​​Dear all,


Many thanks for your advice. Indeed, this is a requirement that comes up frequently and your comments below would be of help to other researchers as well.


All the best,

Rosica




From: Social Networks Discussion Forum <[log in to unmask]> on behalf of David Lockhart <[log in to unmask]>
Sent: 10 November 2014 22:47
To: [log in to unmask]
Subject: Re: [SOCNET] Ethical approval issue
 
***** To join INSNA, visit http://www.insna.org *****
Another approach to anonymization in presentation and publication is to develop a statistical model of the collected data and present a random network drawn from a distribution with the same relevant statistical properties as the collected data.  For examples of this see Faux Mesa High and Faux Magnolia High in the R package 'ergm'.  These are made to have several of the same properties as the friendship networks of rural western (Mesa) and large southern (Magnolia) high schools as collected by the AddHealth Study. The Magnolia high school data has the same distribution and levels of homophily for grade, gender, and race as well as similar shared partner distributions.  However, because the dataset does not represent any actual high school, but is just like a high school in these ways, it is not possible to de-anonymize the data via inside information.  The downside is that there may be relevant properties of the network that were not considered or were not correctly modeled. However, these downsides would in essence occur for any statistical analysis and the conclusions drawn therefrom.

On Mon, Nov 10, 2014 at 2:18 PM, Christian Steglich <[log in to unmask]> wrote:
***** To join INSNA, visit http://www.insna.org *****
The ethical review requirement does indeed come up frequently, and rightly so. Sociocentric network data are extremely sensitive and can be easily de-anonymised with just a little bit of insider info.

The data reduction procedures Brooke mentioned are all solving the ethical problem, but they create a non-negligible missing data problem that medically trained IRBs are typically not aware of because they are used to survey samples. In sociocentric network studies, we often cannot afford even moderate missing data levels. They can lead to severe biases in analytical results, which also has its ethical aspects - you wouldn't want to spend public money on inferior data and publish inferior, maybe misleading statistical results.

It's a dilemma of ethics versus ethics, in the end. In my view, the conclusion is that there is no alternative to high-investment data collection. You need to convince each single respondent (or caretaker of the respondent) of the study you do, and make him/her participate. This is effortful, but research not worth this effort probably shouldn't be done anyway.

Good luck,
Christian





On 10-Nov-14 19:32, Brooke Foucault Welles wrote:
***** To join INSNA, visit http://www.insna.org *****
This question comes up frequently in social network research - who "owns" the relationship, the (consented) respondent, the (non-consented) alter or both? It is pretty common to simply provide the list, and then remove non-consented alters after the fact and/or anonymize any results, as others have suggested.  

If you have ethical concerns about providing the list at all, and you have a good way to communicate with all the potential participants, one way I have dealt with similar issues in the past is to offer an "opt-out" period for individuals to request that their name be removed from the survey instrument before it is ever distributed. I have only used this in cases where being identified on the list would not pose any risk other than what would normally be encountered anyway. Presumably, if you have a relatively small group of adults working in a non-stigmatized profession, minimal risk would apply. This gives people the option not to be included in the research, while retaining the possibility of fixed choice if that's important for your study design. 

Best, 
Brooke

On Mon, Nov 10, 2014 at 2:22 AM, Pachilova, Rosica <[log in to unmask]> wrote:
***** To join INSNA, visit http://www.insna.org *****

Dear all,

 

I am applying for ethical approval to conduct a social network analysis in a hospital department. I would like to run an online survey including all staff members and ask participants to indicate up to 25 people they communicate and rate the frequency and type of interaction. It was drawn to my attention that I might have an issue with this, as individuals may not wish to consent but their name would still appear in the survey list.

 

Could you please advise how I could go around this issue?

 

All the best,

Rosica


Rosica Pachilova 
PhD Researcher
University College London

Mobile: +447414001893
UCL Bartlett    LinkedIn    Twitter   
_____________________________________________________________________ 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.

_____________________________________________________________________ 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.


-- 
_ __ ___ ____ ___ __ _ __ ___ ____ ___ __ _

Christian Steglich, University of Groningen
Faculty of Behavioural and Social Sciences
Grote Rozenstraat 31, NL-9712 TG Groningen
http://www.gmw.rug.nl/~steglich/
_ __ ___ ____ ___ __ _ __ ___ ____ ___ __ _
_____________________________________________________________________ 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.

_____________________________________________________________________ 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.