***** To join INSNA, visit http://www.insna.org *****
I'm not sure if the problem is one of data collection (for which that
really is a pretty big number) or one of data analysis (for which that is
a big number but not quite so insurmountably big -- most stats packages
can manage a million observations reasonably well).
Either way, if what you're trying to estimate is the probability of a
3-party alliance getting formed, you might take a look at Gary King's work
(with Michael Tomz and Langche Zeng) on rare-events logistic regression:
The main substantive context that they talk about this in is about rare
singleton events or dyadic events (states engaging in wars and things like
that), but it seems like that'd be applicable to your setting as well.
They discuss methodologically better ways to analyze contributing factors
to these kinds of events and also how to do some of the "matched sample"
data collection to do such a study.
Hope that helps,
On Mon, 3 Sep 2007, Ralph Heidl wrote:
> ***** To join INSNA, visit http://www.insna.org *****
> I am working on a project that looks at the factors affecting partner
> selection in a multi-partner alliance. As a start I'd like to find a
> workable approach for the formation of triads.
> My data set comprises 3-party alliances formed over 10 years between 87
> firms. The conventional approach would call for the calculation of the
> corresponding risk set (all 3-party alliances that could have formed). Now
> this would mean 87*86*85*10/6 = a painfully large number.
> I have been looking at matched samples but I wanted to ask whether anybody
> out there has tackled a similar problem successfully. Any help is greatly
> Ralph A. Heidl
> Box 353200
> University of Washington
> Seattle, WA 98195
> office: 206 685 2748
> mobile: 206 412 3888
> email: [log in to unmask]
> 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.
Assistant Professor of Strategy
MIT Sloan School of Management
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.