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

Hi Ion,

I was thinking about similar questions and would like to know what you find
out!

I was at Steve Borgatti and Martin Everett's Sunbelt presentation,
Unpacking Burt's Constraint Measure, and you might find this abstract
interesting.

>
> "Ron Burt proposed a number of measures for ego networks in his classic
> book “Structural Holes”. In 1997 Steve Borgatti wrote a short article in
> connections called unpacking Burt’s redundancy measure. The redundancy
> measure is relatively easy to disentangle as it is expressed as a fraction.
> In this talk we attempt to do the same with constraint. Burt’s constraint
> measure is a lot more challenging as it is expressed as a sum and a squared
> term. Burt claims it consists of a number of factors that he deems to be
> important namely size, density and hierarchy. Here I show that if the graph
> is binary then we can re-write the constraint to explicitly see these
> contributions. This allows us to do a number of additional things.* 1
> Calculate max and min for constraint * 2 Re-weight how these terms
> contribute to the measure. 3 Approximate constraint when we know the
> degrees of the alters without knowing the exact structure. In addition it
> provides an opportunity to redefine constraint in a way that more evenly
> balances the three components."

I remember the max is at least contingent on the node's degree value, not a
set number that universally applies to any network.

Good luck!
Ruqin

> ***** To join INSNA, visit https://urldefense.proofpoint.com/v2/url?u=http-3A__www.insna.org&d=DwIFaQ&c=pZJPUDQ3SB9JplYbifm4nt2lEVG5pWx2KikqINpWlZM&r=uXI5O6HThk1ULkPyaT6h2Ws3RKNKSY__GQ4DuS9UHhs&m=ZJFqVtos5jCI7up9gDXXh4f8XIywULd1ObiZgWsbSnI&s=6rQr4xpMZHEFXth7029dhzVqh2RQ_n4y5_H7QRZ_eMU&e=
> *****
>
> Dear All
>
>
>
> When examining structural holes, it is possible to calculate the aggregate
> constraint of an ego. The value of the aggregate constraint is usually
> between 0 and 1, but it can be greater than 1.
>
>
>
> I think I read somewhere that the maximum value possible (greater than 1)
> has been proved. I have mislaid the reference. Might anyone be able to
> inform me of the reference? Of the maximum value?
>
>
>
> With thanks
>
>
>
> Ion Georgiou
>
>
> _____________________________________________________________________
> SOCNET is a service of INSNA, the professional association for social
> network researchers (https://urldefense.proofpoint.com/v2/url?u=http-3A__www.insna.org&d=DwIFaQ&c=pZJPUDQ3SB9JplYbifm4nt2lEVG5pWx2KikqINpWlZM&r=uXI5O6HThk1ULkPyaT6h2Ws3RKNKSY__GQ4DuS9UHhs&m=ZJFqVtos5jCI7up9gDXXh4f8XIywULd1ObiZgWsbSnI&s=6rQr4xpMZHEFXth7029dhzVqh2RQ_n4y5_H7QRZ_eMU&e=). To unsubscribe, send an email
> in the body of the message.

--
Ruqin Ren
ruqinren.wordpress.com/blog

_____________________________________________________________________
SOCNET is a service of INSNA, the professional association for social
network researchers (http://www.insna.org). To unsubscribe, send