you might also see my critique of the small world method,
in m. kochen (ed.) the small world, 1989, ablex.
On Thu, 10 Jan 2002, Eszter Hargittai wrote:
> I thought the following article in today's Chronicle of Higher Ed may be
> of interest to people on this list.
> Eszter's List: http://www.eszter.com/elist
> This article from The Chronicle of Higher Education
> (http://chronicle.com) was forwarded to you from: [log in to unmask]
> Thursday, January 10, 2002
> Professor Challenges Work That Led to Idea of '6 Degrees of
> By CHRISTOPHER FLORES
> Judith Kleinfeld, the iconoclastic scholar best known for her
> controversial criticism of gender studies, has published a
> paper disputing the veracity of the research that led to the
> concept of "six degrees of separation."
> The widely accepted theory that any two individuals are linked
> by an average of six acquaintances is based on the early work
> of the late Stanley Milgram, the psychologist whose
> groundbreaking but ethically suspect work on obedience to
> authority in the 1960s cost him his position at Yale. The
> theory has come under fire by Ms. Kleinfeld, a professor at
> the University of Alaska at Fairbanks, in her article "The
> Small World Problem," in the January/February issue of
> Society. She asserts that upon close scrutiny of Mr. Milgram's
> papers, archived at the Yale University library, it becomes
> clear that the process he used was flawed, and that even
> within that flawed methodology, the results he found did not
> support his theory.
> "'Six degrees of separation' has really gone from the realm of
> science into the realm of pop culture," says Ms. Kleinfeld,
> citing its appearance in everything from high art (the John
> Guare play of the same name) to parlor games (Six Degrees of
> Kevin Bacon). "The theory has been popularly accepted for so
> long, even though there is no real evidence to support it."
> Mr. Milgram's experiment -- commonly referred to as the
> "small-world method" -- entailed randomly selecting people to
> send a folder to a target person unknown to them in a distant
> location by first mailing it to someone they thought might
> know the target. The process was then repeated until the
> target received the folder. According to Ms. Kleinfeld's
> research, the selection process was anything but random,
> drawing participants through advertisements and purchased
> mailing lists, a practice that she asserts would cull mostly
> high-income and highly connected people.
> Even under conditions as favorable to the theory as these, Ms.
> Kleinfeld reports that on average only 30 percent of the
> folders in Mr. Milgram's experiments -- and in most
> replications of the small-world method -- ever reached their
> target, and then through an average of eight people (or nine
> degrees of separation). Ms. Kleinfeld points out that Mr.
> Milgram never published those data.
> "I think Milgram was a fine scientist," says Ms. Kleinfeld,
> whose original intent in exploring the psychologist's work was
> to repeat it with her students. "My own view of this is not to
> debunk Milgram, but to argue that the small-world problem
> continues to fascinate us all, and nothing is so useful as a
> good problem." Ms. Kleinfeld's findings will also appear in
> the next issue of Psychology Today beside a paper by Thomas
> Blass, a leading Milgram scholar who is sympathetic with Ms.
> Kleinfeld's findings.
> Mr. Milgram presented his theory in 1967 in a noted article in
> the first issue of Psychology Today. He rested the theory
> entirely on the instance of one folder that made it from a
> Kansas wheat farmer to its target, the Boston wife of a
> divinity student, in four days and through only two
> intermediate links. The article made no mention, however, of
> the statistical results of the experiment.
> "Why do we want to believe this so much?" asks Ms. Kleinfeld.
> "With such scant empirical evidence?" She suggests that the
> theory is so readily accepted because it bolsters our sense of
> security, confirms our religious beliefs, and makes sense of
> our personal experiences of coincidence. At least one
> scholar, Duncan Watts of Columbia University, believes that
> Mr. Milgram's theory is sound but contends that no individual
> has the capacity to consistently map his links to every other
> person. Mr. Watts is working on a mathematical model to
> explain the theory.
> Ms. Kleinfeld says she does not mean to suggest that Mr.
> Milgram's study is without merit. The findings did indicate
> that people are actually drastically separated by social class
> and, as other experiments using the small-world method have
> shown, by race, she notes. "Rather than living in a 'small,
> small world,'" writes Ms. Kleinfeld, "we may live in a world
> that looks a lot like a bowl of lumpy oatmeal, with many small
> worlds loosely connected and perhaps some worlds not connected
> at all."
> This article from The Chronicle is available online at this address:
> If you would like to have complete access to The Chronicle's Web
> site, a special subscription offer can be found at:
> You may visit The Chronicle as follows:
> * via the World-Wide Web, at http://chronicle.com
> * via telnet at chronicle.com
> Copyright 2002 by The Chronicle of Higher Education
Alden S Klovdahl / [log in to unmask] / fax: +61 2 6125 2222
Social Sciences / Australian National University / Canberra Australia 0200