[igraph] Re:Mean and Variance of Triad Census

[Dai Shizuka]

Hi Gabor, 
Yes, sorry I should've been more specific about the reference. 
The method I am looking into was described here: 
Holland, PW and Leinhardt, S. (1978) An Omnibus test for social structure using triads. Socialogical Methods and Research. 7, 227-256.

It was also described quite thoroughly in Wasserman & Fausts' book "Social Network Analysis" Chapter 14

The general idea is to create a test statistic for deviations in frequencies of triadic configurations as a means to test specific hypotheses about the structure of directed graphs. I think this function was included in PAJEK and UCINET... perhaps as part of the output to triad censuses? Though I never used these programs, I've references to make me suspect this. I hope that helps.

Thanks for your hard work,
Dai

Message: 3
Date: Wed, 8 Dec 2010 13:13:01 +0100
From: G?bor Cs?rdi <address@hidden>
Subject: Re: [igraph] Mean and Variance of Triad Census
To: Help for igraph users <address@hidden>
Message-ID:
       <AANLkTikVtwb14umFaUMYkwTzB+9nzqU=-address@hidden>
Content-Type: text/plain; charset=ISO-8859-1

Hi Dai,

this is not included in igraph. Can you point us to the paper in which
this methodology was published?

Thanks,
Gabor

On Tue, Dec 7, 2010 at 1:59 AM, Dai Shizuka <address@hidden> wrote:
> Hey folks,
>
> Is there an easy way to calculate the variance of the expected values of
> components in a triad census? I am trying to conduct some stats on structure
> based on the Davis-Leinhardt-Holland approach, using expected frequencies of
> different triad configurations based on a U|MAN distribution. I have figured
> out how to calculate mean (expected) triad frequencies of random graphs, but
> I am not sure about the variance of these values in order to calculate the
> test statistic, tau. I believe these functions were standard in PAJEK and
> UCINET, but I'd love to do this in R.
>
> Thanks,
> Dai
> ----------
> Daizaburo Shizuka
> Postdoctoral Scholar
> Ecology & Evolution
> University of Chicago
> 1101 East 57th Street, Chicago, IL 60637
>
>
> _______________________________________________
> igraph-help mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/igraph-help
>
>

----------
Daizaburo Shizuka
Postdoctoral Scholar
Ecology & Evolution
University of Chicago
1101 East 57th Street, Chicago, IL 60637
https://sites.google.com/site/daishizuka/