[Air-L] Inquiry on screen shots

Stuart Shulman stuart.shulman at gmail.com
Wed Feb 27 17:17:18 PST 2019


It violates the Twitter ToS if you even look at, or worse display/present,
the content of deleted Tweets. The practice is one (of many) that cause
Twitter fits with academia in general and it has contributed to current
state of affairs. Every single spreadsheet (or other non-compliant
collection) of Tweets without a live connection to Twitter (to confirm the
Tweet is not yet deleted) violates the Twitter ToS and the basic notion of
the "right to be forgotten" as articulated through European public policy
when a researcher looks at the deleted content. Screenshots of Tweets would
not be amenable to automated systems of checking for deletions (unless you
attach key metadata and build an API-enabled system for checking deletions
in real time). The right to be forgotten is ensured (for now) only at the
instant a user of a compliant system seeks to look at a Tweet live in the
Twitter display and is unable to see deleted Tweet content. If you kill the
connection to live Twitter you preclude, or at least make much more
difficult, the possibility for compliance. Looking at the content of Tweets
as raw data disconnected from Twitter almost certainly ensures ToS
violations and the wrath of Twitter, it also invites a painful future
Napster moment in the courts for a handful of the super-abusers of the ToS.
I am curious if someone can point to a first rate Office of Sponsored
Research document that fully articulates the responsibilities of academic
researchers when using Twitter data. What is the state of the art in
compliance? Conversely, what are the most problematic ways that academics
flout the right to be forgotten by storing, reviewing, and publishing
content, for example, from deleted Tweets? Many are unhappy about the
changing ecosystem; however how many are willing to accept their share of
the responsibility for why things have changed so much so fast? There is
work to be done creating a better shared understanding that recognizes the
risks when you make humans into research subjects without their consent.

On Wed, Feb 27, 2019 at 7:07 PM Amy Mowle <amy.mowle at live.vu.edu.au> wrote:

> I had a conversation regarding violating ToS for research purposes with my
> university’s ethics officer this morning and she made a good point - when
> users sign up to a social network, they do so assuming their data is
> protected by the ToS they personally agree too. Violating the ToS May
> equate to violating the contract of consent between the unknowing subjects
> of the study and the researcher.
>
> It’s a tough one, but necessary to consider! One must tread carefully and
> think ethically if we are to avoid further restrictions on the collection
> and utilisation of publicly available data. I would err on the side of
> caution and not use screen grabs without first requesting the consent of
> the poster.
>
> Amy
>
> > On 28 Feb 2019, at 5:24 am, Livingstone,S <S.Livingstone at lse.ac.uk>
> wrote:
> >
> > Good advice. And I agree: let’s start with first question- why do you
> want to show these images and what would be lost if you didn’t? Best, S
> >
> >> On 27 Feb 2019, at 17:48, Charles M. Ess <c.m.ess at media.uio.no> wrote:
> >>
> >> I'll be eager to hear what others think ...
> >> the problem with such images is that they are easily identifiable
> simply as images, no matter how careful you may have been to hide the
> identifying texts.  And while some here will argue that since the images
> are (quasi-) publicly available - or are they?  That is, are these drawn
> from open sites or sites that require a login?  If the latter, is there any
> guidance from the Terms of Service as to the use of images?
> >>
> >> (Probably forbids them - in which case you then get to enter the
> exciting world of considering violating a ToS for the sake of research ...
> This is its own domain of discussion, especially vis-a-vis Facebook and its
> recent change in the ToS.  Our national ethics board will not give firm
> guidance either way - i.e., yes, it's o.k / no, it's not o.k.: one of our
> researchers is waiting to hear from the data security agency what they
> think of the matter ...  Any updates on how this is faring in the U.S. or
> elsewhere - i.e., whether or not violating the ToS = violating the law?)
> >>
> >> So a first question would be - why do you need to provide the images in
> your publication?
> >> If they are necessary in some form to illustrate your method - o.k.,
> but then consider some additional options.
> >> One would be to ask for consent from the person(s) depicted in the
> image.  Perhaps difficult to do and perhaps not likely to acquire, but it
> is an option some researchers would pursue.
> >> Alternatively, a common technique is to use software to modify the
> images so that they no longer provide enough data for recognition and
> identification, but still provide enough of an outline to suggest /
> illustrate the point(s).
> >> I can't give you specific recommendations, but I've seen examples of
> this any number of times at AoIR and other conferences, so perhaps some
> members of the list will have specific suggestions.  Depending on what
> exactly you want to illustrate / demonstrate with the image will determine
> how far and in what ways you can blur out / modify it.
> >>
> >> My 2 cents.  Hope others will have additional wisdom, guidance, and
> experience to offer.
> >>
> >> and good luck!
> >> - charles ess
> >>
> >>
> >>
> >>
> >>> On 27/02/2019 17:35, evelyne wanjiku via Air-L wrote:
> >>> Greetings members,
> >>> I am conducting a research on dialogues around revenge pornography on
> social media platforms, fb, Twitter and telegram I am using a discourse
> analysis approach.
> >>> As part of the analysis, i have picked out several dialogues and
> screen shot them. My question is would it be ok for me to publish screen
> shots from the various platforms? I have taken precaution to disguise/hide
> the names/identities of those commenting.
> >>> Has anyone engaged in such? And what would be your advice with
> publishing screen shots?
> >>> Sent from Yahoo Mail on Android
> >>> _______________________________________________
> >>> 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/
> >>
> >> --
> >> Professor in Media Studies
> >> Department of Media and Communication
> >> University of Oslo
> >> <http://www.hf.uio.no/imk/english/people/aca/charlees/index.html>
> >>
> >> Postboks 1093
> >> Blindern 0317
> >> Oslo, Norway
> >> c.m.ess at media.uio.no
> >> _______________________________________________
> >> 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/
> > _______________________________________________
> > 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/
> _______________________________________________
> 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/



More information about the Air-L mailing list