same direction as the original result, the replication result was not statistically significant. Researchers said that the lack of effect could be due to a failure to effectively manipulate free will.
WOMEN ARE ATTRACTED TO DIFFERENT MEN WHEN FERTILE
This study claimed that women who are in a relationship prefer single men when they are ovulating and attached men when they are not ovulating. They suggested that attached women would only leave a relationship for a new partner who is capable of providing resources, whereas an ovulating women should be attracted to a single man who would be more available for sex. But a second attempt could not recreate the link between man’s availability, participant’s conception risk and participant’s partnership status.
LONELY PEOPLE ATTACH FEELINGS TO OBJECTS
The original study said when we are feeling lonely, we attach feelings to inanimate objects and pets. The researchers claimed that socially disconnected people show an increased belief in supernatural agents, a greater propensity to attach social traits to pets and be more likely to detect faces in ambiguous drawings. But the second set of researchers could not replicate any of the three claims.
What we can take away from this
Drawing conclusions from such results can be complicated by all sorts of factors, from tendencies to experiment on people from higher socio-economic backgrounds in the West, to different sample sizes or alterations in the testing environment.
Patrick Goodbourn from the University of Sydney was one of the 270 researchers from 17 countries to be involved in the project. He told news.com.au the number of failures didn’t mean the original research was wrong, but scientists needed to be wary of small things that could affect results.
“The more surprising results were less likely to replicate, but you want a balance between innovative results and reliable or solid ones. There’s also a second stage, which is that you might disagree on what a study means for the human mind.
“It’s not the case that 61 per cent of studies are false. We were keen to make sure the original author was involved so there was no sense we were targeting them.”
While psychologists can find it harder to create large sample sizes than physicists looking at particles, m
Many fields deal with variability in results. There’s a high rate of variability in drug treatment, for example.
What we need to see now is other disciplines doing the same replication work, science journals enforcing more rigorous testing and the public responding with more scepticism. Then maybe we can get a little closer to understanding complicated human beings.