Print

Print


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

Hi Thomas,

from what you write, I'd say that a Siena-analysis certainly makes 
sense. Whether overall measures change or not is not decisive here, what 
you need to check is to what degree the networks differ in terms of 
Hamming distance. If there is sufficient distance, there is sufficient 
statistical power to identify parameters.

For networks of size 600, it is very reasonable to include the group 
membership variables (and other covariates) as predictors of network 
evolution because of the homogeneity assumption underlying Siena's 
models: It is unlikely that an actor indeed considers all 599 other 
actors in every decision about changing his personal network, so 
inclusion of proxy variables for interaction opportunity (such as group 
membership) is advisable.

Greetings,
Christian


Thomas Plotkowiak schrieb:
> *****  To join INSNA, visit http://www.insna.org  *****
>
> Hi,
>
> I have a network that I want to investigate for structural effects of
> network dynamics. (Its around 600 nodes, snapshots of the network
> every day, 30days )
>
> - What I have seen from my longitudinal analysis from ORA is that the
> networks overall measures like closeness, betweenness, average
> distance have not changed much in this timeframe (betweenness from
> 0.08 to 0.085, closeness from 0.0008 to 0.0009 and so on...).
> - When investigating sub communities of the network I see more
> significant changes like for group A betweenness from 0.1 to 0.135, or
> group B in closeness from 0.02 to 0.05.
>
> My question is given that i have some additional covariates like
> membership of a group, gender, age etc.. does it even make sense to
> test for their significance in SIENA? I guess I can always test for
> transitive triplets or balance effects and so on, but my question is
> as a rule of thumb how much network dynamics is necessary to raise
> reasonable conclusions.
>
> Cheers and Thanks for your help
> Thomas
>
> _____________________________________________________________________
> 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.
>   


-- 
_ __ ___ ____ ___ __ _ __ ___ ____ ___ __ _

Christian Steglich, researcher
Faculty of Behavioural and Social Sciences
University of Groningen
Grote Rozenstraat 31
9712 TG Groningen
The Netherlands

fon +31-(0)50-363 6189
fax +31-(0)50-363 6226

http://www.gmw.rug.nl/~steglich/
_ __ ___ ____ ___ __ _ __ ___ ____ ___ __ _

_____________________________________________________________________
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.