[Air-L] Accessing data from closed groups on FB?

Wed Nov 8 05:39:18 PST 2017

The only way I know of accessing anonymized, aggregated FB data from private accounts or for private attributes - though probably not for closed groups - is via FB's advertising API and its audience estimates method.
https://developers.facebook.com/docs/marketing-api/reference/ad-account/reachestimate/
Using that method you can ask queries of the type "how many FB users match criteria X". The targeting criteria can include interests on FB, education level, political orientation, income, marital status, ... (see https://developers.facebook.com/docs/marketing-api/targeting-specs/).

I've used this data source quite a bit in my own research (see https://ingmarweber.de/publications/) in particular as a "digital census" where we're working with demographic researchers to estimate stocks of migrants or birth rates or gender inequalities (see http://data2x.org/big-data-challenge-awards/#digital).

If you'd like to use these reach estimates, there's a Python wrapper at https://github.com/maraujo/pySocialWatcher and an R wrapper at https://github.com/ccgilroy/r-estimates-fb-ads.

Also see "Not-so-obvious Online Data Sources for Demographic Research"
https://www.slideshare.net/IngmarWeber/notsoobvious-online-data-sources-for-demographic-research

Best,
Ingmar

-----Original Message-----
From: Air-L [mailto:air-l-bounces at listserv.aoir.org] On Behalf Of Bernhard Rieder
Sent: Wednesday, November 8, 2017 2:54 PM
To: air-l at listserv.aoir.org
Subject: Re: [Air-L] Accessing data from closed groups on FB?

Hello,

I am pretty confident that Facebook doesn't just "sell data", especially not data that has a strong privacy dimension to it. In 2015, all FB APIs implemented a pretty strict distinction between what FB considers to be public and private (with some modulation of that distinction left to users) and I know of no means to pay for access to private data. There are some particular APIs reserved for partners and DataSift sells access to the Topic API (stream of public posts) via their STREAM product (formerly PYLON), but nothing that concerns the "private"-facing side of FB, AFAIK. I would be very, very surprised if there is a commercial pathway to accessing data from a private group, even in aggregate.

I fear that there are only two things you can do:
1) If the group is not too big, you can use manual work or write a small browser plugin to scrape data from the interface.
2) Get the admin to set the group to public => get data => set it to private again.

If there are other ways, I'd be interested to know.

Good luck and cheers,
Bernhard

---
Bernhard Rieder | Associate Professor | New Media and Digital Culture University of Amsterdam | Turfdraagsterpad 9 | 1012 XT Amsterdam | The Netherlands http://thepoliticsofsystems.net | http://labs.polsys.net | https://www.digitalmethods.net | @RiederB

> On 8 Nov 2017, at 09:07, Raquel Rennó <raquelrenno at gmail.com> wrote:
>
> Hi all,
>
> Just a reminder for the near future, the educational sector will have
> to comply with the GDPR http://www.eugdpr.org/ that will be
> implemented in the EU in May 2018.
> http://academytoday.co.uk/Article/gdpr-how-academies-should-prepare-fo
> r-compliance
>
>
>
> On Wed, Nov 8, 2017 at 9:36 AM, Ylva Hård af Segerstad <
> ylva.hard-af-segerstad at ait.gu.se> wrote:
>
>> Thanks, Jeremy, for this clarification.
>>
>> I also think it’s worth pointing out that Facebook - and other
>> gigantosaur platforms that many of us have sold our souls to - is
>> allowed to do for commercial purposes what researchers are not
>> allowed to do for scientific purposes. As researchers we have to make
>> sure we minimize harm, and as the ethics discussions at the recent
>> conference in Tartu, maybe even strive to also do good.
>>
>> We’ve addressed ethical concerns in relation to studying closed
>> groups online using a specific and sensitive case study to discuss
>> these issues in a chapter in the recently published volume "Internet
>> Research Ethics for the Social Age: New Cases and Challenges", edited
>> by Michael Zimmer and Katharina Kinder-Kurlanda:
>>
>>
>> Hård af Segerstad, Y., Kasperowski, D., Kullenberg, C. & Howes, C. (2017).
>> Studying Closed Communities On-line: Digital Methods and Ethical
>> Considerations beyond Informed Consent and Pseudonymity. In M. Zimmer & K.
>> Kinder-Kurlanda (Eds.), Internet Research Ethics for the Social Age:
>> New Cases and Challenges: Peter Lang. (213-225).
>>
>> It is available under a Creative commons license and can be found here:
>> https://www.peterlang.com/view/product/82023?format=EPDF
>>
>> I am indeed a member of the closed group. What we are intending to do
>> in the particular project I was asking for is to invite a number of
>> individual members to participate in a longitudinal study. We’d like
>> to follow their activity in the closed group and possibly also their
>> activity in other groups on FB and on blogs or whatever else they do.
>> This will be accompanied by interviews and maybe also surveys now and
>> then, depending on how many we can recruit and for how long we go on.
>> What I’d like to do is to get the FB-activities for these
>> individuals. And for this, we will be able to get informed consent
>> and active research participation during a longer period of time from these individuals.
>>
>> As it is no longer possible to scrape data from closed groups using
>> the FB API we’d like to explore the possibility of buying the data
>> for these particular individuals from FB. If we get that data it will
>> definitely also lead to other interesting ethical concerns, re commercial powers'
>> possibilities to override researcher obligations but which may be
>> used by researchers. Which is my hidden agenda to look into, I guess.
>>
>> So: does anyone know how to go about buying data from Facebook?
>>
>> Hälsningar
>>
>> Ylva H
>> ____________________________
>>
>> Ylva Hård af Segerstad
>> Docent | Associate Professor
>> Institutionen för Tillämpad IT | Department of Applied IT Avdelning
>> Lärande Kommunikation och IT | Divison of Learning Communication and
>> IT University of Gothenburg
>> SE-412 56 Gothenburg, Sweden
>> +46(0)70 3314739 (mobile)
>> hardy at ait.gu.se<mailto:hardy at ait.gu.se>
>>
>>
>>
>>
>>
>> 8 nov. 2017 kl. 02:08 skrev Jeremy Hunsinger <jeremy at tmttlt.com<mailto:
>> jeremy at tmttlt.com>>:
>>
>> There are innumerable possible models of research design that would
>> not require informed consent in this situation.  Informed consent is
>> usually only required when there is some harm that must be agreed to,
>> but it is easy to conceive of any number of situations where this
>> data collection leads to nothing that requires informed consent.  It
>> isn't ethically anything, until you understand the whole design.
>> Also remember the ethical standards are generated from different
>> principles than harm in many Scandinavian countries as noted in the
>> first set of AoIR guidelines, so your instincts of informed consent
>> might not be quite fitting.  So let's not 'throw the baby out with
>> the bathwater' and instead try to answer the specific question if we can.
>>
>>
>>
>> --
>> jeremy hunsinger
>> Associate Professor
>> Communication Studies
>> Wilfrid Laurier University
>>
>> Collaboratory for Digital Discourse and Culture Virginia Tech
>> www.tmttlt.com<http://www.tmttlt.com>
>>
>> ()  ascii ribbon campaign - against html mail
>> /\                        - against microsoft attachments
>> http://www.stswiki.org/  sts wiki
>> http://transdisciplinarystudies.tmttlt.com/  Transdisciplinary
>> Studies:the book series
>>
>> I am always doing that which I can not do, in order that I may learn
>> how to do it.
>> -Pablo Picasso
>> _______________________________________________
>> 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/
>>
>
>
>
> --
> Raquel Rennó
> Politics of data researcher - Tactical Tech Collective
> https://tacticaltech.org/
> Researcher ID :   F-5319-2014
> Profile URL:http://www.researcherid.com/rid/F-5319-2014
> <http://www.zzzinc.net>
> -------
> _______________________________________________
> 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/
________________________________
        Disclaimer
________________________________

This email message, including any attachments is intended solely for the addressee(s), and may contain information that is confidential or legally protected. Any use, disclosure, copying, printing or distribution of the information contained herein by persons other than the designated addressee is unauthorized and may be unlawful. If you believe that you have received this email in error, please notify the sender immediately and delete this email and any attachments from your system and discard any printout thereof. Qatar Foundation makes no warranty that this email is error- or virus-free .The views, opinions, conclusions, and other information expressed in this email are not necessarily endorsed by Qatar Foundation unless otherwise indicated by an authorized representative. As such, Qatar Foundation does not accept any liability for any emails and attachments that may violate the relevant applicable laws.

________________________________