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

To all data cleaners:

I would like to add that differing reports about the structure of a
social system should not always be regarded as inaccurate information.

In our Sunbelt 2004 presentation (with Patrick Kenis) we argued that
"Inconsistent is different from Inaccurate" in that inconsistent reports
point to personal differences that should be exploited in an analysis
rather than eliminated during data preparation.  

This certainly complicates things, but it might be more important to
know, e.g., who people think (or, for that matter, want you to think)
they share confidential information with rather than with whom they
actually do (on whatever scale).

There seem to be at least three different kinds of skewed data:

 incomplete (missing information)
 inaccurate (unreliable information)
 inconsistent (additional information)

The potential utility of inconsistent data is the main reason for
supporting unconfirmed ties in the visone software (


Prof. Dr. Ulrik Brandes
Department of Computer &      | +49 7531 88 4433 (phone)
Information Science, Box D 67 | +49 7531 88 3577 (fax)
University of Konstanz        | [log in to unmask]
78457 Konstanz, Germany       |

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.