I've had students ask the same thing, and have been equally at a loss for an answer. Looking for an out, in this paper (https://www.msu.edu/~zpneal/publications/neal-emporg.pdf) I proposed an alternative (I called it gamma-centrality) that yields qualitatively similar results and has a tunable parameter, but never yields negative scores. As an added bonus, the "gamma" parameter doesn't have any restrictions on its range.
Zachary Neal, PhD
Assistant Professor, Department of Psychology & Global Urban Studies Program
Editor, Journal of Urban Affairs
Michigan State University
316 Physics Road, Rm 262
East Lansing, MI 48824
On networks - www.routledge.com/books/details/9780415881425/
On system science - www.routledge.com/books/details/9780415843324/
On public space - www.routledge.com/books/details/9780415997270/
On diversity & cohesion - www.youtube.com/watch?v=Az6Uy7p3kas
Quoting Jordi Comas <[log in to unmask]>:
***** To join INSNA, visit http://www.insna.org *****
I mistakenly told students to "play" with beta parameter of Bonacich
Student:"How can they be negative scores?'
Me: "I don't know. I'll get back to you. But, we need to keep Beta below
the inverse of largest eignevalue. Sorry about that. Should have told you
?Inquisitive student: "I still don't understand how it can return negative
Me: "Not sure. I'll get back to you on that. Another cookie?" (Ok,
Ididn't offer cookies).
So, can someone help me out? To my eyes, the Bonacich algoritihim sums up
the centrality score of each node by weighting each node. How can and
additive series get a negative score?
(cookies in the mail for you for good answer. :) )
*"There is nothing so practical as a good theory." Kurt Lewin*
Social Innovation and Entrepreneurship Blog (under construction): Simple
Netcentric <http://netcentric.blogs.bucknell.edu/>: Resources for Teaching
Network Theory and Research
Research and Writing Blog: Nets We Weave <http://netsweweave.wordpress.com>
*Assistant ProfessorSchool of ManagementBucknell UniversityTaylor 112570
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.