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


This is a pretty thorny issue, and any sort of minimum is not 
going to be the same for all quantities!

For example, you mentioned "centrality", and what is reasonable is going 
to be different for different types of centralities.  (I believe that 
there is a recent article by T. Valente and coauthors--- if I recall 
correctly, and forgive me if I am messing up the reference---that 
discusses various centrality measures.  I want to say that it appeared in 
"Connections".  I'm sure there are several other relevant articles as 

Also, I'd in general be cautious about saying that there is _any_ 
minimum with which I'd be definitively happy.  My preference on such 
things is actually to see mathematically rigorous theorems that give guidance...


> Date:    Thu, 6 May 2010 16:52:16 -0400
> From:    Allan Clifton <[log in to unmask]>
> Subject: Minimum network size
> *****  To join INSNA, visit  *****
> Hello all- is there an accepted minimum/desirable network size required
> to get accurate measures of network variables like centrality, cliques,
> density, etc.? Although you could compute these measures in a network of
> 5 people, it seems like a less reliable value than you would get from a
> network of 100 (although that might just be my impression, generalizing
> from non-network statistical analyses). Any suggestions for literature
> on this topic would be appreciated.
> Many thanks,
> Allan Clifton

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.