*****  To join INSNA, visit  *****

> -----Original Message-----
> From: Social Networks Discussion Forum [mailto:[log in to unmask]]
> Behalf Of Cora Schaefer
> Sent: Saturday, September 30, 2006 4:25 AM
> To: [log in to unmask]
> Subject: AW: Validity of network ties
> *****  To join INSNA, visit  *****
> Dear all,
> thanks for the quick replies! I realize I should have been more
> Without having read the suggested literature yet (so please forgive me
> the issues raised are answered in there), I try to explain more
> what I meant.
> There's a lot of literature about the outcomes of networks, e.g. the
> classic
> example of finding a new job. So, by "real networks" I meant to pose
> questions if these conclusions could be assumed as well for networks
> constructed from online data.
> There two concerns with network data gathered from social network
> such
> as LinkedIn that occurred to me so far: first, I expect there to be
> network ties in this kind of online data than in questionnaire studies
> ties accumulate in social network sites. Probably very few people
> out" their ties in their profile. Yet, when asked I don't expect these
> very
> weak or maybe "old" ties to be mentioned.
> Second and this is mentioned by danah boyd, there are some people who
> collect ties as an end in itself. Besides the point of asking about
> motivation to do so, I wonder whether these persons can be compared to
> hubs
> who know (as in face-to-face knowing or through more extensive
> communication
> than the message asking for the tie to be confirmed) their alters.
> Regards,
> Cora Schaefer

I think that a lot of this LinkedIn phenomenon you are seeing can be
cast mainly as a flaw in the model that LinkedIn has adopted: the notion
that a tie is "binary", that is, it either exists or it does not, and
that the only thing that one needs to do is point and click an
invitation at someone easily found in the system to go from "we have
absolutely no tie" to "we have the same kind of tie as people I've known
all my life." There is no notion of a "cost" per tie, for instance, nor
the notion that one needs to have laid down *some* sort of groundwork
before a tie can become "operationalized" (that is, sufficiently
"strong" for it to be a source of traditional "networking"). The lack of
a feel of "fidelity" when you use their product (at least, this happens
to me) is, I think, a very good practical example of what happens when a
model is not a very good match to the reality that it is attempting to

I think there are better models for a LinkedIn company to adopt, models
that better match reality; of course I do, since my full disclosure is
that I work for a company that is attempting to do that very thing. But
even outside my employment-inspired interest, I think it is going to be
interesting to see how an attempt to more accurately model the
real-world social network within social networking software will impact
the user experience. Maybe it just doesn't matter: maybe social
networking software is just a tool that provides some services and
doesn't depend on really being a "model" of the real world at all. *I*
don't think so, but I want to see this all in side-by-side action just
for curiosity's sake. It's also going to be a neat testbed for SNA


Dr Andrew J. Cleary
Director of Algorithms R&D
Visible Path Corp.

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