« Finite Horizon models of inflation as the horizon goes to infinity | Main | Time use: what's the Big Question? »

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

Anecdotes aren't singulars for data but still.
I am a fan of warships and I visit everyone I can when there visits. European sailors are noticeably taller than americans. Unless you argue that OS (Ordinary Seaman) OR-2 is an unusually prestige job in Germany compared to the US that it attract upper-class men...
I have heard reports that European men (esp. Dutch, Germans and Scandinavians complain that north american hotel bed are too short.
On photographs from WWII, it's easy to differentiate men from officers, especially in british troops, as well as distinguishing Americans, Canadians and Australians from various europeans. In his memoirs,French fighter ace Pierre Clostermann, himself from a high socio-economic background (his father was an ambassador) often reffered to his Canadian and Australian colleagues height, which he never mentionned for the Brits.
And just try to find a good italian silk ties. Today ties are simply too long but fit just right when Italian men wear them.

Jacques Rene - that's very to the point, as much of the male height data comes from data on military personnel/conscripts.

Very true in the case of the Australian's of the 1st AIF in France during WWI. The only data for the 1896 cohort available for Australia with any sort of ease of access would be that from the Australian Army's records, in particular their Attestation Forms upon enlistment, with the 1896 cohort of course being 18, and the minimum age to volunteer for the AIF in 1914.

But the minimum height requirement for the AIF was 5'7", which was more than 2 inches greater than the average height of the British Army during WWI. This despite the fact that the British Army found that when conscription was introduced their 18 year olds actually grew significantly (on the order of one and three quarter inches) after joining the Army and going onto Army rations which included a greater proportion of red meat than they consumed in civilian life.

Well, I live in Korea and see thousands of them each day. Older people are way shorter than younger people. One tall woman told me that it was only recently that milk was widely consumed in Korea.

David, the height study contains good, credible evidence that S. Korean heights have increased substantially in the past 50 years. What is far less credible is the idea that heights were increasing rapidly between 1896 and 1916, long before S. Koreans started drinking milk and eating substantially more meat etc. But those years between 1896 and 1916 are generating a good chunk of the growth in height that is reported by the article - and in the popular media.

Firstly, I agree with your view on the growing pressures on researchers to demonstrate impact, but your point regarding peer review is not correct in this specific instance. If you read the article towards the end it quite clearly states that the 'article has been reviewed by two peer reviewers, and the evaluation has been overseen by Eduardo Franco, as a Reviewing Editor, and Prabhat Jha, as the Senior Editor. M Dawn Teare, a Member of eLife's Board of Reviewing Editors served as one of the reviewers and agreed to reveal her identity.' A 'lightly edited' version of the review is also provided.
Secondly I disagree with your comment that it's impossible to tell what is the cause of the stagnation in US height without a much more sophisticated analysis. It isn't the analysis that's the problem it's the study design. Causal inference from observational studies is extremely problematic, no matter what analytical approach is used.[1]

1. Judea Pearl. Causal Inference in the Health Sciences: A Conceptual Introduction. Health Services & Outcomes Research Methodology 2:189–220, 2001

3948_Bohr: I did not say that the article hadn't been peer reviewed. I said that it had not been through a traditional peer review process - and I stand by that claim. Take a look at the eLife website and the description of the peer review process that the journal uses - it's definitely non-traditional.

If you are - as I guess you might be - a serious academic, surely you'll agree that any journal that promises "pain-free publishing" can't be using a traditional type of peer review ;-)

I very much agree with your comments on the difficulty of inference - however it would be possible to do some back-of-the-envelope calculations along the lines of "if we assume that all immigrants are of average height for their country of birth, how much would we expect average American heights to change over time solely as a result of immigrant flows?" If the predicted effect of immigration on mean height is -10%, and in fact we observe a decrease in height of -2%, it's pretty hard to claim that we have evidence of declining nutritional quality - though as you say that possibility can't be ruled out.

Frances, thanks for responding. You're right, I am an academic and I do agree that "pain-free publishing" rings alarm bells. Still, I think you're splitting hairs a bit on the peer review issue. I just looked at eLife's site and, despite the advertising puff, the process they engage in looks pretty much like traditional peer review to me.

I agree that the authors could have done more to assess the influence of immigration on the observed height patterns had that been a goal of the paper; however, I don't think it was. As far as I can see the question of nutrition in the US doesn't feature in the article - it was raised by Ezzati in interviews, in an interview in the Telegraph, he admitted that immigration might play a role.

You can criticize Ezzati for those comments perhaps, but I feel to label the article 'bad science' (implying to me, politically or commercially motivated bias) seems unwarranted.

3948_Bohr: " I feel to label the article 'bad science' (implying to me, politically or commercially motivated bias) seems unwarranted."

I don't think the authors here are venal - if they have a political motivation, it's a fairly benign agenda of trying to improve human health. I don't think they're any more publicity seeking and self-promoting than the average academic - they've just happened across a great venue for self-promotion.

But I stand by my position that their paper is bad science in the sense of being thoughtless and ahistorical. Take, for example, those S. Korean numbers. Over the past century S. Korea has experienced colonization, war, (WWII and the civil war), etc. To take that turbulent history and fit a standard growth curve on top of it is just thoughtless. It would be foolish to expect S. Koreans who were doing their growing during the difficult periods of WWII and and the civil war to experience the same kind of height growth as those fed on milk and meat and late 20th century prosperity.

In addition to stats issues, we have the astonishing fact that in many cases average national height declined when people who were shorter than the previous average immigrated.

The comments to this entry are closed.

Search this site

  • Google

    WWW
    worthwhile.typepad.com
Blog powered by Typepad