[Air-L] Ethical challenges in qualitative research using Facebook

[Charles Ess]

Hi Emily and colleagues,

First of all, kudos for your taking such good care at looking into these
ethical dimensions of your research.

Secondly, a couple of questions: are the posts drawn exclusively from
Australian FB users - and if so, what is the Australian law regarding what
appears in a public context online?  In the U.S., last I knew, such
postings are considered automatically copyrighted, and so one ethical-legal
issue is, should this be the law in Australia as well, to recognize the
posters as copyright holders.

Thirdly, the methodological approach and its affiliated ethical challenges
as you describe them are, in my somewhat limited experience, rather
standard.  This means that I've seen similar projects - including ones
involving far more sensitive expression - receive IRB and REB approval, so
I would not stop with your worst-case scenarios (though these are clearly
and carefully thought out - again, kudos).

Some possibilities.  In terms of dissemination / publication - would it be
possible to (a) disguise many, if not the majority of posts by way of a
paraphrase and/or aggregate identity (if appropriate), thereby avoiding
direct citations, while (b) using some direct citations when absolutely
necessary - and then requesting permission to do so?  While there are
difficulties with requesting such permission, as you point out, in my
experience (i.e., reading about and/or discussing similar cases in a
variety of contexts - the U.S., the E.U., and Scandinavia), these are not
insurmountable. To be sure, one needs to exercise caution regarding
identity - though these days, it seems that the vast majority of FB
identities are more or less authentic.  Ditto in the case of a vulnerable
group, beginning with legal minors.  But with care, permission can be
requested when necessary in ethically sensitive and responsible ways.  Of
course, there will likely be a few instances of "no, thank you" (or less
politely) - but with any luck, having to omit perhaps a handful of choice
citations (and relegating them back into more anonymized form if possible /
necessary), will not prevent you from having sufficient evidence and
analyses to effectively respond to your research questions.

That's at least my first take - and I only offer these as first thoughts,
knowing that there are real experts on this list, including several
colleagues in Australia who will have better information about the Aussie
context, as well as our colleagues Elizabeth Buchanan, Michael Zimmer,
Annette Markham, and many others in the northern hemisphere with extensive
experience in these domains as well who will have much more to add.

Hope this helps, then, as at least a helpful start in the discussion.

Again, kudos and all best wishes,
- charles
-- 
Professor in Media Studies
Department of Media and Communication
University of Oslo
<http://www.hf.uio.no/imk/english/people/aca/charlees/index.html>

Editor, The Journal of Media Innovations
<https://www.journals.uio.no/index.php/TJMI/>

Postboks 1093
Blindern 0317
Oslo, Norway
c.m.ess at media.uio.no

On Tue, Jul 26, 2016 at 7:06 AM, Emily Wolfinger <emilywolfinger at hotmail.com
> wrote:

> Hi AoIR Community,
>
>
> I am a PhD Candidate and I have run into some ethical issues in my
> research, which I am looking for some guidance on.
>
>
> I am exploring Facebook user perceptions of sole mother poverty and
> welfare in Australia, focusing on a period of welfare debate in which sole
> parent pension amendments were introduced (May 2012-January 2013).  I'm
> undertaking a discourse analysis of a subset of comments across four
> categories of public pages and groups- media, political parties and
> ministers, welfare/charity organisations and sole mother networks- that
> were published in response to the amendments.
>
>
> Although Facebook Data Policy considers this information public, Internet
> research ethics guidelines and other academic papers point to a number of
> ethical issues around publishing the comments of users without obtaining
> their consent.  As I am doing a post-structural discourse analysis this is
> unavoidable unless I consider paraphrasing or similar techniques that aim
> to protect the identity of users, however this strategy raises questions of
> scientific rigor and does not seem to be one that is widely used or indeed
> fool proof.  There are also issues around contacting users for consent, for
> example users could be underage or belong to other vulnerable groups.
>
>
> I am left with two options if I take a worst case scenario approach to
> these dilemmas - either tweak my research question (for example to look at
> the posts of public figures and organisations such as media outlets) or
> consider alternative research methods which do not present the ethical
> challenges of discourse analysis or other methods of close analysis, but
> allow me to answer my research question (What were the Facebook user
> perceptions of sole mother poverty and welfare in Australia between May
> 2012 and January 2013?).
>
>
> Any suggestions for readings, tips or advice regarding ways forward
> including methods would be most appreciated.
>
>
> Many thanks in advance,
>
> Emily
> _______________________________________________
> The Air-L at listserv.aoir.org mailing list
> is provided by the Association of Internet Researchers http://aoir.org
> Subscribe, change options or unsubscribe at:
> http://listserv.aoir.org/listinfo.cgi/air-l-aoir.org
>
> Join the Association of Internet Researchers:
> http://www.aoir.org/