Pink vs Blue: how much a social construct?

A nice post from Ben Goldacre's Bad Science column - Out of the blue and in the pink - tackles the recent universally-reported story that "scientists have cracked the pink problem" (about a recent paper Biological components of sex differences in color preference by Anya C. Hurlbert and Yazhu Ling). The BBC's Why girls 'really do prefer pink' is characteristic.
      One problem with the reportage is its assumption that such a preference can be taken for granted, which Goldacre blows out of the water with a little research. We only have to go back to the early 20th century to find the pink/blue convention reversed, suggesting a strong cultural component. The Ladies' Home Journal wrote in 1918: "...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 colour is more suitable for the boy, while blue, which is more delicate and dainty, is prettier for the girl".
      More detail at the Bad Science website - Pink, pink, pink, pink. Pink moan - which delves into another aspect of the story,. Particularly, the press filtered the contents of the original paper, which did appreciate the cultural aspects and tried to test that by using subjects from a different culture - into a 'just so' story that could easily be ridiculed for stating something perceived as obvious. The Guardian's Zoe Williams took the bait and did exactly that, in Stop this idiocy now.
      It highlights a general problem for anyone researching anything so well-known that it's an unquestioned truism. If you confirm it, it'll be ridiculed as a waste of money. If you disprove it, people will doubt the quality of the research.

Innumeracy again

The Times yesterday had an interesting suite of articles by Michael Blastland and Andrew Dilnot, promoting their book The Tiger That Isn't: Seeing Through a World of Numbers, a spinoff from their Radio 4 More or less programme.
      The common thrust of More or Less and the articles is critique of a general blindness to misuse of numbers, either by mistake or as deliberate obfuscation, by goverments and media. Go figure deals with mistakes in magnitude. For instance, a scary Telegraph story reported falsely that 1 in 5 men of 65 would die before reaching 67. The 1 in 5 was the figure for all men up to the age of 67, not those in that two-year slot.
      The others are about counting and comparisons: how slack categorisation and insufficiently controlled comparison can produce false conclusions. Yob UK? But the sums don’t add up describes a survey that counted childhood rough-and-tumble toward being a "serious offender". It’s a fair cop: the comparison that wasn’t is about interpetation of a survey comparing recidivism in two groups of ex-offenders, where a major problem is the bias inherent in the groups having been selected by assessed probability of reoffending.

See also ABC News' long-standing Who's Counting column by the mathematician John Allen Paulos. In the UK, Ben Goldacre's Bad Science column in the Guardian commonly tackles similar issues over the interpretation of medical data.

Science brands

Via Bad Science, Branded with science, a compilation of examples of scientists who have tattoos related to their subjects: DNA, dinosaurs, mathematics, etc. This is a topic at Carl Zimmer's interesting The Loom ("A blog about life, past and future"), one of many worth reading of the 60+ hosted at ScienceBlogs.