====== Ideas, problems ======
[[:net|Back to networks]]
* [[:net:time|Temporal networks]]
* Feb 20, 2013. Tipa blokov: row-structural (1 ali 0 vrstice), col-structural (1 ali 0 stolpci)
* Feb 20, 2013. Ocena povprečne stopnje, povprečne razdalje.
* Slovenski stemming in lemmatization \\ http://oznacevalnik.slovenscina.eu/Vsebine/Sl/ProgramskaOprema/Navodila.aspx \\ http://oznacevalnik.slovenscina.eu \\ http://lemmatise.ijs.si/Software/
* http://blog.stephenwolfram.com/2013/04/data-science-of-the-facebook-world/
* http://clair.si.umich.edu/homepage/bib2html/webgraph-bib.html
* http://www.nap.edu/openbook.php?isbn=0309089522
* http://www.danah.org/researchBibs/sns.php
* http://en.wikipedia.org/wiki/Non-negative_matrix_factorization
* http://cs.stanford.edu/people/jure/talks/
* http://www.infovis.net/printMag.php?num=202&lang=2
===== Name generators =====
A few others have pointed out the origin of the name generator technique,
such as Coleman, Laumann and Wellman. If you are interested in Laumann's
instrument, you can find it from his journal article (in one of the
footnote). Fischer's approach is also very important and need to be
mentioned here.
Laumann, E. O., & Pappi, F. U. (1973). New Directions in the Study of
Community Elites. American Sociological Review, 38(2), 212-230.
Fischer, C. S. (1982). To Dwell Among Friends: Personal Networks in Town and
City. Chicago: University of Chicago Press.
In my opinion, the "name generator",as the name describing this technique,
came to light when Burt used the "important matters" question in the General
Social Survey. Here are the three articles detailing the development of the
question, and how it was used.
Burt, R. S. (1984). Network Items in the General Social Survey. Social
Networks, 6(4), 293-339.
Burt, R. S. (1985). General Social Survey Network Items. Connections, 8,
119-123.
Marsden, P. V. (1987). Core Discussion Networks of Americans. American
Sociological Review, 52(1), 122-131.
Note that some researchers (such as Fischer) prefer using multiple
questions to elicit social ties in one's personal networks than using just
one question. See Marsden (2003) and Marin and Hampton (2007) for more details.
Marsden, P. V. (2003). Interviewer effects in measuring network size using a
single name generator. [doi: DOI: 10.1016/S0378-8733(02)00009-6]. Social
Networks, 25(1), 1-16.
As you could imagine, the difference in wording (i.e. the criteria of the
personal network of interest) would lead research participants to report
different types of social ties. Comparing generated personal networks across
studies may be difficult. One of the first study that seeks to find out
whether there is any commonalty between different name generators is
Campbell and Lee (1991).
Campbell, K. E., & Lee, B. A. (1991). Name generators in surveys of personal
networks. [doi: DOI: 10.1016/0378-8733(91)90006-F]. Social Networks, 13(3),
203-221.
Finally, to the original question, I think it is nice to be able to find
some benchmark from UK's Government Statistical Service, and say something
about how good your results are. But I don't think "direct" comparison is
necessary or appropriate. My main concern is the original wording of the
question.
Strictly speaking, it is a "triple-barrel" question: it asked respondents to
recall and report three different types of social ties. And respondents only
answered the question once. If you break down the question and ask
respondents to report these three types of social ties separately (i.e.
Fischer's approach), you will definitely obtain a larger personal network
size. See Hlebec et al (2006). for an recent example.
Hlebec, V., Manfreda, K. L., & Vehovar, V. (2006). The social support
networks of internet users. New Media & Society, 8(1), 9-32.
Hope this helps,
Patrick
For an older paper see:
Social relations and innovation in the medical profession:
the epidemiology of a new drug.
H Menzel, E Katz - Public Opinion Quarterly, 1955
Vlado
One of the first studies that used affective and/or role relation criteria
as name generators was the 1966 Detroit Area Study (Laumann 1973). He asked
respondents to name their three "best friends". Wellman's 1979 study asked
about "persons outside your home that you feel closest to".
Laumann EO. Bonds of pluralism: the form and substance of urban social
networks. New York: Wiley, 1973. http://www.getcited.org/pub/101325959
Wellman B. The community question: the intimate networks of East Yorkers.
AJS 1979;84:1201-31. http://www.jstor.org/stable/2778222
I'm not quite sure if this will be useful, but the General Social Survey in the U.S.
(currently out of the U of Chicago and funded by the NSF) has used questions of this
type for decades. http://www3.norc.org/GSS+Website/
The Framingham Heart Study is also considered foundational in the use of name
generators for personal social network research.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2600606/
===== Twitter =====
Bruno,
Yes, RTs are important Twitter links, as are @ tweets (conversations) on Twitter.
IMHO, looking at follower/following is useless at best, misleading at worst -- so
many false followers. Remember GIGO.
Conversations can often be captured in specific twitter chats on topics that attract
those of similar interests. Here is a quick SNA I did of a conversation on
"serendipity" amongst those interested in "innovation". This is just a quick peek
at the group as it was structured on that day and on that topic. (I have more data
on this group and will do a more in-depth look at some future date).
http://www.thenetworkthinkers.com/2013/02/arrows-on-twitter.html
Of course, this was not the only convo on serendipity on Twitter, and many who had
an interest in this topic did not know the convo was going on.
Wincent,
I would look on Twitter where your NGO groups are active... do they partake in
Twitter Chats? If so, do not do your own data gathering... Twitter only provides
limited data via their API. Find an organization that gets data from the Twitter
"fire hose" (all of the data). I have used Socialping, Inc. http://socialping.com
and was happy with their timeliness and competence. Then they can filter the data
with your list of NGOs ... looking for @ conversations, RTs, MTs, and HTs. This
should give you a good first draft to start with.
Good Luck!
Valdis Krebs
http://orgnet.com
http://thenetworkthinkers.com
Ian:
The earliest empirical name generator work I'm aware of is Coleman and
Katz's diffusion study:
Coleman, James, Elihu Katz, and Herbert Menzel. "The diffusion of an
innovation among physicians." *Sociometry* 20, no. 4 (1957): 253-270.
Coleman talked more generally about it later:
Coleman, James S. "Relational analysis: the study of social organizations
with survey methods." *Human Organization* 17, no. 4 (1958): 28-36.
Jim
Ronggui,
You might look at this one:
Playing in the Same Twitter Network: Political information seeking in
the 2010 US gubernatorial elections
Itai Himelboim, Derek Hansen & Anne Bowser
http://www.tandfonline.com/doi/abs/10.1080/1369118X.2012.706316#.UZJSTLXijK0
A reasonable approach to understanding what a "Follow" relationship
means is to consider it from the perspective of the person following
the another account. When doing so, it quickly becomes clear that
there are many factors at play. For example, it's clear that if I
follow some account I am "aware" of it (though there are some accounts
I don't follow that I'm aware of, so this is only part of the story).
If I'm an active user who actually reads Tweets that come to my stream
then I follow accounts so that I can receive information from them
(i.e., "information flow"). Finally, since my Follow network is
public, there is a sense in which I may Follow accounts of
organizations that I tend to agree with or at least that wouldn't
reflect poorly on my "identity". In some yet-to-be-published work Jen
Golbeck and I have some evidence that, on average, people tend to
follow politicians who they agree with (as opposed to those who they
don't agree with), perhaps for all of the reasons listed above or some
subset of them.
I agree with Bruno that Re-tweet and Mention networks are more
indicative of active connections between people. Inevitably, the
follow networks are much more dense than those networks (as you can
see on many of Marc Smith's Eventgraphs: e.g.,
http://www.nodexlgraphgallery.org/Pages/Graph.aspx?graphID=4569 )
Great question, and one that I'd love someone to tackle directly
through some interviews and/or survey research about why people follow
others - though I haven't looked for such an article, so I suppose it
may already exist...
Derek Hansen
Brigham Young University
Dear Dr. Huang,
This paper might be of interest to you:
*Partisan Asymmetries in Online Political Activity *
M. Conover, B. Gonçalves, A. Flammini, F. Menczer
*EPJ Data Science* *1*, 6 (2012)
http://www.bgoncalves.com/component/jdownloads/finish/3/38.html
In this paper (and a couple others that you can find here:
http://www.bgoncalves.com/publications.html ) we look at the
follower-followee, retweet and mention networks and find that the retweet
network is the most informative if you are trying to cluster users
according to their political leanings (or activity, or collaboration,
etc...) since a retweet implies a stronger connection (agreement) than just
following or mentioning.
Best,
Bruno
===== Blockmodelling medium size two-mode networks =====
Subject: [Pajek] blockmodelling medium size two-mode networks
From: "Jef Vlegels"
Date: Thu, May 16, 2013 11:33
To: pajek@list.fmf.uni-lj.si
Is there any way to run a two-mode blockmodel on a network with more than
256 vertices (in Pajek or maybe alternative software)?
Pajek returns the error message: 'number of vertices in both modes must be
lower than 256', even though calculation time is still acceptable.
Thanks
===== What is an old/dormant strong tie? =====
Subject: What is an old/dormant strong tie?
From: "Valdis Krebs"
Date: Tue, August 13, 2013 21:15
John,
Yeah, for the first iTunes phone (ROKR), design by committee (across boundaries
between Apple and Motorola) probably included many weak ties. After that failure,
for the iPhone, Jobs went back to his strong ties, especially Jonathon Ive, to whom
he trusted many key Apple designs. Ive in turn had his own strong ties within
team/company to execute the design.
The iPod was also designed by strong ties within Apple and outside of Apple... an
"old strong tie" came to call on Apple -- "Hey, I have an idea."
Hey Academics... what is an "old strong tie"???
Old strong tie is where A and B worked together at Company X, but now A works at
Company Y and B at Company Z (and naturally they interact less now) and they
re-connect to collaborate (re-animating their prior tie) for a new product at
Company Z. Is an old strong tie, even an inactive/dormant strong tie, still a
"strong tie"? Or does it become a "weak tie" while in it's dormant phase???
Would love to hear many viewpoints on this!
Valdis Krebs
Orgnet,LLC
Twitter: orgnet
http://orgnet.com
http://thenetworkthinkers.com
On Aug 13, 2013, at 2:34 PM, John T. Maloney wrote:
> Apple rejected the "design by committee" approach that had yielded the Motorola
ROKR E1, a largely unsuccessful collaboration with Motorola.
Subject: Re: What is an old/dormant strong tie?
From: "Kenny Joseph"
Date: Tue, August 13, 2013 21:24
While not directly referring to the concept of an "old strong tie",
Caroline Haythornthwaite provides the very useful concept of a latent tie.
If I recall correctly, the term is used to distinguish social ties that
may develop when a new media is introduced into a social system.
Haythornthwaite, Caroline. 2002. “Strong, Weak, and Latent Ties and the
Impact of New Media.” *The Information Society* 18 (5): 385=401.
Kenny Joseph
Subject: Re: What is an old/dormant strong tie?
From: "Joe Labianca"
Date: Tue, August 13, 2013 23:04
Hi Valdis,
I'd look at Daniel Levin's work:
Levin, D. Z., Walter, J., & Murnighan, J. K. (2011).
“Dormant Ties: The Value of
Reconnecting.”
*Organization Science, 22(4),* 923-939.
Levin, D. Z., Walter, J., & Murnighan, J. K. (2011).
“The Power of Reconnection — How Dormant Ties Can Surprise
You.”
*MIT Sloan Management Review, 52(3),* 45-50.
Joe
Giuseppe (Joe) Labianca
Gatton Endowed Professor of Management
Gatton College of Business & Economics
LINKS Center for Social Network Analysis
University of Kentucky
Lexington, KY 40506
859-257-3741 (office)
404-428-4878 (mobile)
http://linkscenter.org/
tinyurl.com/JoeLabianca
Subject: Re: What is an old/dormant strong tie?
From: "John McCreery"
Date: Wed, August 14, 2013 01:37
Another old, strong tie is family. Family members may scatter to the four
corners of the earth and have little to do with one another. Then a
central, usually senior, member dies, and the clan reassembles for the
funeral. The need to cooperate in disposing of legacies may bring brothers
together, who usually have little to do with each other. Quarrels over
inheritances may remind us of Simmel's observation that hostility is a
social tie.
While attending INSNA 2013 in Xi'an I was struck by the difference between
the ideas of participants from Europe or the US and those from China. The
former expounded theories grounded in the mathematics of random graphs,
which ideologically speaking, coincide with market fundamentalism —
autonomous actors bumping into each other and engaging in transactions. In
contrast, when the Chinese talked about *guanxi,* they constantly made
reference to pre-established ties created by kin or classmate relations and
assumed that hierarchy and emotion are essential aspects of networks and
not just outcomes of dyadic transactions.
John
Subject: Re: What is an old/dormant strong tie?
From: "Pryke, Stephen"
Date: Wed, August 14, 2013 13:46
To: SOCNET@LISTS.UFL.EDU
***** To join INSNA, visit http://www.insna.org *****
Some interesting discussion about tie strength and we have been giving this some
thought to tie strength in the context of project based organisations here at UCL.
My view is that tie strength is transient and not necessarily reciprocal. I have
difficulty with the notion that family is a strong tie, any more than marriage or
other forms of contracts create strong ties. In fact tie strength, like beauty, is
'in the eye of the beholder'. We all have our own value systems that we apply to
tie strength measurement.
My father, as part of his business activity, called at an address in London one day;
a man of approximately the same age answered the door; during the brief
conversation that ensued the two men established that they were brothers. They had
made no contact for 30-40 years. Once their business was completed they shook hands
and wished each other well. And never saw each other again! Nothing malicious
here; no dispute over inheritances. Just a very weak tie!! [large family; very
wide age range amongst siblings; lots of relocation caused by WW2; natural tendency
not to maintain family ties etc]
So I think as academics we need to do a lot more work on tie strength; it is
transient and therefore needing maintenance; related to actor characteristics and
the environment in which the networks function; it is also possible to have several
different tie strengths present between actors concurrently, based on a variety of
network functions.
Thoughts of others welcome
Stephen
UCL
Subject: Re: What is an old/dormant strong tie?
From: "John McCreery"
Date: Wed, August 14, 2013 14:20
To: SOCNET@LISTS.UFL.EDU
Stephen, speaking as an anthropologist I respect your experience. But I
also recognise how specific it is. It strikes me that increasingly the
world is divided between those who, for whatever reason, live lives engaged
with networks whose ties are always transient, and those of us whose for
whom relationship as social fact seems more factual.
On a personal note, my father was one of seven children, my mother one of
four. She was disappointed that she was only able to have two, my younger
brother and me. We grew up attending annual family reunions, as well as
graduations, weddings and funerals. My parents were able to buy the
Virginia waterfront property on which I grew up because my maternal
grandfather provided the necessary loan. They lived to celebrate their
sixtieth wedding anniversary, and my wife and I have been married
forty-four years. Oddly enough, I, my brother, my wife and her four
brothers are still married to the people we married for the first time.
And when my Chinese friends talk about guanxi, I take them seriously. Why?
Earlier this year, I was invited to a conference in China. To get to the
university that paid for my air fair to give a talk in exchange, I had to
fly from Japan to Shanghai, get across Shanghai from one airport to
another, fly to a city in Western China, and then take a train to the city
where that university is located. Arriving at my first airport in Shanghai,
I was greeted by a young woman who took me in hand and led me through the
subway maze to get to the second airport. I asked if she were a student of
the professor who had invited me to the conference. No, she said, she was a
younger sister of a classmate of the assistant professor who was handling
the organisational details for my trip. Arriving at the airport in the city
in Western China, I was picked up by another woman, who, it turned out, was
another classmate of the organiser, who got me to a hotel for the night and
was there in the morning to pick me up and put me on the train to the city
where the university where I spoke is located. Neither of these women knew
me from Adam. They had gone out of their way to pick up and take care of a
perfect stranger because they felt a strong obligation/emotional tie to the
person who asked the favour. The first case was particularly striking,
since I was, in effect, a stranger at two removes, with no direct
relationship either to the woman who met me at the airport or the sister
who had asked the favour. In China, family members and classmates can ask
this kind of favour, expecting that, given a chain of pre-existing
relationships, those relationships will be activated to deliver the favour
in question.
In both these cases, one can argue that the relationships in question are
based on reciprocal expectation of support as needed — but to try to
account for them in terms of transient transactions is a very long stretch,
indeed. I would argue, instead, that we inhabit a world in which some
people get by with transient relationships in fluid networks; others enjoy
the privilege of relationships they can count on, whether or not they
interact directly with all of the individuals involved. Each sees the
other's world as artificial and unnatural — but both are part of the real
world and must be understood as such.
John
Subject: Typology of relational embeddedness - Old/dormant strong ties & latent embeddedness
From: "Julie Hite"
Date: Wed, August 14, 2013 14:28
Great discussion!! Building on Stephen's comments regarding tie multiplexity and tie
evolution, my articles on the typology of relational embeddedness (Hite, 2003, 2005,
2008) identify different types of relational embeddedness. This typology expands
our notions of strong ties and also suggests that some types of relationally
embedded ties may have previously been identified as weak ties. This typology helps
to move us from a dichotomous view of weak/strong ties and explores the continuum in
between, enabling a broader range of identification of types of ties.
The typology suggests that ties can have high extents of personal relationship,
social capital and dyadic interaction - or any one or combination of these three
social components (aligned with affect, reciprocity and frequency). Ties with all
three social components reflect "full embeddedness," governed by three different
types of trust. However, ties with a high degree of only one of these social
components can still be relationally embedded (affecting decision making;
maintaining relationship over economic decision making, using trust-based
governance, resource accessibility, etc.). Many of these ties with only one type of
social component may often be identified as weak ties in research based on the
dichotomy of weak/strong ties.
One type of relationally embeddedness that the typology identifies is "latent
embeddedness" in which ties have a high extent of personal relationship and social
capital, but do not have a high extent of dyadic interaction. These ties may be
overlooked in network research, particularly work- or organizational-based networks,
as there may not currently be direct network content flowing (due to low interaction
or frequency); they may also be labeled as weak ties in one given network context
(relation) while being strong ties in another network context (relation).
Yet, given the multi-dimensionality (e.g. multiplexity) and evolutionary (transient)
nature of ties, these latent ties can often be activated quickly based on their high
degree of personal relationship and social capital. Thus, what may appear to be a
weak (or absent) tie in one network relation may actually also have characteristics
of a strong tie in another relation. Without accounting for the multiplexity of
network relations, and the multiple dimensions of relational embeddedness (strong
ties), the weak/strong tie research may be missing critical components of the
nature, content and structure of ties.
Hite, Julie M. (2003). Patterns of multi-dimensionality among embedded network ties:
A typology of relational embeddedness in emerging entrepreneurial firms. Strategic
Organization, 1(1), 9-49.
Hite, Julie M. (2005). Evolutionary processes and paths of relationally-embedded
network ties in emerging entrepreneurial firms. Entrepreneurship Theory and
Practice, 29, 113-144.
Hite, Julie M. (2008). The role of dyadic multi-dimensionality in the evolution of
strategic network ties. In J. A. C. Baum & T. J. Rowley (Eds.), Advances in
Strategic Management: Network Strategy (Vol. 25, pp. 133-170). Oxford UK: Emerald
Group Publishing Ltd.
Dr. Julie M. Hite
Brigham Young University
Educational Leadership & Foundations
801-422-5039 (voice mail comes to my email)
Julie_hite@byu.edu
Subject: Re: What is an old/dormant strong tie?
From: "John T. Maloney"
Date: Wed, August 14, 2013 17:24
Hi �
still a "strong tie"? Or does it become a "weak tie"
Thanks for the discourse on ties. Further idle August musings on ties.
The temporal notion of tie strength is very important. Latent ties are just
absent (potential) ties in a rationally-bounded network or �tie space.�
Latent is probably a better word than absent.
Social networks are dynamic, complex systems. Ties come and go. You could
probably make the case for �old absent ties� or �future latent ties� just as
much as �old strong ties.� There is probably a magic ratio of strong ties to
other ties for the circumstances of healthy, prosperous social networks.
It is pretty safe to say that too many strong ties leads to dysfunctional
networks, group think and other network pathologies. E.g. it�s common to
�refresh� corporate boards, management teams, and govt cabinets when there
are too many strong (stale) ties�
BTW, sorry, but according to my wife, one of my most important strong ties,
former girlfriends are positively not �old strong ties.� In fact, if I know
what is good for me, they are not weak or even latent ties. They are
permanently absent ties. And you better believe it! J
Strong ties evolve from propinquity. There is certainly an ebb and flow of
propinquity, particularly in the fast-moving world of industrial design of
consumer products like Apple� and Corporate America overall.
Jobs had propinquities w/Ive and a generation early w/Woz. They were strong
ties. However, autocracy and strong ties are not mutually exclusive.
Recall the Apple corporate-era, the Sculley Model, when its so-called
�strong-ties� produced bombs like Lisa and Newton.
This Apple �corporate� epoch was particularly indifferent to the unique
pathology of the fervent Apple community, and its concomitant
weak/absent/latent ties. It was utterly dismissive of Jobs� apparent messiah
complex disorder. (Steve Jobs use and advocacy of entheogens, to discover
the divine within, is legendary.)
During this time, Apple cratered, was lost in the wilderness, nearly gone,
until they remediated their commitment to weak/absent/latent ties vis-à-vis
the archetypical Apple zealot (customers) and once again gave their messiah
absolute control.
Anyway, the prevailing theory of the firm, transaction cost economics (TCE),
depends on �less-weak� ties inside the firm to lower costs. External ties
are weaker/absent, thus creating greater friction and costs. It�s the
overall rationale for the firm. IMO, it is ambitious to call corporate ties
to lower transaction costs �strong ties.�
Meanwhile, a new view of the firm, the knowledge-based view (KBV), depends
on far �less weak� (stronger) ties than TCE. Some of these may in fact be
conventional strong ties, but they are rare in large organizations.
KBV firms are required to exchange sticky or tacit knowledge like designs,
inventions and creativity. These complex interactions of �The Social
Enterprise� must be conducted quickly, with as little cost or friction as
possible.
http://colabria.com/2013/07/14/the-social-enterprise/
The KBV sees the firm simply as a unique collection of fluid, multi-valent,
high-velocity �stronger� ties that conspire to create unfair competitive
advantage.
-j
P.S. Useless ties: http://bit.ly/1cNBn1A
Subject: Re: What is an old/dormant strong tie?
From: "John McCreery"
Date: Wed, August 14, 2013 23:11
To: SOCNET@LISTS.UFL.EDU
Dan,
I just downloaded your dissertation and read the introduction. You have me
hooked. Both as an advertising copywriter and as a veteran of several
volunteer organizations mired in old divisions and quarrels, I responded
instantly to your title "Ghosts of Organization Past." The introduction has
drawn me in. Yes, yes, yes....I find my head nodding yes over and over
again as I read.
You point to another important aspect of real world social networks that
distinguish them from the mathematical models we construct based on random
graphs that mirror free market assumptions. I am tempted to call it
innocence. We chatter on about homophily and brokerage, for example,
without considering the social debris, old quarrels, grudges, unhealed
wounds, etc., that affect how ties are may or may not be activated and,
perhaps more important, how they will be activated—in warm support, chilly
alliance, bitter factionalism or vendetta. Not saying the models are
wrong. The math works really well. What it has to do with the messiness of
real life and real politics? That is an interesting question, indeed.
John
Subject: Re: What is an old/dormant strong tie?
From: "Ian McCulloh"
Date: Thu, August 15, 2013 02:26
To: SOCNET@LISTS.UFL.EDU
I am disappointed that Matt Brashears has not commented on the tie strength
issue. I like his theories on tie strength so much I published it in my
book.
Matt argues that there are several components to the strength of a tie, for
example intimacy and frequency of contact. He illustrates this with the
example of an old friend (high intimacy, low frequency), a spouse (high
intimacy, high frequency), a co-worker (low intimacy, high frequency), and
a stranger (low intimacy, low frequency). The tie strength is a diagonal
line where the spouse quadrant is the strong tie and the stranger quadrant
is a weak tie. Strong ties often require more effort and resources to
maintain and cultivate. A person utilizes different dimensions of strength
depending upon the nature of the need. For example, if you need to borrow
a few dollars for lunch, you won't ask an old friend in a different town.
You'd ask a co-worker. If you were faced with a personal tragedy, you may
not rely on a co-worker, but would call an old friend.
Look for Matt's paper when he publishes it.
As for John Maloney's point, I think the main idea was that there are
people using SNA for pretty important things (military targeting, criminal
investigations, etc) that don't really know what they are talking about.
Regardless of the specific point (tie strength, centrality, etc), I think
he is totally right. Perhaps we are at the point as an academic field,
where we should consider certifications. Should there be some kind of
certification process for practitioners? We require certification for
teachers, lawyers, medical doctors, even hair stylists. If people in
government are going to use SNA for military targeting and criminal
investigation, I think it would be a good idea to certify them to make sure
they know what they are doing.
ian
Subject: Re: What is an old/dormant strong tie?
From: "John McCreery"
Date: Thu, August 15, 2013 02:35
To: SOCNET@LISTS.UFL.EDU
Ian,
I look forward to reading Matt Brashears' work. In your description of it,
I still detect, however, that innocence I mentioned in earlier posts. The
relations envisioned are friends, spouses, strangers. What about rivals,
enemies, estranged spouses, strangers regarded as potentially or inherently
hostile? I recall Georg Simmel's observation that hostility is also a
social relationship and that only indifference is truly asocial. We can
capture that in signed networks (+, –, 0). Bashears appears to distinguish
different types of positive relationships, but what about the negative
ones? There is certainly some difference between, for example, mild
distaste and murderous rage.
John
Subject: Re: What is an old/dormant strong tie?
From: "Matthew Edward Brashears"
Date: Thu, August 15, 2013 02:40
To: SOCNET@LISTS.UFL.EDU
Hi All,
I appreciate Ian’s kind words, but my paper on the subject is still in development
(owing mostly to some simulations that are taking awhile to complete). As he
suggests, I have my doubts about tie strength and think we’d do better to use a
different system that captures the latency of tie usage (i.e., average time between
activations), the capacity of the tie (i.e., how much stuff you can cram down it),
and its redundancy. My conception is closely tied to Aral and Van Alsytne’s (2011)
work, but I don’t completely agree with them. Those of you at the Hamburg Sunbelt
may have seen me discuss this in my session. In any event, the paper is still rough,
but if you email me I’m willing to give advance looks at everything but the
simulation results.
But definitely look for the complete publication once it’s out! ;)
-Matt
Subject: Re: What is an old/dormant strong tie?
From: "Loet Leydesdorff"
Date: Thu, August 15, 2013 08:15
To: SOCNET@LISTS.UFL.EDU
Dear John,
Why would there only be two dimensions? (+, -, 0). It seems to me that this
design is also factor-analytical (positional). However, the factor-analysis
would provide dimensions to the nodes (variables) and not to the
interactions. Perhaps, one can use factor-scores for this.
The positional analysis assumes a vector-space that is constructed on the
basis of the correlations (= distributions of relations). One can normalize
the adjacency matrix using the cosine and then the network derived from that
matrix represents the vector space. The vector space has a different
topology from the network graph.
Best,
Loet
Subject: Re: What is an old/dormant strong tie?
From: "John McCreery"
Date: Thu, August 15, 2013 14:14
To: SOCNET@LISTS.UFL.EDU
Dear Loet,
I do not, for a moment, believe that relationships have only two
dimensions. Most of mine, at least, are far more complicated than that. The
(+1, 0, -1) coding is nothing more than a pointer to Simmel's classic
observation that relationships can be positive, negative, or non-existent,
i.e., friendly, hostile, or asocial, i.e., indifferent. Still, however, I
retain my impression that much of the social network analysis to which I
have been exposed displays that innocence I mentioned.
Were I asked to explain why, I would hazard the speculation that network
studies are fundamentally biased toward emphasizing the positive. Your work
on citation networks and mine on teams that win ad contests share this
bias: our data are derived from groups whose members have not only worked
together, they have worked together successfully, thus their publications
and awards. But what of the many whose projects have failed? I have seen
numerous studies that begin with such questions as "Who are your friends?"
"Would would you ask for advice?" "Who would you ask for help?" Though it
may be my own ignorance, I cannot recall one in which the relevant
questions were "Who are your enemies?" "Whose advice would you avoid like
the plague?" or "Who would you never, ever turn to for help?" Yet these
sorts of relationships are also part of social and organizational life.
But I freely admit I know very little. Nothing would please me more than a
rush of citations proving me wrong.
John
Subject: Negative Ties ...
From: "Edmund Chattoe-Brown"
Date: Thu, August 15, 2013 16:30
To: SOCNET@LISTS.UFL.EDU
Dear All,
Although I am far from expert, it also struck me, reading:
Evans, K. M. (1962) Sociometry and Education, The International Library
of Sociology and Social Reconstruction (London: Routledge and Kegan
Paul).
that the routine collection of negative tie information seems to have
declined. For school children, I can see that there would be ethical
issues. (What do you do if you discover everyone hates Johnson?)
It is hard to see that these negative ties would not have significant
effects in real networks and thus that models not dealing with them
would be in some way inadequate.
All the best,
Edmund
Edmund Chattoe-Brown (Department of Sociology, University of Leicester, UK)
edmundchattoebrown@fastmail.fm
http://www.fastmail.fm - A no graphics, no pop-ups email service
Subject: Re: Negative Ties ...
From: "Joe Labianca"
Date: Thu, August 15, 2013 17:08
To: SOCNET@LISTS.UFL.EDU
Amen on the need for more negative tie research. This article lays out why:
Labianca, G., & Brass, D.J. (2006). “Exploring the Social Ledger: Negative
Relationships and Negative Asymmetry in Social Networks in
Organizations.” *Academy
of Management Review*, 31: 596-614.
Here are a few examples from organizational and political research:
Venkataramani, V., Labianca, G, & Grosser, T.J. (forthcoming). “Positive
and Negative Workplace Relationships, Social Satisfaction, and
Organizational Attachment.” *Journal of Applied Psychology*.
Smith, J.M, Lopez-Kidwell, V., Halgin, D.S., Labianca, G., Brass, D.J. &
Borgatti, S.P. (forthcoming). “Power in Politically Charged Networks.” *Social
Networks*, special issue on Political Networks.
Surprisingly, there is a lot going on with negative ties in schools (e.g.,
bullying, disliking). A recent paper is an example of this work:
Huitsing, G., van Duijn, M.A.J., Snijders, T.A.B., Wang, P., Sainiod, M.,
Salmivalli, C., & Veenstra, R. (in press). Univariate and multivariate
models of positive and negative networks: Liking, disliking, and
bully–victim relationships. *Social Networks*.
If done with the proper considerations for confidentiality, there is
nothing inherently unethical about studying negative ties.
Anyone interested in learning more about what's happening in negative tie
research can also join the LinkedIn group "Negative Ties and Networks".
Best,
Joe
Giuseppe (Joe) Labianca
Gatton Endowed Professor of Management
Gatton College of Business & Economics
LINKS Center for Social Network Analysis
University of Kentucky
Lexington, KY 40506
859-257-3741 (office)
404-428-4878 (mobile)
http://linkscenter.org/
tinyurl.com/JoeLabianca
Subject: Re: Negative Ties ...
From: "Bruggeman, J.P."
Date: Thu, August 15, 2013 17:33
To: SOCNET@LISTS.UFL.EDU
...and if you forget the negative ties:
Jeroen Bruggeman, V. A. Traag, and Justus Uitermark (2012) Detecting Communities
through Network Data. American Sociological Review December 77: 1050-1063.
http://asr.sagepub.com/content/77/6/1050.full.pdf+html
Best,
Jeroen.
Subject: Re: Negative Ties ...
From: "Rich Persaud"
Date: Thu, August 15, 2013 17:24
To: SOCNET@LISTS.UFL.EDU
These may be relevant..
Rich
> Ilany, Barocas, Koren, Kam, Geffen (2013) Structural balance in the social
networks of a wild mammal. Animal Behaviour.
> http://www.nimbios.org/press/FS_hyrax
>
> Srinivasan (2011), Local balancing influences global structure in social networks
> http://www.pnas.org/content/108/5/1751.full
>
> The friend of my enemy is my enemy: Virtual universe study of structural balance
> Szell, Lambiotte, Thurner (2010). Multirelational organization of large-scale
social networks in an online world. Proceedings of the National Academy of
Sciences
> http://www.sciencedaily.com/releases/2010/07/100719162510.htm
> http://www.pnas.org/content/107/31/13636.full.pdf
>
> Leskovec, Huttenlocher, Kleinberg (2010), Predicting Positive and Negative Links
in Online Social Media
> http://www.cs.cornell.edu/home/kleinber/www10-signed.pdf
> http://www.bcs.org/upload/pdf/jkleinberg-slides.pdf
>
> Overview: Social Balance on Networks: The Dynamics of Friendship and Hatred
> http://www.r it.edu/~w-cmmc/research/Tibor_Antal.pdf
>
> Brandes, Lerner (2008), Visualization of Conflict Networks
> http://www.inf.uni-konstanz.de/algo/publications/bl-vcn-08.pdf
>
Subject: Malignant Ties
From: "John T. Maloney"
Date: Thu, August 15, 2013 17:57
To: SOCNET@LISTS.UFL.EDU
Hi ?
Re: "Ghosts of Organization Past"
Link vestiges and network provenance are essential properties and critical
enablers of Hastily Formed Networks ?
http://colabria.com/2013/08/15/hastily-formed-networks/
Perpetual drills and well-conceived simulations nurture latent links that
are key to creating positive outcomes vis-à-vis Hastily Formed Networks.
Concerning malignant ties, please consider some observations.
Chronic (old) strong-tie dysfunction is broadly manifest in central
government agencies and unions. Remember, the bureaucratic mission is to
preserve the bureaucracy, by definition. These rancid, unelected
organizations have impenetrable strong-tie patterns. They are a cancer on
civil society.
Criticizing strong-tie central govt bureaucracy for not being responsive or
competent is ridiculous. Service not their objective. It makes for good
political theater though. It?s why 9 out of 10 Federal employees summarily
reject the ACA (ObamaCare). No surprise that Federal employees pursue
enlightened self-interest in the ?real-world? for themselves. They
vehemently reject their own dog food!
When criticized, the conditioned motor-response of these intransigent
strong-tie networks is ?more money.? Recall, strong-ties take enormous
effort and resources to maintain. More money for entrenched bureaucrats and
strong-tie networks makes them even more insular.
Take the loathsome teachers unions for example. They have led an
unprecedented 40-year decay of K-12 education. Even with the highest level
of per-student funding in the world, these feculent strong-tie networks
cannot produce competitive students.
Strong-tie network patterns can produce great things like breakthrough
science, prosperous families and flourishing communities. They also produce
the disease Eisenhower warned of in 1961, the military-industrial complex
(MIC). It is now a monolithic and sinister Iron Triangle ?
http://en.wikipedia.org/wiki/Iron_triangle_(US_politics)
Note the treacherous ?Love Triangle? of strong-ties too ?
http://upload.wikimedia.org/wikipedia/en/2/29/The_Leveson_Love_Triangle_Leve
son-2.jpg
-j
Subject: Re: What is an old/dormant strong tie?
From: "Ian McCulloh"
Date: Thu, August 15, 2013 21:51
To: SOCNET@LISTS.UFL.EDU
I think there is a subtle difference between tie strength and the idea of positive,
negative, null ties. I don't think Matt claims there are only two aspects of tie
strength, but rather more than one. While this may seem obvious, it has not been
formalized. I think this is central to whether a strong tie is useful or not,
because someone must ask what is meant by strong and what is the purpose of the tie
under what context.
Ian McCulloh
Subject: Re: What is an old/dormant strong tie?
From: "Loet Leydesdorff"
Date: Fri, August 16, 2013 08:13
To: SOCNET@LISTS.UFL.EDU
Dear John,
It seems to me that there are two themes:
1. the bias in the data collection: one cannot ask so easily for latent
relations; but psychologists may have developed smart strategies for this
such as scales;
2. the data-analysis in SNA is focused on relations (graph-analysis) whereas
relations may mean different things in terms of latent dimensions in the
vector space.
Both issues point to the limits of using SNA other than as a methodology.
Best,
Loet
Subject: Re: What is an old/dormant strong tie?
From: "Ian McCulloh"
Date: Fri, August 16, 2013 13:02
To: SOCNET@LISTS.UFL.EDU
I'm not sure the definition of SNA is that tight. There is a methodology, using
mainly visualizations, math, and some other stuff. There is also theory, such as
social circles, social forces (homophily, proximity, etc), and tons of stuff. I
would argue that is SNA too.
It is the development of social theory that should precede the development of
mathematical techniques to test them.
Also, do not forget that many of the techniques developed in this manner find
application in other non-social domains.
Ian McCulloh
Subject: Re: Negative Ties ...
From: "John P. Boyd"
Date: Fri, August 16, 2013 20:26
To: SOCNET@LISTS.UFL.EDU
I can’t help but point out that the only cases that really fit the theory of
structural balance were known in anthropology as “moiety systems”: the tribe was
divided into two halves, with the rule that people had to marry into the other half
and that children belonged to the same half as their father in “patrilineal” moiety
systems (or as their mother in the “matrilineal” kind). This implies that marriage
is a negative relation. (No jokes allowed here.)
This brings up the question of how do you know if a relation R is positive or
negative? One way is to look at its psychological properties, but another way is by
its algebraic properties: positive ties are idempotent R^2=R, while negative ties
are involutions: R^2≠R, but R^3=R.
Subject: Re: What is an old/dormant strong tie?
From: "Michel Grossetti"
Date: Fri, August 16, 2013 21:25
To: SOCNET@LISTS.UFL.EDU
Hello,
For those who can read French, Claire Bidart, Alain Degenne and me
gathered in a book the results of several surveys of personal networks,
including a longitudinal study. One chapter focuses on ties creation,
another of their changing over time, another on
their disappearance, etc. We also
conclude that dormant ties are certainly an
important issue, not enough studied.
The book :
Claire Bidart, Alain Degenne, Michel Grossetti, 2011, La vie en réseau.
Dynamique des relations sociales, Presses Universitaires de France,
Collection « Le lien social ».
Michel Grossetti
Subject: Re: What is an old/dormant strong tie?
From: "John McCreery"
Date: Sat, August 17, 2013 01:25
To: SOCNET@LISTS.UFL.EDU
Ian,
I agree. SNA is a mixture of theory and methodology. To me the methodology
provides the social microscope I mentioned (Vladimir Bategelj says I ought
to say "telescope"). I t is the theory where I see the innocence I talk
about. I recall seeing references to the old maxim: "The enemy of my enemy
is my friend," but I don't see much done with this idea.
I wonder what might be done with, for example, segmentary opposition, a
classic notion from British social anthropology and also said to be a
common feature of societies throughout North Africa and the Middle East.
The idea is a simple one. Every individual belongs to a nested series of
groups. Let's call them for sake of illustration family
Subject: Re: Negative Ties ...
From: "Taha Yasseri"
Date: Mon, August 19, 2013 15:01
To: SOCNET@LISTS.UFL.EDU
In this work (Section 4.2.) we study anti-collaboration network of
Wikipedia editors (based on "reverts" in contrast to "co-authorship"):
Value Production in a Collaborative Environment, Taha Yasseri, János
Kertész Journal of Statistical Physics, 2013, 151(3-4), pp 414-439.
I thought it might be helpful.
bests,
Taha
===== Resources =====
Subject: Re: "Resources" in SNA
From: "David Shoham"
Date: Thu, September 12, 2013 17:23
To: SOCNET@LISTS.UFL.EDU
Abigail,
It's a more general definition, but Coleman offers the following which is very
useful for the SNA conceptualization of resources: "The resources each actor has
which are of interest to others include a wide variety of things. The most obvious
of these are what economists call private goods... But private divisible goods are
only one of several kinds of things over which actors have control and in which they
are interested... [T]here are several properties that distinguish types of
resources... These properties are divisibility, alienability, conservation, time of
delivery, and absence of externalities." (Foundations of Social Theory, pp.33-34)
One of Coleman's examples is information, which does not exhibit conservation. Eg,
if I pass along a piece of information, the information remains with me. This is not
true for private economic goods: if I pass along a piece of fruit, I no longer
possess it.
David
****************************
David Shoham, PhD MSPH
Dept. Public Health Sciences
Phone: 708-327-9006
Email: dshoham@lumc.edu
****************************
>>> Abigail McHugh-Grifa 9/12/2013 9:37 AM >>>
***** To join INSNA, visit http://www.insna.org *****
Dear SOCNET community,
I am a Ph.D. candidate in Music Education at the Eastman School of Music in
Rochester, NY. I am working on my dissertation, which uses a qualitative approach
to SNA. However, there is one piece of information that I am having a surprisingly
difficult time finding, and I was wondering if you might be able to help me: How is
the term resource typically defined in SNA research? Is there some article or book
that you are aware of that explicitly defines what resources are from a SNA
perspective? Any guidance you might provide would be greatly appreciated.Thank you
for your time,
Abigail McHugh-GrifaEmail: abigailmchugh@hotmail.com
Blog: happilyeducated.com
Subject: Re: "Resources" in SNA
From: "Malick Faye"
Date: Fri, September 13, 2013 09:41
To: SOCNET@LISTS.UFL.EDU
Priority: Normal
Options: View Full Header | View Printable Version | Download this as a file | Whitelist Sender
***** To join INSNA, visit http://www.insna.org *****
Abigail,
I don't know if you understand French, but you will find seven
categories of ressources in Lemieux (1999:17-18): ressources
matérielles, ressources normatives, ressources statutaires, ressources
actionneuses, ressources humaines, ressources informationnelles et
ressources relationnelles.
He notes, further, as David wrote, three types of ressources with regard
to their transmission: deperditive, semi-deperditive, and non déperditive.
Hope it helps
Malick
===== Pathfinder =====
5. nov 2013
Can it be combined with
ftp://ftp.etsii.ull.es/pub/PRGCOM/trabajos/caminos_min/fast_shortest_path.pdf
?
===== Networks and maps =====
Subject Re: Network and Geographic maps
From Tracy Van HoltAdd contact
Sender Social Networks Discussion ForumAdd contact
To SOCNET@LISTS.UFL.EDUAdd contact
Date Wed 14:47
Hi Michelle,
ArcGIS is widely used in spatial analyses, but you probably can’t just overlay these datasets and you may need transform these data. Also, ORA has a geospatial component and may get you what you want—it may be more user friendly. Do you have the city map layers? You will need to make sure that all of your data points are in the same datum and projection and these data are entered in a way that the software can recognize in ArcGIS.
Viewing your data points on a map is the first step, but all gets very exciting when you can actually run spatial analyses and integrate landscape-level information with your network data. For that, you need ArcGIS or some other spatial analysis software. You can contact me if you want to discuss more.
—Tracy Van Holt
East Carolina University-Geography
Swedish Royal Academy of Sciences-Visiting Scientist
On Mar 19, 2014, at 2:27 PM, Michelle Anne Birkett wrote:
> ***** To join INSNA, visit http://www.insna.org *****
> Hi Folks,
>
> Is anyone familiar with combining network plots with some sort of GIS or geographic mapping software? I’m working with neighborhood by neighborhood connections in Gephi and would love to visualize this on an actual City map. Any resources you can point me to would be appreciated. Software outside of Gephi is fine too.
>
> Thanks!
> Michelle
>
>
> Michelle Birkett, PhD
> Research Assistant Professor
> Department of Medical Social Sciences
> Northwestern University Feinberg School of Medicine
> 625 N. Michigan Avenue, 27th Floor
> Phone: 312.503.3538; Fax: 312.503.4800
> Email: birkett@northwestern.edu
> www.feinberg.northwestern.edu
Subject Re: Network and Geographic maps
From Loet LeydesdorffAdd contact
Sender Social Networks Discussion ForumAdd contact
To SOCNET@LISTS.UFL.EDUAdd contact
Date Wed 14:49
Dear Michelle,
You may find the interface between Pajek and Google maps at http://www.leydesdorff.net/gmaps/index.htm useful.
Within a Pajek file one can also use the geographical coordinates for positioning the nodes. It is a bit of puzzling. One can read the Pajek files thereafter into other visualization tools (e.g., VOSviewer) and keep the coordinates.
Best,
Loet
Loet Leydesdorff
Professor Emeritus, University of Amsterdam
Amsterdam School of Communications Research (ASCoR)
Subject Re: Network and Geographic maps
From Alla G. KhadkaAdd contact
Sender Social Networks Discussion ForumAdd contact
To SOCNET@LISTS.UFL.EDUAdd contact
Reply-To Alla G. KhadkaAdd contact
Date Wed 15:09
Hi Michelle,
I use ORA (already mentioned by Tracy). All you need is longitude latitude info, and ORA
puts your locations on the map and connects them (if your data is relational). It also
allows coloring your nodes by attribute values. The advantage of ORA is that first you
can do social network analysis of your network and then visualize it on a map. Feel free
to e-mail me if you need more info.
Alla
Subject Re: Network and Geographic maps
From Kathleen CarleyAdd contact
Sender Social Networks Discussion ForumAdd contact
To SOCNET@LISTS.UFL.EDUAdd contact
Date Wed 15:10
ORA also will export to ARCGis and will link to googlemaps
Subject AW: [SOCNET] Network and Geographic maps
From Dr. Murat ÜnalAdd contact
Sender Social Networks Discussion ForumAdd contact
To SOCNET@LISTS.UFL.EDUAdd contact
Date Wed 15:11
Dear Michelle
Thanks for your e-mail. You can generally use ORA- Net Scenes developed by people from Carnegie Mellon where you have the ability to create geospatial networks (using latitudes and longitudes). It is also linked to Nasa World Wind technology and could potentially be the solution you look for as ORA is quite a comprehensive solution.
All the best Murat
Subject Re: Network and Geographic maps
From Craig AllenAdd contact
Sender Social Networks Discussion ForumAdd contact
To SOCNET@LISTS.UFL.EDUAdd contact
Date Wed 15:15
Michelle,
I use R which has some really easy-to-use mapping tools. In particular the packages "maps" and "maptools" should be a good start. If you want to incorporate street-level background maps, then I suggest the packages "OpenStreetMap" together with "ggmap" - once you learn these tools you can pull in address lists.
In fact at http://stackoverflow.com/questions/3257441/geocoding-in-r-with-google-maps
there is a nice little discussion of how to extract the latitude and longitude from any address in the world using the API at maps.google.com. You can only make 2,500 requests per day, but with a bit of planning, you can work through an entire data set fairly quickly.
- Craig Allen
Subject Re: Network and Geographic maps
From Jan RieblingAdd contact
Sender Social Networks Discussion ForumAdd contact
To SOCNET@LISTS.UFL.EDUAdd contact
Date Wed 15:16
Dear Michelle,
we had much success in combining the Networkx and Basemap packages in Python. This may not be the easiest solution but it allow you to tackle the problem within one framework. You can look at our code here: http://www.sociology-hacks.org/?p=67
Kind regards,
Jan Riebling
Subject Re: Network and Geographic maps
From Sean EvertonAdd contact
Sender Social Networks Discussion ForumAdd contact
To SOCNET@LISTS.UFL.EDUAdd contact
Date Wed 15:29
Michelle:
ORA works quite well in this regard. It can export to Google Maps and ArcGIS. I discuss how to do so in Chapter 10 of my book (pp. 333-341).
Sean
Sean F. Everton, PhD
sfeverton18@gmail.com
http://godpoliticsbaseball.blogspot.com/
Disrupting Dark Networks (book)
https://sites.google.com/site/sfeverton18/research (book companion site)