Pink, pink, pink, pink. Pink moan.

August 25th, 2007 by Ben Goldacre in bad science, evolutionary psychology | 103 Comments »

Ben Goldacre
The Guardian
Saturday August 25 2007

I want you to know that I love evolutionary psychologists, because the ideas, like “girls prefer pink because they need to be better at hunting berries” are so much fun. Sure there are problems, like, we don’t know a lot about life in the pleistocene period through which humans evolved; their claims sound a bit like “just so” stories, relying on their own internal, circular logic; the existing evidence for genetic influence on behaviour, emotion, and cognition, is coarse; they only pick the behaviours which they think they can explain while leaving the rest; and they get themselves in massive trouble as soon as they go beyond examining broad categories of human behaviors across societies and cultures, becoming crassly ethnocentric. But that doesn’t stop me enjoying their ideas.

This week every single newspaper in the world lapped up the story that scientists have cracked the pink problem. “At last, science discovers why blue is for boys but girls really do prefer pink” said the Times. And so on.

The study took 208 people in their twenties and asked them to choose their favourite colours between two options, repeatedly, and then graphed their overall preferences. It found overlapping curves, with a significant tendency for men to prefer blue, and female subjects showing a preference for redder, pinker tones. This, the authors speculated (to international excitement and approval) may be because men go out hunting, but women need to be good at interpreting flushed emotional faces, and identifying berries whilst out gathering.

uk.JPG

Now there are some serious problems here. Firstly, the test wasn’t measuring discriminative ability, just preference. I am yet to be given evidence that my girlfriend has the upper hand in discriminating shades of red as we gambol foraging for the fruits of the forest (which we do).

But is colour preference cultural or genetic? Well. The “girls preferring pink” thing is not set in stone, and in fact there are good reasons to suspect it is culturally determined. I have always been led to believe by my father – the toughest man in the world – that pink is the correct colour for mens’ shirts. In fact until very recently blue was actively considered soft and girly, while boys wore pink, a tempered form of fierce, dramatic red.

There is no reason why you should take my word for this. Back in the days when ladies had a home journal (in 1918) the Ladies Home Journal wrote: “There has been a great diversity of opinion on the subject, but the generally accepted rule is pink for the boy and blue for the girl. The reason is that pink being a more decided and stronger color is more suitable for the boy, while blue, which is more delicate and dainty, is prettier for the girl.”

The Sunday Sentinel in 1914 told American mothers: “If you like the color note on the little one’s garments, use pink for the boy and blue for the girl, if you are a follower of convention.” Some sources suggest it wasn’t until the 1940s that the modern gender associations of girly pink became universally accepted. Pink is, therefore, perhaps not biologically girly. Boys who were raised in pink frilly dresses went down mines and fought in World War 2. Clothing conventions do change over time.

But within this study, was the preference stable across cultures? Well no, not even in this experiment, where they had some Chinese test subjects too. For these participants, not only were the differences in the overlapping curves not so extreme; but the favourite colours were a kind of red for boys and a bit pinker for girls (not blue); and they had more of a red preference overall. Red, you see, is a lucky colour in contemporary Chinese culture.

china.JPG

And also snuggled away in the paper was the information that femininity scores on the Bem Sex Role Inventory correlated significantly with colour preference. Now the BSRI is a joy from the 1970s, a self-rated test explicitly designed to measure how much you adhere to socially desirable, stereotypically masculine and feminine personality characteristics (do it on yourself here).

You mark on the score sheet from one to seven how much you feel you suit words like theatrical, assertive, sympathetic, adaptable, or tactful; and then your score is totted up at the end. So women who describe themselves as “yielding” “cheerful” “gullible”, “feminine” and “do not use harsh language” also prefer pink. Thanks for the warning, I’ll try and use that to avoid them in future.

It’s worth being critical and thoughtful about these stories, not because it’s fun to be mean: but because that’s what the authors would want, and also because stories about genes and culture are an important part of the stories we tell ourselves about who and what we are, our sense of personal responsibility, and the inevitability in our gender roles.

· Please send your bad science to bad.science@guardian.co.uk
References:

The academic paper is here:

Biological components of sex differences in color preference
Current Biology, Volume 17, Issue 16, 21 August 2007, Pages R623-R625
Anya C. Hurlbert and Yazhu Ling

dx.doi.org/10.1016/j.cub.2007.06.022

Unless you have an Athens login, you are not allowed to read what the researchers actually said, instead of what the media said they said. Because although they are publicly funded academics at the University of Newcastle, and although this work has been publicised in every major mainstream media outlet in Britain and the US, and although the journal is edited by academics you fund, and paid for by subscriptions from university libraries… the actual academic article is behind a paywall, with a payment model geared towards institutions, rather than interested individuals.

Bad luck you. I guess you have to rely on journalists.

Jo Paoletti has written at length on gender, colour and clothing, an accessible article is here:

www.gentlebirth.org/archives/pinkblue.html

A pair of free-to-access vaguely bookish reviews of colour and gender here and here.

And here is an instructive vintage clipping courtesy of Prof Paoletti:

pinkorblue.JPG


++++++++++++++++++++++++++++++++++++++++++
If you like what I do, and you want me to do more, you can: buy my books Bad Science and Bad Pharma, give them to your friends, put them on your reading list, employ me to do a talk, or tweet this article to your friends. Thanks! ++++++++++++++++++++++++++++++++++++++++++

103 Responses



  1. ForeverAutumn said,

    August 28, 2007 at 2:33 pm

    The basic premise of evolutionary psychology – that the human mind and many of its resultant behaviours are products of evolution – is pretty non-controversial (assuming of course you accept the basic premises of evolution in the first place).

    However, the mere existence of a specific behaviour or preference says nothing about whether it results from genetic hardwiring, cultural programming, or some mixture of both. Certainly, if a preference is widespread across many unrelated cultures –for example, a liking for sweet food – that is a good indication that it has some evolutionary underpinning. But the existence of a preference in two small groups encompassing two cultures that have heavily influenced each other – for example, a liking for raw fish among both Japanese and Californians – shows nothing of the sort.

    This nature/culture issue seems to be particularly fraught and politicized when it comes to differences between the sexes. Certainly it is silly to ignore evidence of such differences because of “political correctness”. If a trait is shown to be widespread across many different cultures – for example, a tendency for males to be more physically aggressive – then there is likely to be some evolutionary reason. But the existence of a trait among related cultures – for example, a tendency for women in both Afghanistan and Saudi Arabia to appear in public less often than their menfolk – then culture is a far more likely explanation.

    There seems to be a new kind of “political correctness” arising. There is a tendency in certain parts of the media to leap on studies that conform to their “men are from Mars, women are from Venus” thesis. Any suggestion that humans have in fact rather a moderate sexual differentiation, and that culture may actually have some impact, is shouted down with charges of “political correctness” and “feminist opinion”. Compare for example the publicity that followed this “girls love pink cos of the berries” story with another recent study that found the average number of words per day spoken by women was not significantly different to the number spoken by men.

  2. ForeverAutumn said,

    August 28, 2007 at 2:47 pm

    Report about the male/female chattiness study:
    www.sciencedaily.com/releases/2007/07/070705152953.htm

  3. NYC EMS said,

    August 28, 2007 at 4:13 pm

    Wrap your baby up in the lugerious color…….jet black.

    I wounder what the psychological impact that would have on our little bundle of joy.

  4. Eddie said,

    August 28, 2007 at 4:43 pm

    Excellent use of Nick Drake in the title!

  5. MT said,

    August 28, 2007 at 5:11 pm

    Sorry, it’s long – I’m a twat.

    O – Enlighten me; what are these “…very well-known, long-standing associations between red hues and sex” other than red-light districts. I thought the red light was to help maintain anonimity – I accept I could be wrong. Just a note though, female humans are not baboons. They have pretty concieled ‘bits’ so how they would act as a red rag to a bull I’m not sure. (Apologies for the red rag allusion in this particular context;)

    I think your claim of a *female* gender bias is a bit of a stab in the dark to be honest, unless you want to back it up with evidence.

    FA – “Universal variability
    One implication of the evolutionary psychologists’ view that the human mind took its current form in the Pleistocene is that all modern humans share a universal human nature. But how do evolutionary psychologists reconcile this claim with the manifest diversity of behaviour and culture found around the world? The answer is that psychological mechanisms are ‘condition dependent’ — that is, the behaviour they produce will be different under different conditions. Suppose, for example, that human psychology operates in part according to the rule “If resources are scarce, then adopt a more aggressive approach to acquiring them”. On this basis one might expect overt levels of aggression to vary according to the current economic or ecological circumstances of the people under consideration. Evolutionary psychologists look at, amongst other things, permutations in behaviour in order to work out what the underlying rules are and how they operate. This research, which necessarily involves cross-cultural studies, commits evolutionary psychologists to a strongly “environmentalist” position: the idea that differences in behaviour are largely the product of differences in environmental — physical, social or cultural — factors.” from Oliver Curry’s review – link above

    The practice of coveting women (and their sexuality) is explicable via an evolutionary lens and is a pan-cultural phenomenon. It is only very recently that women have gained unlimited freedon of movement in the West and dress codes were relaxed and women allowed protection under the law as individuals, don’t forget.

    I really don’t know what you mean by “a new kind of political correctness” – You may be confusing popular discourse with scientific discourse, where no such thing exists. The ‘men are from mars’ stuff is really insidious rubbish that has nothing to do with the scientific study of gender difference, especially in Ev theory. Men and women are not evolved from different planets, I can assure you and the battle of the sexes arises out of the pretty great cooperative endeavour that is human sexual reproduction (I’m paraphrasing Cronin there).

    And there is no biological/environmental (inc.cultural) dichotomy – genes are influenced by both! Culture does have an impact – no one, I assure you who knows what they are talking about, would say that it didn’t! Gods teeth man!

  6. CaptainKirkham said,

    August 28, 2007 at 5:55 pm

    How about a slightly more basic question – is there any actual reason to suspect that there is a fundamental biological colour preference rule based on gender? Why would there be? What purpose would it serve?

    For the record, pink is one of my least favourite colours. Yet somehow I still have two X chromosomes.

  7. MT said,

    August 28, 2007 at 7:36 pm

    I think the paper make it clear what purpose it would serve – or more correctly, what purpose it perhaps *did* serve Whether it is accurate remains to be seen.

    Goldacre should probably explain about statistical significance; I am only a lay person myself here.

    All I would broach is that if pink was found to be preferred by females in statistically relevant numbers, that isn’t the same as asserting that all females *should* like pink, something I don’t think this paper does anyway. Like I say, it is descriptive not prescriptive.

  8. MT said,

    August 28, 2007 at 7:44 pm

    ..and for the record, I too have two X chromosomes..and never liked pink but don’t mind it now – in small doses:)

  9. A.S. said,

    August 28, 2007 at 7:45 pm

    @ 46 Littleshim

    “one problem is that they haven’t actually compared different cultures very well – a very limited sample, and what you might call cross-cultural interference because the Chinese group are apparently immigrants to the UK.”

    True, but you could argue that since they did find a statistically significant difference, the samples were large enough and the cross-cultural interference cannot have been very strong. I agree that it would be desirable to use truly “monocultural” subjects.

    “But why should the Chinese preference be the ‘culture specific aspect’? How very ethnocentric. In fact, from the descriptions above (having not been able to read the article) I’d be tempted to suggest quite the opposite: a ‘general preference for red/pink’ (prevalent across gender and culture boundaries), but with a culture-specific shift toward blues in Western men.”

    Fine by me. I followed the argument from the original paper that a preference for blue is the default and some posts that mentioned a special cultural significance of the color red in China. However, the direction you propose might well have a more plausible explanation.

  10. otheus said,

    August 28, 2007 at 7:57 pm

    MT – Hrm, I thought it would have been obvious. Okay, perhaps the lay scientists need to get out more…. Q: What type and color of flower are girls given by their beaus on special occasions? Red roses. Q: Of what shape and color are the icons that fill love notes? Red-filled “hearts”. Q: The most common shade of lipstick? A: Green. No, just kidding. Can anyone else think of a color more associated with sex and sexual desire?

    As for gender bias, I also thought it would have been obvious that females would not relate in the same way a man would in seeing or dreaming about a woman’s pussy labia.

  11. clobbered said,

    August 28, 2007 at 10:34 pm

    @39 – I am also wondering whether it is valid to screen out colour blindness, especially given the huge gender imbalance between rates of colour blindness (we are talking about factors of 10-50 more colour blindness in men, no?).

    And I would really be a lot more impressed if their cultural control was some isolated tribe rather than Chinese people already exposed to Western cultural norms.

  12. MT said,

    August 28, 2007 at 10:35 pm

    Hmm, well you have a unique perspective on that – I’ll leave you to your daydreams.

  13. Despard said,

    August 29, 2007 at 9:11 am

    #68, 39: it’s perfectly valid to screen out colour-blindness in this case because it’s a confound. If you’re investigating colour vision, you do it in that part of the population that has normal colour vision! If someone can’t see the full colour spectrum, how are you supposed to get meaningful data from them?

    Many of my experiments involve reaching movements, and usually I screen out left-handed people. Now about 10% of the population is left handed, but if I’m looking for something that right-handed people do then of course it makes sense to screen out the left-handers.

  14. ForeverAutumn said,

    August 29, 2007 at 10:03 am

    @62 MT:”I really don’t know what you mean by “a new kind of political correctness””

    You brought up the feminist opinion of ev psych, although no-one else had mentioned it and Ben’s criticism was from a scientific rather than an ideological point of view.

    Yes, the problem is probably more in the media’s interpretation of ev psych rather than the field itself. Certain sections of the media *love* these “men are from mars” explanations. Any criticism, however reasonable, is seen as ideologically motivated and therefore suspect.

    E.g.
    Study finding: Women have a greater average number of shoes than men.
    Ev Psych explanation: Women have a tendancy to gather things, be they berries or shoes.
    Media response: Study Shows Why Women Love Shoes!!!!
    Critic: But this study was only done on a few hundred people living in New York. Also, individual differences were far stronger than group ones. The explanation does not necessarily follow from the study, at least not until you have replicated the findings in various cultures including ones isolated from Western influence.
    Media response: What are you, some kind of Birkenstock-wearing lesbian?

  15. MT said,

    August 29, 2007 at 2:19 pm

    FA – yes sorry, wires crossed. But I just wanted to make clear that though some branches of feminism seem to distrust science (I’m sure you have read Zoe Williams response to the paper), Darwinian feminism puts it at it’s centre. I had a too reactionary response to Ben’s article I do admit – I just saw all the work that had been done attempting to cajole academic feminists to at least view ev theory with an open mind, to begin to accept science as an ally not an enemy being put back to square one. Life began to look too short at that moment and I nearly gave up and got a job in Primark.

  16. ShatterFace said,

    August 29, 2007 at 3:03 pm

    ”But I just wanted to make clear that though some branches of feminism seem to distrust science (I’m sure you have read Zoe Williams response to the paper)”

    I suspect that some writers simply look at a scientific report and ask themselves ‘in what way can I find this offensive to myself or, better still, on behalf of others?’ rather than ‘does this report demonstrate what it claims to demonstrate?’

    Instead of a column assessing the claims of the report we ended up arguing about Zoe Williams’s interpretation of it.

    A week earlier Williams had attacked what she saw as sexism in another report.

    ”Research has suggested women prefer girly-looking fellas to the alpha-male variety. Surprise, surprise – another study that works in men’s favour.”

    commentisfree.guardian.co.uk/zoe_williams/2007/08/so_macho.html

    Frankly I found it hard to see how this report ‘works in men’s favour’ since if women are attracted to one type of men, another class of men are obviously missing out.

    I read an article once attacking binary code as ‘sexist’ as it privileges the phallic ‘1’ over the vulvic ‘0’.

    You can’t reason with some people no matter how feminist you attempt to be.

    By the way, my messages keep disappearing.

    Am I doing this wrong?

  17. ShatterFace said,

    August 29, 2007 at 3:06 pm

    Ah, I see – they reappear when I log back on.

    Curious!

  18. MT said,

    August 29, 2007 at 4:33 pm

    It is utterly insane SF, and is maybe one of the reasons so many people are ashamed to call themselves feminists these days. Unfortunately however, the media is principle vehicle by which many people are introduced to these issues and debates, and most don’t seek to verify the facts for themselves but take the columnists opinions as being facts. That’s why I was so p’ed off at Goldacre’s, what seemed to me, weasel words re ev psych at the beginning of this piece.

    Columnists are very powerful people when it comes to the dissemination of ideas – I despair at the power people like Zoe Williams has.

    Richard Dawkins’ (I think it was in The Extended Phenotype) related a story of the harm misrepresentations, such as those that pour out of the Roses and are passed on via Williams and Bunting et al, can cause. A woman stood up at a conference he was speaking at; she was visibly upset at the consensus about gender difference because she believed that such differences actually sentenced her to a life of inferiority and servitude. Now this doesn’t come from the science, as I’m sure we all know here; such fear mongering comes from anti-ev theory columnists like Williams. They seem to have absolutely no idea of the harm their bloody mindedness and ignorance is capable of.

  19. PRL1973 said,

    August 29, 2007 at 9:42 pm

    Hello from a newbiegeek. I have come home!

    I do stats for money (cause I couldnt get a job as a journalist) and nearly wept on the train reading the original story in the paper – the two professions colliding like planets in a 1950s a-movie to create such a compact little vortex of mind evaporating cack. Did a single paper include a simple “but this of course is boll0cks” before just publish the email some bored students sent them?

    Why wasnt everyone else on the train screaming at the soul crushing inanity (is that a word? I coulda bin a jerno…) of it all?

    100% of the people in this room have just decided the moon is made of stilton.

    It’s all enough to make me want to tickle Richard Dawkins.

    Pete
    x

  20. Robert Carnegie said,

    August 30, 2007 at 1:54 am

    How prevalent does something have to be to avoid being excluded from normal? BBC radio show [Am I Normal?] has interestingly and often inconclusively considered several such issues. Sometimes a lack of conclusions is the best answer you’ve got.

    Incidentally, not all women are multi-orgasmic. I do what I can. 🙂 I think surveys still say from time to time that a proportion of women don’t have orgasms or aren’t sure, which of course does not mean no pleasure in the act at all but arguably missing out nevertheless, somewhat. It’s not as easy as it is in some of the popular media, and it still can be difficult to talk about.

  21. MT said,

    August 30, 2007 at 9:52 am

    Well..from an evolutionary perspective there is a hypothesis that the female orgasm has something to do with ‘female choice’ – this about orgasm during sex not masturbation as I’d think that many women know how to bring themselves to orgasm on their own.

  22. HowardW said,

    August 30, 2007 at 1:14 pm

    Just wanted to say: one of the best Bad Science article titles ever!

    (though does it need just one more “pink” in there? need to dig out the CD…)

    Howard

  23. RS said,

    August 30, 2007 at 3:29 pm

    “I think surveys still say from time to time that a proportion of women don’t have orgasms or aren’t sure”

    There was one of those terrible sensationalist TV documentaries a few years ago about women having spinal electrodes implanted to give them orgasms because they were anorgasmic.

    When they discussed with some of them whether they had tried masturbation or oral sex they responded ‘eugh – oh no, that’s disgusting’.

  24. RS said,

    August 30, 2007 at 3:35 pm

    “How about a slightly more basic question – is there any actual reason to suspect that there is a fundamental biological colour preference rule based on gender? Why would there be? What purpose would it serve?”

    Well given their explanation that it is something to do with berry collecting or flushed faces the obvious experiment that grown-ups would do is to see whether women have increased sensitivity or discriminative ability in the red spectrum. Because the EP explanation actually suggests that as a hypothesis, whereas the colour preference question has no conceptual connection with the explanation, which is pretty much ad hoc anyway.

  25. Littleshim said,

    August 30, 2007 at 4:38 pm

    @66 A.S.

    “True, but you could argue that since they did find a statistically significant difference, the samples were large enough and the cross-cultural interference cannot have been very strong.”

    Quite possible – having not been able to read the paper I can’t really say much.

    Littleshim: “But why should the Chinese preference be the ‘culture specific aspect’? How very ethnocentric…”

    “Fine by me. I followed the argument from the original paper that a preference for blue is the default and some posts that mentioned a special cultural significance of the color red in China. However, the direction you propose might well have a more plausible explanation.”

    Apologies for lack of clarity – I was assuming you followed the paper’s arguments, and my comments were directed at their reasoning rather than at you.

  26. MT said,

    August 30, 2007 at 6:16 pm

    …erm, I just realised looking at Ben’s first sentence to the piece…but wasn’t this a biology paper and not a psychology paper?

    I have it now actually. Am just off the read it myself. Not bragging or nowt.

  27. Dr T said,

    August 30, 2007 at 8:23 pm

    I’ve been watching this with interest, and there are a couple of things that occur to me. I have had a good look at the paper, (in “current Biology” which isn’t a mickey mouse journal) and it all seems fair enough. However, I happened to catch a Radio4 programme, ” the material world” this afternoon where the ethics of deceiving participants in psychology expts was discussed. The point was that if the participants *know* what they are being asked to do, then that adds bias to the results. Some relevance here shurely? The detailed methods say the participants were encouraged to choose quickly, but that might be hard to do. Would you not get a better measure of preference by designing an experiment that tested colour preference covertly? Other than that the paper is fine except for the leap of interpretation that selection for good colour discrimination in the red spectrum “WOULD emerge as a stable colour preference” I think I would change that to “might”. And thus have a working hypothesis that you could test in more detailed experiments – for example that preference is actually linked to colour discrimination…?

    I would also be curious to see a copy of the press release sent out by the journal. The journos don’t have time to scan the real thing much, so my guess is that the press releases are where this comes from. How common is it I wonder, that press release from the journal makes the flamboyant suggestions and is the main cause of the media nonsense, rather than the news journos themselves? So Ben… if you have it, do please post it up – I’d love to see it! And if there is anyone out there with a window on how papers are chosen for the PR treatment that would be very interesting.

    Sorry this is too long 🙁

  28. MT said,

    August 30, 2007 at 8:50 pm

    Eye tracking maybe?

    The paper is only three pages long BTW – with references.

    And scratch my post above please – my supervisor has put me right.

  29. Dr T said,

    August 30, 2007 at 9:29 pm

    Aha thanks Ben – what an interesting site!
    If anyone else wants a look (They appear in pop-ups without book marks), search Eurekalert www.eurekalert.org/index.php for “girls prefer pink”and it comes up top. It’s all pretty reasonable until the speculation at the end – but I guess that is what people latch on to. I suspect that most of the paper articles are cribbed from this.
    MT: I could believe eye-tracking just about, for New Scientist and Nature, though I know the latter press releases. Current Biology would be to obscure to look at regularly, I reckon.

  30. buffalo66 said,

    August 30, 2007 at 10:00 pm

    @ 78 Robert Carnegie : “How prevalent does something have to be to avoid being excluded from normal?”

    I suppose it depends on why you might want to exclude it. In the case of Hurlbert & Ling’s paper, the issue is not how prevalent colour blindness is, but whether it would be expected to affect the subjects’ colour preferences. People with color blindness have vastly different colour perception from normal trichromats, and this is likely to have a huge effect on their colour preferences. As Despard said (#70), colour blindness is confounded with gender so, if colour blind subjects were not removed from the sample, any differences between the colour preferences of men and women could be due to the larger number of colour blind subjects in the male group. Hurlbert & Ling were absolutely correct to remove colour blind subjects.

  31. buffalo66 said,

    August 30, 2007 at 10:01 pm

    I think the reaction to this paper has been a little harsh. The methodology and statistics appear sound. I agree that the evolutionary stuff is probably nonsense, but the authors make clear that this is speculation. My main criticism is that the title of the paper is quite misleading: “Biological components of sex differences in color preference”. The paper doesn’t provide any evidence whatsoever that the difference that they found is due to biological differences, as opposed to cultural influences. In fact, their data seem to suggest the opposite, given that the Chinese subjects, despite being sufficiently westernized to have moved to the UK, showed a substantially different pattern of results.

    The title is probably referring to the fact that the “principal components” that explain the most variance in the data correspond well to the “red-green” and “blue-yellow” mechanisms that have been identified physiologically. But I’m not sure how significant this is. I think you’d always expect these components to emerge from this sort of analysis, because they provide an efficient representation of the cone responses. That’s probably why these mechanisms evolved in the first place.

  32. Dr T said,

    August 30, 2007 at 10:34 pm

    Good point, buffalo66. Principal components analysis is just a way of finding from a ‘noisy’ data set, (and in this case there is loads of noise from culture and other environmental effects) the strongest signal. I am prepared to accept from this paper that there is something in the idea, but a significant relationship between a principal component and another mechanism is NOT a causal relationship – that remains to be demonstrated.

  33. buffalo66 said,

    August 31, 2007 at 12:14 am

    I think the problem is that there’s enormous pressure on scientists to publish in high-impact journals, and these journals love stuff on evolutionary psychology, because it gets picked up by the media, and generates huge amounts of publicity. Rather than being “bad scientists”, Hurlbert & Ling have actually played the game spectacularly well. They’ve conducted a perfectly good study on colour preferences but, instead of sending it to a bog-standard journal, they’ve tacked on a completely spurious and speculative evolutionary “explanation”, and – bingo – it gets into Current Biology.

  34. Ben Goldacre said,

    August 31, 2007 at 12:33 am

    fyi

    Public release date: 20-Aug-2007
    [ Print Article | E-mail Article | Close Window ]

    Contact: Nancy Wampler
    nwampler@cell.com
    617-386-2121
    Cell Press
    Girls prefer pink, or at least a redder shade of blue

    A study in the August 21st issue of Current Biology, a publication of Cell Press, reports some of the first conclusive evidence in support of the long-held notion that men and women differ when it comes to their favorite colors. Indeed, the researchers found that women really do prefer pink—or at least a redder shade of blue—than men do.

    “Although we expected to find sex differences, we were surprised at how robust they were, given the simplicity of our test,” said Anya Hurlbert of Newcastle University, UK. In the test, young adult men and women were asked to select, as rapidly as possible, their preferred color from each of a series of paired, colored rectangles.

    The universal favorite color for all people appears to be blue, they found. “On top of that, females have a preference for the red end of the red-green axis, and this shifts their color preference slightly away from blue towards red, which tends to make pinks and lilacs the most preferred colors in comparison with others,” she said.

    Overall, the differences between men and women were clear enough that the seasoned researchers can now usually predict the sex of a participant based on their favorite-color profile.

    To begin to address whether sex differences in color preference depend more on biology or culture, the researchers tested a small group of Chinese people amongst the other 171 British Caucasian study participants. The results among the Chinese were similar, Hurlbert said, strengthening the idea that the sex differences might be biological. The explanation might go back to humans’ hunter-gatherer days, when women—the primary gatherers–would have benefited from an ability to key in on ripe, red fruits.

    “Evolution may have driven females to prefer reddish colors–reddish fruits, healthy, reddish faces,” Hurlbert said. “Culture may exploit and compound this natural female preference.”

    She said another way to separate “nature versus nurture” when it comes to favorite colors will be to test the preferences of infants. The researchers have plans to modify the color-choice test for use in young babies and hope to have some answers on that front soon.

    About the universal preference for blue, “I can only speculate,” said Hurlbert. “I would favor evolutionary arguments again here. Going back to our ‘savannah’ days, we would have a natural preference for a clear blue sky, because it signaled good weather. Clear blue also signals a good water source.”

    ###

    The researchers include Anya C. Hurlbert and Yazhu Ling of Newcastle Univesity in Newcastle upon Tyne,UK.

    Hurlbert and Ling: “Biological components of sex differences in colour preference.” Publishing in Current Biology, 21 August 2007, R623-625. www.current-biology.com

  35. lauredhel said,

    August 31, 2007 at 7:46 am

    ” I have had a good look at the paper, (in “current Biology” which isn’t a mickey mouse journal) ”

    It wasn’t exactly in the Current Biology journal proper. It was in the “vibrant magazine section, that includes news, analysis and opinion, profiles of leading scientists and institutions, and informative, accessible guides to notable topics in biology.” No mention of whether the section is even peer-reviewed.

  36. Dr T said,

    August 31, 2007 at 9:53 am

    Fair point, lauredhel.
    That’s the curse of internet journals – you don’t get a feel for the ‘context’ in which it is presented, though the page numbers should have alerted me to that. I checked the guide to authors though,
    www.current-biology.com/misc/page?page=authors#Research
    and the magazine bit at the bottom says that they are peer reviewed.

    buffalo66
    “I think the problem is that there’s enormous pressure on scientists to publish in high-impact journals, ”

    ain’t that the truth!

  37. Robert Carnegie said,

    September 1, 2007 at 1:03 am

    If several of your friends are colour-blind, then won’t that influence your community consensus on colours?

  38. Garote said,

    September 2, 2007 at 11:48 am

    For those of you who wish to take the BSRI (Bem Sex Role Inventory) test for yourselves, but don’t wish to break out pen and paper to learn your score, I’ve whipped up a self-scoring version using JavaScript. Took me about half an hour.

    garote.bdmonkeys.net/bsri.html

  39. RS said,

    September 2, 2007 at 6:52 pm

    “In other words, the genetic influence becomes stronger and the environmental influence weaker as the person ages, at least with regard to IQ.”

    And yet, high IQ heritability enthusiasts interpret that as great evidence for the genetic basis IQ, rather than evidence of the lability of IQ.

  40. kim said,

    September 2, 2007 at 9:08 pm

    I’m entirely with Zoe Williams on this. A piece of research that does nothing but find out that women tend to share a preference for one colour of another, and men likewise, is a waste of time and public money. The hypothesis that colour preference is genetically coded is equally pointless: a, you can’t prove it, b, even if you could prove it, what purpose would it serve? Seriously?

  41. ShatterFace said,

    September 3, 2007 at 2:56 pm

    ”I’m entirely with Zoe Williams on this. A piece of research that does nothing but find out that women tend to share a preference for one colour of another, and men likewise, is a waste of time and public money. The hypothesis that colour preference is genetically coded is equally pointless: a, you can’t prove it, b, even if you could prove it, what purpose would it serve? Seriously?”

    There are serious issues involved if there are sexual differences in colour preferences.

    Imagine, for instance that me find blue more eyecatching and women find pink more eye catching.

    Imagine that a safety panel decides, based on it’s study of *male* employees, that safety warnings should be written in blue in order to be noticed.

    Wouldn’t pointing out that this might put female employees at risk be important enough?

    Or is simply pretending that men and women are exactly the same – whatever the consequences – more important to you?

    ”I suppose that if there is a genetically founded sex bias in our choices of colour of important objects, such as the Houses of Parliament, it would amount to discrimination… hmm, does any country have a pastel legislature?”

    I believe there’s a Whitehouse somewhere.

  42. RS said,

    September 4, 2007 at 5:15 pm

    “There are serious issues involved if there are sexual differences in colour preferences.

    Imagine, for instance that me find blue more eyecatching and women find pink more eye catching.”

    Gosh yes, imagine that someone had actually done a study that showed that – rather than an utterly pointless study of expressed colour preferences.

  43. Ben Goldacre said,

    September 4, 2007 at 5:18 pm

    i certainly don’t think the study was pointless, i just think the interpretation was a little wishful.

  44. RS said,

    September 4, 2007 at 5:57 pm

    Well ok, not pointless, but not exactly setting out to answer a major scientific question – for all the pretty graphs it is only one step up from a survey.

  45. Amy H said,

    September 4, 2007 at 8:22 pm

    Thanks for posting the comments from 1914 and 1918 about pink being the appropriate color for a boy. This sheds light on a fascination I’ve had for a long time about the colors used in men’s professional cycling in Europe and how at odds they seem with what (in the US anyway) would be considered masculine colors.

    Yellow for the leader’s jersey in Tour de France. Red polka dots for King of the Mountains jersey. White for the young rider’s jersey. Pink leader’s jersey in the Giro D’Italia. And lots of men’s teams with bright but light colors such as pink, yellow, lime green, sky blue — T-Mobile calls its color magenta but c’mon it’s really bright pink! And this is in a sport which is unabashedly tough and masculine.

    At any rate I know the leader’s jerseys in the Tour and Giro are related to the companies that sponsor them (e.g., the pink pages of the Italian newspaper that sponsors the Giro) so maybe it’s just a matter of capitalism at work. On the other hand, the Tour, for one, is over a hundred years old so maybe it’s only a reflection of earlier masculine tastes in color. Then again, dark colors absorb more heat — so maybe it’s just a matter of practicality. At any rate, whatever the explanation, I know you would never see a pink jersey as the ultimate prize in the average US men’s sports competition!

  46. coracle said,

    September 4, 2007 at 11:13 pm

    Uh oh, hope the site survives, badscience just got slashdot‘d

  47. buffalo66 said,

    September 15, 2007 at 12:03 pm

    @ 100 Kim: “I’m entirely with Zoe Williams on this. A piece of research that does nothing but find out that women tend to share a preference for one colour of another, and men likewise, is a waste of time and public money.”

    Aesthetic preferences play an important part in most peoples’ lives. They influence our choice of music, films, books, art, clothes, hairstyle, car, interior decoration, and sexual partner, among many other things. Therefore, to understand aesthetic preferences is to understand an important part of the human psyche. Surely this is not a waste of time and money? Hurlbert & Ling’s study is just a tiny piece in an enormous jigsaw puzzle. It provides some good quality data on what peoples’ colour preferences are.

    You ask what purpose this research could serve. One possible application would be to help manufacturers decide what colours to make their products. But the primary goal of science is to find stuff out, not to be useful. Finding stuff out does tend to have practical benefits but, in many fields of science, if you take an individual paper, it is often difficult to think of a convincing practical application of the findings it describes. Most individual findings are pretty insignificant on their own but, taken together, they can lead to a useful increase in our understanding of something.

    But, even if something has virtually no practical benefit, it can still be worth funding. That’s why we have government funding of the arts. In the case of science, if the findings are interesting, then the research was worth doing. And given the amount of media coverage of Hurlbert & Ling’s research, it’s clear that many people were interested in it. In fact, research with no immediate practical benefit is just the sort of thing that should be funded by public money – research with obvious practical benefits is likely to find an industrial sponsor.

  48. buffalo66 said,

    September 16, 2007 at 4:25 pm

    @ RS (#105): “Well ok, not pointless, but not exactly setting out to answer a major scientific question – for all the pretty graphs it is only one step up from a survey.”

    Why does every study have to answer a major scientific question? Surely there’s room for the less important questions as well? As I argued in my previous post, the general topic of aesthetic preferences is far from trivial. Answering major scientific questions takes a lot of time and effort, and usually involves many false starts, so it’s a good idea to have some easy side-projects on the go, to keep your publication rate up. Anya Hurlbert’s main research area is the more serious issue of colour constancy, i.e. the tendency for things to look the same colour after a change in the illuminant (e.g. from daylight to light bulbs).

    As for the study being “one step up from a survey”, it’s true that the subjects were shown things and asked questions about them, but that’s the most straightforward way to study the perceptual experience. It is a very well-established paradigm to show subjects two stimuli, and get them to say which one looked brighter, longer, faster, further away, etc. Asking subjects which of two colours they prefer is formally equivalent to any of these. I can see no problems with either the subject matter or the methodology of this study.

    The authors have been rightly criticised for their rather implausible evolutionary interpretation of the results, but in their paper the evolutionary stuff is clearly described as speculation rather than a firm conclusion.

  49. mickjames said,

    September 27, 2007 at 4:49 pm

    Is it me or is the extreme “pinkisation” of female childhoood (and, more distubingly, adulthood) not a very recent phenomenom? It’s difficult to recall, as I grew up in an all-boy household in the ’60s (so our dominant colour was camouflage green) but I thought all that “painting the nursery pink” thing only happened in Doris Day movies.
    Nowadays everything seems to have a “female,pink” version: pink bikes, pink DVD players, even pink computers. Friends with female children claim they choose pink spontaneously, but I can’t see what other choice they are offered.

  50. Neil... said,

    March 7, 2009 at 10:17 pm

    Well the marketing departments of all these companies who make a pink digital camera think that girls like pink, and continue to do so into adulthood (at least until they are old enough to buy digital cameras and laptops, mobile phones even cars…