Print

Print


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

By way of amplifying Zuckerman's cautionary comments:  

 

1) There is a voluminous literature on group cohesion and performance, which
has long acknowledged that the relationship between the two depends upon the
content of group norms. (Of course, structural density and psychological
measures of cohesion are not identical, but it seems likely that they will
tend to be correlated.) 

 

2) It also seems likely that structural density and performance co-evolve
over time, rather than one driving the other in a purely one-way direction--



but see the forthcoming meta-analytic piece by Balkundi and Harrison: 

 

Balkundi, P. & Harrison, D. A. (2005, forthcoming). Ties, leaders, and time
in teams: Strong inference about network structure's effects on team
viability and performance. Academy of Management Journal.

 

 

Ajay Mehra

U. of Cincinnati

 

 

 

.      

 

-----Original Message-----
From: Social Networks Discussion Forum [mailto:[log in to unmask]] On
Behalf Of Ezra Zuckerman
Sent: Thursday, July 07, 2005 10:33 PM
To: [log in to unmask]
Subject: Re: Density...optimal and otherwise...

 

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

 

Another word of caution:  Density has very different implications (as does

performance) depending on whose perspective you take.

 

Consider Coleman's (AJS, 1988) classic notion of "intergenerational
closure"--

i.e., the extent to which parents in a community have ties amongst
themselves

and thus have the necessary "social capital" to keep their kids in line.
This

story makes it sound like density is great for "performance" (i.e.,
achieving

desired ends) and it has often been (mis)interpreted as antithetical to the

idea that social capital is about low density among one's contacts (or

"structural holes").  But now consider the situation from the perspective of

the kids.  Assuming that the kids do not want to listen to the adults (which
is

presumably Coleman's assumption since otherwise the adults wouldn't have to

worry about keeping them in line), density among the adults is *bad*, not
good,

for their "performance" (and density among the kids is good for the kids but

bad for the adults performance, as any parent who has ever separated his
kids

at the dinner table knows).

 

Of course, this example assumes a zero-sum game and life is more interesting

that that.  The more general point is that whether ties between a pair of

actors improves your performance depends on whether those actors have the
same

interests as you (tends to be good, though it's more complicated than that)
or

not (tends to be bad, though it's more complicated than that).

 

A very good exposition of these points is in Burt's Toward a Structural
Theory

of Action (Academic Press, 1982).  My recent papers with Ray Reagans and
Bill

McEvily (Org Science, 2001 & ASQ 2004) also use them to try to clarify some

confusion in the demographic diversity literature.

 

Best,

 

Ezra Zuckerman

 

Quoting "Johnson, Jeffrey C" <[log in to unmask]>:

 

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

> 

> One needs to be theoretically cautious in using density as a measure.
Just

> like the mean of a population, density only tells you a portion of the
story.

>  Two networks can have identical densities but have very different
structures

> given the distribution of links.  In a sense, this is like the
relationship

> between the mean and the standard deviation. This is particularly crucial
if

> one is interested in linking structure to outcomes (e.g., performance).  A

> discussion of this can be found in the following:

> 

> J.C. Johnson, J.S. Boster, and L. Palinkas.  Social Roles and the
Evolution

> of Networks in Isolated and Extreme Environments.  Journal of Mathematical

> Socilogy.  Vol. 27/number 2-3 (2003): 89-122.

> 

> 

> Jeff Johnson

> 

> ________________________________

> 

> From: Social Networks Discussion Forum on behalf of David Lazer

> Sent: Thu 7/7/2005 5:56 PM

> To: [log in to unmask]

> Subject: Re: Density...optimal and otherwise...

> 

> 

> 

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

> 

> One more addition to this thread-- Allan Friedman and I have been doing

> work using agent-based models that suggests, consistent with Labianca,

> Uzzi, and others (notably, work by Kratzer, Leenders, and van Engelen on

> creative teams), that increased density in collaborative networks can

> result in inferior outcomes.  Our analysis suggests that, however, that

> what really matters is not so much density as how rapidly the structure

> disseminates information (obviously, there is a relationship between the

> two, but one can have sparse networks that are very effective at

> disseminating information, as well as fairly dense networks that have

> multiple components and thus do not spread info effectively).  We also

> found that networks that disseminate information quickly do best given

> short time horizons.  We also found a curvilinear relationship between
long

> run performance and density in random nets.

> 

> We will have a revised version of our paper ready shortly (we are

> presenting it at ASA)-- will post at www.ksg.harvard.edu/netgov.

> 

> In addition, related to this thread conceptually, Maria Binz-Scharf has

> done some work on project teams and the density of their informal

> connections, arguing that dense connections are good for exploitation, and

> sparse networks for explorations, suggesting a task and/or temporal

> contingency with respect to the impact of network density on performance.

> 

> chrs,

> 

> David

> 

> 

> __________________________________________

> __________________________________________

> 

> David Lazer

> Associate Professor of Public Policy

> Director

> Program on Networked Governance

> Kennedy School of Government

> Harvard University

> 

> 

> 

>              [log in to unmask]

>              A

>              Sent by:                                                   To

>              [log in to unmask]         [log in to unmask]

>              EDU                                                        cc

> 

>                                                                    Subject

>              07/07/2005 03:58          [SOCNET] Density...optimal and

>              PM                        otherwise...

> 

> 

>              Please respond to

>              [log in to unmask]

>                      A

> 

> 

> 

> 

> 

> 

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

> 

> Soc-netters:  Here is a compendium of responses to my e-question(s)
earlier

> today.  Thanks for your input!

> 

> cdr

> 

> QUESTION: Can anyone point to specific theories or studies wherein

> increased density is assumed to lead to increased output or improved

> performance? I assume

> that if one is studying communication networks that this assumption might

> hold.  In other cases, maybe not (I am thinking of the value of 'weak
ties'

> and

> 'structural holes').

> 

> Alternatively, has there been work done (comparatively speaking) to
uncover

> 'optimal' densities as it relates to networks?

> 

> Any help that you can offer in this would be great!

> 

> ***********************************************************************

> 

> IBM Global Services and I did some research on adaptive

> organizations... we found some high correlations [> 0.55] between some

> network metrics and high scores in 'managing change'/adaptability --

> those orgs who managed change well had different network patterns than

> those that did not.  Density was NOT one of the key metrics...

> 

> Valdis

> 

> ***********************************************************************

> Cami:

> We presented a paper on this topic in the area of community coalitions

> expecting increased density to lead to increased uptake of prevention

> programs and policies.  We found this not to be true, however, and found

> that increased density was associated with less adoption. It surprised

> us, but was consistent with 2 other presentations at the conference.

> The paper reporting our results is currently under review, but I can

> send you a copy if you'd like.

> - Tom Valente

> 

> ***********************************************************************

> Cami,

> 

> There's an article by Podolny and Baron in the ASR, 1997, that shows

> that certain kinds of egocentric networks are conducive to upward

> mobility in a firm.  My recent stuff on bankers (ASR, 2001) has shown

> that sparse networks facilitate successful deals, but I have an

> in-progress paper that shows that high density approval networks among

> the same bankers are associated with higher year-end bonuses.  I should

> have a version of that paper posted on my website within the next month

> or so, but if you send me a reminder in early August I'll send you a

> copy, since it should be revised by then.

> 

> Mark S. Mizruchi

> Professor of Sociology and Business Administration  / University of

> Michigan

> ***********************************************************************

> 

> Cami,

> 

> there is a paper [1] I published at the 2004 P2P Knowledge Management

> Workshop which makes some observations about query routing performance
when

> a self-organized P2P network assumes states with different clustering

> coefficients (with a fixed maximum outdegree -- i.e. routing table size --

> per participant).

> 

> The bottom line is that you can "over-cluster", leading to what Duncan

> Watts (I think it was him) has dubbed "caveman worlds" -- dense clusters

> which are poorly connected to each other, making it difficult to get

> messages across at all.

> 

> Best regards,

> 

> Christoph

> 

> [1] Christoph Schmitz. Self-organization of a small world by topic. In

> Proc. 1st International Workshop on Peer-to-Peer Knowledge Management.

> Boston, MA, August 2004.

> <http://www.kde.cs.uni-kassel.de/schmitz/publ/p2pkm.pdf>

> 

> --

> -- Christoph Schmitz <[log in to unmask]>

> -- FG Wissensverarbeitung, FB 17, Universitšt Kassel

> 

> ***********************************************************************

> 

> Cami,

> 

> This paper by Oh might be helpful regarding your question on "optimal"

> networks.  He points to a middle ground rather than

> maximized density as best for performance.

> 

> Oh, H., Chung, M.-H., & Labiance, G. (2004). Group social capital and
group

> effectiveness: The role of informal socializing

>       ties. Academy of Management Journal, 47(6), 860-875.

> 

> There's also Coleman's work on network closure:

> 

> Coleman, J. S. (1988). Social capital in the creation of human capital.

> American Journal of Sociology, 94, S95-S120.

> 

> Jean

> 

> ***********************************************************************

> 

> Hi Cami,

> 

> I have an empirical piece where we demonstrate an inverted U effect of

> density on performance:

> Oh, H., Chung, M-H., & Labianca, G.  (2004) "Group Social Capital and

> Group Effectiveness:  The Role of Informal Socializing Ties." Academy of

> Management Journal, 47: 860-875.

> and a theoretical piece:

> Oh, H., Labianca, G., & Chung, M-H.  (forthcoming).  "A Multilevel Model

> of Group Social Capital."  Academy of Management Review. (Available on my

> website -- see signature below for URL).

> 

> I would also refer you to Ray Reagans' work in this area:

> Reagans R., & Zuckerman, E. W. 2001. Networks, diversity, and

> productivity: The social capital of corporate R&D teams. Organization

> Science, 12: 502-517.

> Reagans, R., & McEvily, B.  2003.  Network structure and knowledge

> transfer: The effects of cohesion and range. Administrative Science

> Quarterly, 48: 240-267.

> Good luck with your project,

> 

> Joe

> 

> ***********************************************************************

> 

> Hi,

> There is also:

> OBSTFELD D. (2005), Social networks, the Tertius Iungens orientation, and

> inovolvement in

> innovation, Administrative Science Quarterly, 50,, p. 100-130.

> ... if I am correct

> 

> He shows that density positively impacts individual innovation involvement

> (arguing about

> Tertius Iungens as an alternative to the Tertius Gaudens strategy)

> 

> 

> For work on the negative effects of a too-dense network on organizational

> performance, see:

> 

> Uzzi, B. 1997. Social structure and competition in interfirm networks: The

> paradox of embeddedness. Administrative Science Quarterly v42, p35-67.

> 

> Kari

> 

> _____________________________________________________________________

> 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.

> 

> _____________________________________________________________________

> 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.

> 

> _____________________________________________________________________

> 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.

> 

 

_____________________________________________________________________

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.

_____________________________________________________________________
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.