Print

Print


*****  To join INSNA, visit http://www.insna.org  *****

David Ouellette asks for advice.

The first piece of advice I would give is to first have a private
correspondence with the authors before putting such a note up on the
Internet.

 Barry Wellman
 _____________________________________________________________________

  Barry Wellman   S.D. Clark Professor of Sociology   NetLab Director
  Centre for Urban & Community Studies          University of Toronto
  455 Spadina Avenue    Toronto Canada M5S 2G8    fax:+1-416-978-7162
  wellman at chass.utoronto.ca  http://www.chass.utoronto.ca/~wellman
        for fun: http://chass.utoronto.ca/oldnew/cybertimes.php
 _____________________________________________________________________


On Mon, 16 Oct 2006, SOCNET automatic digest system wrote:
> ----------------------------------------------------------------------
>
> Date:    Sun, 15 Oct 2006 14:19:43 -0400
> From:    "David M. Ouellette" <[log in to unmask]>
> Subject: Possible Error in a Published Paper
>
> *****  To join INSNA, visit http://www.insna.org  *****
>
> Hey All,
>
> I am working on my dissertation in social psychology, and in the course
> of reading Baker and Faulkner (1993), I suspect an error in computation
> of graph closeness centralization. Their actor closeness (Eq. 3, p. 848)
>   they call "farness" and later refer to it in note 8 as Sabidussi's
> index. Wasserman and Faust (1994, p. 184) indicate that Sabidussi's
> index is actually the reciprocal of this value. Further, their equation
> is not standardized by multiplying by g - 1 (Wasserman &  Faust, Eq.
> 5.8), causing a problem for comparing across networks of different size.
>
> One of their hypotheses was that the network with high information
> processing (turbines conspiracy) should have the lowest graph
> centralization, which I understand. They give Freeman's equation in
> their Equation 4, but they compute closeness graph centralization using
> their farness index rather than the standardized closeness index. I
> interpret their number as unstandardized graph decentralization, but
> they interpret it as regular centralization. In Table 1, they give
> "Sabidussi graph centralization (farness)" numbers, and the turbines
> network has the highest value. They conclude, "This illegal network
> should be sparse and decentralized. This expectation is not supported by
> the data. The turbines conspiracy network exhibits the highest density
> and is the most centralized" (Baker & Faulkner, p. 850).
>
> Am I correct in reinterpreting their number as unstandardized graph
> decentralization? If so, am I correct in surmising that their conclusion
> is backwards? If this is the case, is there a way to take their number
> and the network size and algebraically obtain proper graph centralization?
>
> I am rather confused by this and would greatly appreciate any advice you
> can offer.
>
> David M. Ouellette
> Psychology Department, Social Division
> Virginia Commonwealth University

_____________________________________________________________________
SOCNET is a service of INSNA, the professional association for social
network researchers (http://www.insna.org). To unsubscribe, send
an email message to [log in to unmask] containing the line
UNSUBSCRIBE SOCNET in the body of the message.