|
***** To join INSNA, visit http://www.sfu.ca/~insna/ *****
Dear Ilan and others,
For multilevel analysis = hierarchical linear modeling, the standard errors of regression coefficients and variance parameters are not determined/restricted by the number of cases per cluster - in this case, number of alters per ego, but in the first place by the total number of clusters, here egos = respondents.
The limited number of 6 alters per ego implies only that the estimation of the "random coefficients" (by the posterior means = empirical Bayes estimates), that convey information about individual respondents, is unstable; but this usually is not a primary focus of analysis.
For those wishing an introduction to the use of multilevel analysis of ego networks, you could consult:
Van Duijn, M.A.J., Van Busschbach, J., and Snijders, T.A.B., Multilevel analysis of
personal networks as dependent variables, Social Networks, 21, 1999, 187-209.
Best wishes,
Tom
-----Original Message-----
From: Ilan Talmud <[log in to unmask]>
To: [log in to unmask]
Date: Mon, 8 Dec 2003 22:25:13 +0200
Subject: Problems of ego-tie cross-level analysis
***** To join INSNA, visit http://www.sfu.ca/~insna/ *****
Hi:
I use an ego-network survey, with a name generator in which each respondent
names up to six contacts. The data contains details on those contacts.
I wish to cross-level each cluster of embedded ties (as level1) with each
ego (level 2) in which they are nested.
Yet, I can't make a usual cross-level analysis, as I have only up to six
(individual level) ties per ego, standard errors deem unstable.
Any solution ? How can I analyze these two-levels, nested observations ?
_____________________________________________________________________
SOCNET is a service of INSNA, the professional association for social
network researchers (http://www.sfu.ca/~insna/). To unsubscribe, send
an email message to [log in to unmask] containing the line
UNSUBSCRIBE SOCNET in the body of the message.
Tom A.B. Snijders
ICS / Dept. of Sociology
University of Groningen
Grote Rozenstraat 31
9712 GC Groningen
tel. +31-(0)50-3636188 (6469)
http://stat.gamma.rug.nl/snijders/
_____________________________________________________________________
SOCNET is a service of INSNA, the professional association for social
network researchers (http://www.sfu.ca/~insna/). To unsubscribe, send
an email message to [log in to unmask] containing the line
UNSUBSCRIBE SOCNET in the body of the message.
|
