[Air-L] Downloading Twitter feeds

Peter Timusk peterotimusk at gmail.com
Sun Apr 23 19:46:20 PDT 2023 

I work in government research with massive surveys and tax data in Canada.
Privacy and protecting confidentiality is a large part of my job. This
topic is also well developed in census, health research and other areas in
many countries in the USA, Europe and elsewhere. You can look for these
best practices in these fields where you will find simple to complex
methods to assess privacy invasion risk. I have to analyse researchers
models to assess them for publication. This is generally quantitative
models but that does not change the nature of the rights to privacy. There
are methods of aggregating data or anonimizing data   even around
qualitative methods. My experience attending AoIR and related conferences
is some new researchers are not very informed of privacy rights of
respondents to the same degree as I am. I have legal obligations to the
resondents of our surveys. Example what may be publicly on the Internet may
also seem to useful for research purposes. When in fact the author of the
data may have felt a sense of privacy when writing it. All subjects of
research should be informed that they are subjects of research.

A researcher finding an interesting medical self help Facebook group or
vulnerable population may find a lot to theorize about. It may seem a great
research topic. But if the membership of these people in this group were
disclosed perhaps someone's life or rental contract or job may be at risk.
This needs to be thought about.


Again look for topics on anonymizing data, disclosure analysis, synthetic
data, masking data, suppression of data, confidentiality protection.

On Sun., Apr. 23, 2023, 9:58 a.m. Marcelo Santos via Air-L, <
air-l at listserv.aoir.org> wrote:

> Good point Floriana, but again: are a private company's terms of service a
> legal boundary, sufficient to prevent researchers to deal with issues of
> public interest. If you ask me, no. But I know  many IRBs follow the terms
> of service, which is a very lazy shortcut, IMHO, to solve complex
> ethical issues.
>
> I would very much like to know what others think, though.
>
> Best
> Marcelo
>
> Em sáb., 22 de abr. de 2023 às 14:11, Grasso, Floriana <
> floriana at liverpool.ac.uk> escreveu:
>
> > Hi Rose
> >
> > I suspect you will not be able to collect *any* data from Facebook, no
> > matter what ethic approval you have, as their terms of service
> > <https://www.facebook.com/legal/terms> say (in 3.2.3):
> >
> > You may not access or collect data from our Products using automated
> means
> > (without our prior permission) or attempt to access data that you do not
> > have permission to access. We also reserve all of our rights against text
> > and data mining.
> >
> >
> > which basically says you cannot scrape content, but I would be interested
> > to know if the legal team at your institution thinks otherwise?
> >
> > Floriana
> >
> > ------------------------------
> > *From:* Air-L <air-l-bounces at listserv.aoir.org> on behalf of Marcelo
> > Santos via Air-L <air-l at listserv.aoir.org>
> > *Sent:* 22 April 2023 18:50
> > *To:* Butler, Rose <R.Butler at shu.ac.uk>
> > *Cc:* air-l at listserv.aoir.org <air-l at listserv.aoir.org>
> > *Subject:* Re: [Air-L] Downloading Twitter feeds
> >
> > Hi Rose, I am very interested in how you solve that puzzle.
> >
> > I have been working lately with WhatsApp groups that are publicly
> available
> > and anonymizing and performing aggregate analysis have gotten me through
> > ethics approval. If you are going to perform more qualitative research it
> > might be very advisable that users are aware of it and consent, which
> might
> > prove very complicated.
> >
> > Let me know how this moves forward.
> >
> > Best wishes
> > Marcelo
> >
> > Em sex., 21 de abr. de 2023 às 09:32, Butler, Rose via Air-L <
> > air-l at listserv.aoir.org> escreveu:
> >
> > > Hello
> > > I have an ethical question about scraping data from FB (that is already
> > in
> > > the public domain)
> > >
> > > I am preparing a project for ethical clearance. The research focuses on
> > > two community FB groups of 3000 and 7000 members respectively and of
> > which
> > > I am a member of both.
> > > As part of the research methods I will automate the collection of data
> > for
> > > analysis (partly on spreadsheets), all of this would be possible
> manually
> > > if I had the time – automation allows me to collect a lot of data
> > quickly –
> > > it is not extending access.
> > >
> > > If this data is anonymised and not published would that be enough to
> > > protect the privacy of members within the group?
> > > I wondered if anyone had experience of doing this and what additional
> > > safeguards they needed to implement for ethical clearance. The problem
> > here
> > > is that a closed group is semi-private, individuals may feel like it is
> > > private even if it spans the whole community (this time a small town).
> > >
> > > Thanks
> > >
> > >
> > >
> > > From: Air-L <air-l-bounces at listserv.aoir.org> on behalf of Stuart
> > Shulman
> > > via Air-L <air-l at listserv.aoir.org>
> > > Date: Tuesday, 18 April 2023 at 12:10
> > > To: M.E.Sciubba <mesciubba at gmail.com>
> > > Cc: air-l at listserv.aoir.org <air-l at listserv.aoir.org>
> > > Subject: Re: [Air-L] Downloading Twitter feeds
> > > CAUTION: This message was sent from outside the University, purportedly
> > > from air-l-bounces at listserv.aoir.org .
> > >
> > > Please check the sender is legitimate before responding. Please treat
> any
> > > links or attachments with care -  do not follow or open them unless you
> > are
> > > sure they are genuine.
> > >
> > >
> > >
> > >
> > > The answer is complex, uncertain, and shifting. It is more than just a
> > > technical issue of "Where can I get data now?" There are fundamental
> > > legal, jurisdictional, market, equity, and political questions
> > > unresolved. Resolution seems unlikely in the short term. As a
> > > political scientist interested in election interference, I do not
> > > anticipate a positive resolution before the 2024 U.S. election. I
> > > would love to be wrong on that prediction. I have heard there are a
> > > few firms that retain agreements to market real time and historical
> > > Twitter data. You might ask your librarian about access to Crimson
> > > Hexagon, Meltwater, or other similar services with long term
> > > agreements Mr. Musk has not broken, yet. Nobody on the AIR-L list has
> > > confirmed whether the academic API is shut down. Is it? I don't have
> > > those credentials to check.
> > >
> > > As I have written previously, there are massive troves of Tweets in
> > > every computer science department and many social science departments.
> > > Any one of the raw JSON Tweet archives could be loaded into a free
> > > account on DiscoverText and then shared for teaching or research. Talk
> > > to the data folks on your campus and across your discipline about what
> > > is extant. Do not look for spreadsheets; find the raw JSON.
> > > Spreadsheets are significantly degraded historical objects that cannot
> > > be considered accurate representations of Tweets. Tweets live in the
> > > Twitter display and die in spreadsheets. I remain befuddled by the
> > > Python/R nexus. Those spreadsheets of Tweets simply are not Twitter.
> > > Any qualitative researcher knows what I mean. Meaning is produced in
> > > the interaction of the display elements. Meaning is destroyed,
> > > diluted, and bent out of shape in spreadsheets.
> > >
> > > The loss of real time access to new data is a problem and dangerous
> > > for democratic systems. This does not change the fact that so much
> > > data has been gathered and stored, like so many old newspapers. For
> > > example, here is a list of the Top 50 datasets in my DiscoverText
> > > account that any academic on this list or anywhere in the world can
> > > access via the "peer network" that enables collaboration, annotation,
> > > measurement of inter-rater reliability, and the creation of gold
> > > standard training sets for machine-learning via a graphical user
> > > interface that features the Twitter display.
> > >
> > >
> > >
> >
> https://docs.google.com/spreadsheets/d/1j-Y40WmwFIX8pidfAcxgAt5JB3A3UyB2psh_ULQdynA/edit?usp=sharing
> > >
> > > "The Case Against Spreadsheets as a Primary Twitter Research Tool"
> > > https://vimeo.com/526218014
> > > Question: When is a Tweet not a Tweet? Answer: When it appears in a
> > > spreadsheet, or is deleted, or the account has been suspended.
> > >
> > >
> > >
> > > On Tue, Apr 18, 2023 at 5:57 AM M.E.Sciubba via Air-L <
> > > air-l at listserv.aoir.org> wrote:
> > >
> > > > Hi Rotem and All,
> > > >
> > > > I am also looking for solutions to this. I hope somebody has an idea
> of
> > > how
> > > > to keep doing research on Twitter.
> > > >
> > > > Eleonora
> > > > *--*
> > > >
> > > > Dr. Maria Eleonora Sciubba
> > > > (she/her)
> > > > (2022): “Adesso m’incazzo!”: Swearwords as resources for managing
> > > negative
> > > > emotions in interaction <
> > https://mediazioni.unibo.it/article/view/15263>
> > > > (2021)
> > https://rolsi.net/2021/06/02/guest-blog-em-ca-for-racial-justice/
> > > >
> > > > *TSHD - Grant Design & Writing *
> > > > *Department of Culture Studies - Senior Researcher*
> > > > Editor, Internship Organizational Supervisor *Diggit Magazine*
> > > > <https://www.diggitmagazine.com/>
> > > > Tilburg University
> > > > Twitter: @LolaSciubba
> > > >
> > > >
> > > >
> > > >
> > > >
> > > > *Be green. Keep it on the screen.*
> > > >
> > > >
> > > > Il giorno mar 18 apr 2023 alle ore 11:15 Rotem Perach via Air-L <
> > > > air-l at listserv.aoir.org> ha scritto:
> > > >
> > > > > Hi All,
> > > > >
> > > > > Can anyone recommend a way to download the feeds (historical tweets
> > and
> > > > > re-tweets) of Twitter users? Is scraping for example using Data
> > Miner a
> > > > > good idea?
> > > > > With the changes to Twitter's API rules, API-based websites such as
> > > > > Vicinitas.io have stopped working (or will stop soon), and I'm
> > looking
> > > > for
> > > > > other approaches.
> > > > >
> > > > > Thank you for your help.
> > > > >
> > > > > Kind regards,
> > > > >
> > > > >
> > > > > Rotem
> > > > >
> > > > > -----------
> > > > >
> > > > > Dr Rotem Perach (he/him)
> > > > > Senior Research Fellow
> > > > > School of Social Sciences
> > > > > University of Westminster
> > > > >
> > > > > ORCID<https://orcid.org/0000-0002-8647-4367> | Twitter<
> > > > > https://twitter.com/DrRotemPerach/><
> > > https://twitter.com/DrRotemPerach/%3e>
> > <https://twitter.com/DrRotemPerach/%3e%3E>
> > > > >
> > > > >
> > > > > The University of Westminster is a charity and a company limited by
> > > > > guarantee. Registration number: 977818 England. Registered Office:
> > 309
> > > > > Regent Street, London W1B 2HW.
> > > > >
> > > > > This message and its attachments are private and confidential. If
> you
> > > > have
> > > > > received this message in error, please notify the sender and remove
> > it
> > > > and
> > > > > its attachments from your system.
> > > > > _______________________________________________
> > > > > 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/
> > > _______________________________________________
> > > 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