LISTSERV mailing list manager LISTSERV 16.0

Help for SOCNET Archives


SOCNET Archives

SOCNET Archives


SOCNET@LISTS.UFL.EDU


View:

Message:

[

First

|

Previous

|

Next

|

Last

]

By Topic:

[

First

|

Previous

|

Next

|

Last

]

By Author:

[

First

|

Previous

|

Next

|

Last

]

Font:

Proportional Font

LISTSERV Archives

LISTSERV Archives

SOCNET Home

SOCNET Home

SOCNET  February 2004

SOCNET February 2004

Subject:

reply: missing data problem

From:

Anabel Quan-Haase <[log in to unmask]>

Reply-To:

Anabel Quan-Haase <[log in to unmask]>

Date:

Wed, 11 Feb 2004 09:25:45 -0500

Content-Type:

text/plain

Parts/Attachments:

Parts/Attachments

text/plain (130 lines)

*****  To join INSNA, visit http://www.sfu.ca/~insna/  *****

Dear Orly Tenne Gazit,
I have put together some replies that I received to an earlier post
addressing the issue of how to deal with missing data. I hope this is
helpful.
Cheers,
Anabel

The University of Western Ontario
Information and Media Studies           Department of Sociology

                                Responses
---------------------------------------------------------------------
Arent Greve, dr.oecon., professor
The Norwegian School of Economics and Business Administration
Department of strategy and Management
Breiviksveien 40
5045  Bergen
Burt & Ronchi (1994) discuss using other people as respondents on behalf
of respondents that they chose not to talk to, when they capture a large
network.

Burt, Ronald S. and Ronchi, Don. 1994. Measuring a Large network
quickly. Social Networks. Vol. 16: 91-135.
Also see:
Greve, A. & Salaff, J W. Research in the Sociology of Organizations:
Social Capital of Organizations. Vol. 18: 107-134, 2001. Editors: Shaul
M. Gabbay and Roger Th. A.J. Leenders.
------------------------------------------------------------------------
Thomas W. Valente, PhD
Director, Master of Public Health Program
http://www.usc.edu/hsc/medicine/preventive_med/ipr/mph/
Department of  Preventive Medicine
School of Medicine
University of Southern California
1000 Fremont Ave.
Building A Room 5133
Alhambra CA 91803

We have a paper forthcoming in Social Networks on this topic.  We find
that, in general,centrality measure accuracy decrease linearly with
nodes missing at
random.  Some measures do better than others.  In degree, for example is
quite
stable retaining a high correlation with the full sample when as much as
50% of
the data (respondents) are missing at random.  A simple eigenvector, as a
measure of centrality, seems to do quite well.  The more a centrality
measured
tapped the structure of the network, the more unstable it became. We
assessed 11
centrality measures on 59 social networks measured in 7 different studies

Costenbader, E. & Valente, T. W. (in press).  The stability of
centrality when
networks are sampled.  Social Networks.


------------------------------------------------------------------
Ravi Jasuja
Although you can perform some interesting simulations regarding the
robustness of your outcomes (compare your results when you fill in
averages or percieved numbers using information from other responses).

You may find this paper interesting :
http://www.psych.unimelb.edu.au/staff/gr/missing5.pdf
-----------------------------------------------------------------
Kathleen Carley:
Carnegie Mellon University
Pittsburgh, PA  15213-3068

we have been trying to work this problem
we find that a combination approach works best
1) make limited changes based on known features of the relationship
2) where there is not enough data of confidence is low use an average or
leave the relationship as 0
in general - anything you do will impact standard social network
measures in a non-linear fashion


---------------------------------------------------------------------
Ju-Sung Lee
Ph.D. Candidate
Social and Decision Sciences
Carnegie Mellon University
Pittsburgh, PA  15213-3068

In the non-network statistics world, a procedure called 'multiple
imputation' (or MI) is gaining acceptance.  If your missing responses
are non-sociometric, then MI may be appropriate.  The idea is that you
can infer a distribution for your missing responses from the data you
have available.  For instance, if all males below the age of 20 answered
a response with a tight standard deviation, you can infer that other
males below the age of 20, who haven't answered, will have their
responses lie in the same distribution.  Of course, the reason for
non-response is important as well.  Many imputation procedures prefer
this reason to be essentially random and not dependent on the data or
the other correlates.

Once the imputation process figures out a distribution for each of the
variables with missing responses, it creates *multiple* full data sets
that include random values picked from these informed distributions.  So
in each of these newly generated data sets, your missing responses will
have different random values, that come from their own distributions.
Then you would run your regression (or other statistical procedure), one
for each of these data sets, and take your beta coefficients, get their
means, and determine how significant they are based on not just each
data set (with missing values filled-in) but also on the distribution of
betas.  The MI software takes care of calculating all this.
Some authors associated with MI are Schafer, Gary King, and Ed Rubin.
Some of their websites have free software.  MI has also been
incorporated into SAS as of version 8.

If your missing data is sociometric, you could think about using the MI
approach, but some changes would have to be made; I'm not yet familiar
with the network version of imputation.

Hope this helps.  If my memory is correct, you attended Sunbelt in
Charleston, SC; but, I don't think we met.

Juice

*****************************

_____________________________________________________________________
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.

Top of Message | Previous Page | Permalink

Advanced Options


Options

Log In

Log In

Get Password

Get Password


Search Archives

Search Archives


Subscribe or Unsubscribe

Subscribe or Unsubscribe


Archives

June 2021
May 2021
April 2021
March 2021
February 2021
January 2021
December 2020
November 2020
October 2020
September 2020
August 2020
July 2020
June 2020
May 2020
April 2020
March 2020
February 2020
January 2020
December 2019
November 2019
October 2019
September 2019
August 2019
July 2019
June 2019
May 2019
April 2019
March 2019
February 2019
January 2019
December 2018
November 2018
October 2018
September 2018
August 2018
July 2018
June 2018
May 2018
April 2018
March 2018
February 2018
January 2018
December 2017
November 2017
October 2017
September 2017
August 2017
July 2017
June 2017
May 2017
April 2017
March 2017
February 2017
January 2017
December 2016
November 2016
October 2016
September 2016
August 2016
July 2016
June 2016
May 2016
April 2016
March 2016
February 2016
January 2016
December 2015
November 2015
October 2015
September 2015
August 2015
July 2015
June 2015
May 2015
April 2015
March 2015
February 2015
January 2015
December 2014
November 2014
October 2014
September 2014
August 2014
July 2014
June 2014
May 2014
April 2014
March 2014
February 2014
January 2014
December 2013
November 2013
October 2013
September 2013
August 2013
July 2013
June 2013
May 2013
April 2013
March 2013
February 2013
January 2013
December 2012
November 2012
October 2012
September 2012
August 2012
July 2012
June 2012
May 2012
April 2012
March 2012
February 2012
January 2012
December 2011
November 2011
October 2011
September 2011
August 2011
July 2011
June 2011
May 2011
April 2011
March 2011
February 2011
January 2011
December 2010
November 2010
October 2010
September 2010
August 2010
July 2010
June 2010
May 2010
April 2010
March 2010
February 2010
January 2010
December 2009
November 2009
October 2009
September 2009
August 2009
July 2009
June 2009
May 2009
April 2009
March 2009
February 2009
January 2009
December 2008
November 2008
October 2008
September 2008
August 2008
July 2008, Week 62
July 2008
June 2008
May 2008
April 2008
March 2008
February 2008
January 2008
December 2007
November 2007
October 2007
September 2007
August 2007
July 2007
June 2007
May 2007
April 2007
March 2007
February 2007
January 2007
December 2006
November 2006
October 2006
September 2006
August 2006
July 2006
June 2006
May 2006
April 2006
March 2006
February 2006
January 2006
December 2005
November 2005
October 2005
September 2005
August 2005
July 2005
June 2005
May 2005
April 2005
March 2005
February 2005
January 2005
December 2004
November 2004
October 2004
September 2004
August 2004
July 2004
June 2004
May 2004
April 2004
March 2004
February 2004
January 2004
December 2003
November 2003
October 2003
September 2003
August 2003
July 2003
June 2003
May 2003
April 2003
March 2003
February 2003
January 2003
December 2002
November 2002
October 2002
September 2002
August 2002
July 2002
June 2002
May 2002
April 2002
March 2002
February 2002
January 2002
December 2001
November 2001
October 2001
September 2001
August 2001
July 2001
June 2001
May 2001

ATOM RSS1 RSS2



LISTS.UFL.EDU

CataList Email List Search Powered by the LISTSERV Email List Manager