darb.ketyov.com

Caveat lector: This blog is where I try out new ideas. I will often be wrong, but that's the point.

Home | Personal | Entertainment | Professional | Publications | Blog

Search Archive

7.10.10

Fraud in science: What's in a name?

Bradley Voytek Hello my name is


So everyone's all atwitter about this NYT article, "Rampant Fraud Threat to China’s Brisk Ascent".

I find it interesting that no one's been mentioning the feature in Nature a few years back about China's "identity crisis".

The NYT piece highlights a lot of issues in scientific (mis)conduct, but the Nature feature was far more fascinating in terms of how scientists can game the flaws in the scientific publishing industry.

In the Metafilter thread on the Nature piece, one of the commentors linked to a few news articles about Chinese names. According to Xinhua:

Only 4,100 different Chinese surnames have been found to exist out of a sample of 300 million Chinese people, according to a survey by the Chinese Academy of Sciences (CAS).

This issue is apparently quite problematic for law enforcement and, according to Rueters, the Chinese police have been considering alternatives to traditional naming conventions:

At least 100,000 people share the name "Wang Tao", the China Daily said, citing the Chinese Academy of Social Sciences... China has 1,601 surnames in total. According to the new regulations, Arab numbers, foreign languages and symbols that do not belong to Chinese minority languages would all be banned.

But, as Nature pointed out, this isn't just a law enforcement issue, it's also an easy way to commit academic fraud.

Jia Wei, associate dean at the pharmacy school of Shanghai Jiao Tong University can remember hundreds of metabolic pathways by heart, but he gets confused by his graduate students' publications. Three of his students—Wang Xiao-yan, Wang Xiao-rong and Wang Xiao-xue... have completely different two-character given names in Chinese, but all publish under the abbreviated name X. Wang. “I really have a hard time sorting out who has published what,” Jia sighs.

After setting an establishing example for the problem, they continue:

The [naming] problem is sufficiently widespread that some researchers have taken advantage of the ambiguity. Surgeon Liu Hui, who padded his CV with publications by another researcher who shared his surname and initial, rose to become an assistant dean at the prestigious Tsinghua University. But the discrepancies were noticed and he was dismissed by the university in March 2006.

As science becomes less WEIRD and more globalized these kinds of issues become more problematic.

In south India, women may not have surnames and are forced to adopt them to publish in science (with this caveat):

Indians from the south traditionally do not have surnames. It is only when forced to comply with Western naming standards that they use their father's given name as a substitute. As a consequence, journal rules require them to publish research under the fathers' given names (with which we—Nalini, Jeevananthinee and Sujatha—also sign this Correspondence letter). Obviously, as young south Indian scientists making a contribution to science, we would prefer to be identified with our first names and not by our fathers' given names.

Naming issues also affect citation rankings.

On the one hand, it surprises me that a technological solution hasn't been widely instantiated to address this. On the other, this is obviously a complex issue and getting a wide array of publishers, databases, etc. all to cooperate and adopt the same standards would be a ridiculously impossible job. As my wife would tell me, database standards are not quick to get adopted. People are slow to change.

EDIT: Note my wife's comment (and correction to my assumption about what she would say) below.

8 comments:

  1. Your wife would in fact tell you that it requires much more than a technological solution. Names, specifically authoritative names have stymied librarians for centuries. How do you assign an author to a work when he/she uses a pen name? Do you use the author's actual name or his pen name? What is the "authoritative" name of the person who wrote "Adventures of Huckleberry Fin", Mark Twain or Samuel Clemens? Where do you put the book in the library, under "T" or "C"? Technological solutions allow for multidimensional organization (the object can be in two places at once, under "T" and "C"), but they may not satisfy all of humanities naming needs. For example, we could (and do in the United States) assign each person a unique number, but names are for more than identification. Consider the South Indian women in the post above, who are required to use their father's names as their authoritative names when publishing. Not only is it inaccurate, but the women find it insulting not to be able to use their own names when signing their own work. In Germany, first names must be approved by government authorities and last names may only contain one hyphen (a law that was challenged and upheld in 2009). Suffice it to say, people tend to balk when external authorities regulate their names. These examples tend to suggest that names are much more than identifiers, they are identities and technological solutions tend to systematize what would otherwise be very personal.

    ReplyDelete
  2. This kind of thing is why we're married. Thanks for expanding on the thoughts I mis-attributed to you :D

    So if you were reworking this from the ground up, how would you approach the scientific naming issue?

    ReplyDelete
  3. I suppose you could set up some sort of assymetric or symmetric cryptographic identity verification system. Public/private key pairs like those used for PGP (email authentication), and private keys like they do for SSL certificates (the things that verify the identity of a website that wants to have your credit card number). Individuals who want (and this is the key point here) to be identified in a certain domain, say science, register and are issued a public and private key. They use whatever human readable name they prefer in print/text, but when they publish (I guess this would only work for electronic publishing, but who doesn't now-a-days) they must also verify their identity by attaching a hash that can only be generated using the private key, but can be decoded using the public key.

    ReplyDelete
  4. Wonder how difficult that would be to implement on a wide scale?

    ReplyDelete
  5. Anonymous04:58

    I worked with a neuroscientist from latin america who had 5 names (typical in latin countries, instead of being, say, John Edward Smith, people there can be John Edward Michael Thomas Smith). He permutated his names to perform multiple submission to low-impact journals. It is amusing to find the same work in two different places by a " John Smith", then by an " Edward Michael Thomas" He actually succeeded in doing so, but he was trying to move his way up to stronger journals, not sure how he is handling that. A true sociopath in other respects of life, a horrible researcher. As an advice to young people: avoid shady people and researchers.

    ReplyDelete
  6. Wow, Anonymous, that's incredible. And yeah, good advice. The best thing a young scientist can do is work with someone they respect both scientifically and personally.

    ReplyDelete
  7. Anonymous11:21

    Yes, and I post anonymously because I'm yet to get my phd and working for two years with this guy cost me a lot of time and stress. My early research life was a nightmare, all because of this one person. He does worse things with data too: for example, he needed the weights of all animals, and simply made the numbers up (the animals were dead and he never kept track of weights). He had one good image of cell-recovery in the area under study, and he used to add that hemi-image to one from another animal, to show that the treatment worked on the site of the lesion only. And ask him any questions....all you get are accusations that you never read such and such papers that don't seem to exist. Neuroscience is growing a lot, and it attracts some bad people.

    Young people, just never follow someone's word because they are senior to you, make sure you feel the right course is being taken.

    ReplyDelete
  8. I'm sorry to hear that... and that's very concerning. Have you thought about blowing the whistle on this and anonymously reporting him?

    ReplyDelete