***** To join INSNA, visit http://www.insna.org *****
Sorry for being a little late in responding to the thread, but this is an
interesting discussion and highlighted many of the issues that my colleagues
and I are trying to address by setting up the EUSN Working Group on SNA and
Research ethics following the EUSN conference earlier this year, so thank
you for raising this.
I work in the healthcare field and for the last 7/8 years or so my
co-researchers and I had come up with similar problems / queries to yourself
when it comes to applying for approval through hospital RECs. We have
written up on our experiences of conducting SNA in a number of healthcare
settings, and this issue of lack of understanding of SNA that Christian
raised earlier is very pertinent when it comes to clinical RECs not used to
seeing SNA applications.
For a dilemma such as yours about the use of names in your SNA surveys, it
would be very helpful to have a dialogue with your REC about their concerns
(usually it should be possible to have a chat with the executive officer of
the REC who could facilitate correspondence with the chair) and to try to
work out a constructive solution that would respect privacy/informed consent
concerns as well as to minimise missing data.
In my own SNA studies on multidisciplinary interactions in healthcare teams
(which sounds like that kind of research you're hoping to conduct too?), we
had been successful in persuading 1 REC to allow the use of names as it was
about a "hidden" population of collaborating clinicians where the condition
has no set team structure, and both the clinicians and the REC saw the value
of SNA for this study even though alter anonymity cannot be protected (but
this would be the case with any contact-tracing epidemiological study, so
clinicians can relate to this kind of studies if they are persuaded of their
In other SNA studies on healthcare teams, we conducted whole network
sampling (where the staff membership of teams are already known) with
gatekeeper permission, and only those who consented explicitly are included
in the sample. For the intearction data, instead of a named individual,
clinicians chose specific named disciplines instead - i.e. "position
generators" (cf Knoke and Yang) were used as opposed to "name generators"
for particular types of multidisciplinary team work. There may be downsides
to this from analytical points of view, and we had to make sure that our
conclusions were pitched at the level of inter-disciplinary interactions as
opposed to inter-personal interactions.
Usually as part of the study protocols the researcher should highlight the
difference between SNA survey and standard surveys - e.g. not judging its
validity by sample size but in its coverage of the target network members.
There are also presentational strategies you can do to anonymise /
de-identify as much as possible individuals from the network maps, and
releasing only group-based statistics rather than ego-based statistics, this
has been the case for a number of SNA studies in the healthcare as well as
the educational fields, and my colleague Filipa Ribiero (University of
Porto) has some resources put up on this on our website below.
I've learnt a lot from discussions on this thread, so thanks Rosita again
for raising this. You and other SOCNET members may be interested in the
below survey on SNA and Research Ethics Practices in the health and social
care field that my colleague Filipa and I are conducting:
More information about our study and the EUSN Working Group on SNA and
Research Ethics can be found here:
We have started this site only recently and have set up resources pages on
SNA and research ethics, and of ethical governance in the areas of health
and education in particular, that may be useful to SOCNET members. Any
comments and feedback and tips most welcome.
Health Policy and Management
Trinity College Dublin
SOCNET is a service of INSNA, the professional association for social
network researchers (http://www.insna.org). To unsubscribe, send
an email message to [log in to unmask] containing the line
UNSUBSCRIBE SOCNET in the body of the message.