Elizabeth A. Buchanan
Associate Professor and Director,
Center for Information Policy Research
School of Information Studies
University of Wisconsin-Milwaukee
Principal Investigator
Internet Research Ethics Commons
elizabeth@internetresearchethics.org
Support for this research comes from the
National Science Foundation.
Presentation to the Secretary’s Advisory
Committee on Human Research
Protections, July 2010
Internet Research Ethics
and IRBs
How Does the Internet Fit In To
Research?
Internet as a TOOL FOR research or…
Internet as a MEDIUM/LOCALE OF research
TOOL=search engines, databases, catalogs, etc…
MEDIUM/LOCALE=chat rooms, MUDs, MOOs,
newsgroups, home pages, MMORPGs, blogs, skype,
tweeting, online course software, etc
Cases to Consider
We had a researcher using the website "Gay Bombay" to study gay Indian men's attitudes, and
the board was worried that since homosexuality is illegal in India, would participation get the
respondents in trouble somehow? (Transborder issues, risk, consent)
A student wishes to analyze blog postings as part of her Master’s thesis. Must she seek IRB
review? If she does not, might she face journals who will not publish her work because it was
not approved?
A researcher wanted to use a public list archive, but—in order to post, membership was
required. Must he gain consent? (No longer fits the “public park” analogy?)
Can a researcher use mechanical turks ( to complete research related tasks, eg, survey responses
without IRB oversight?
Is aggregated facebook data really anonymous? (Zimmer, 2009/2010)
(Interesting reinterpretation: Virtual Milgram)
Emergence of IRE Awareness…
(and Some Confusion!)
1999, Frankel and Siang report
2002, AoIR “Ethical Decision Making…” (Watch for 2010
revision!)
2003, Buchanan, Readings in Virtual Research Ethics; Chen and
Hall, Online Social Research
Scattered literature across disciplines
IRBs facing new lexicon and challenges in their charge to protect
human subjects
A redefinition of what counts as a “human subject” (avatars, turks,
etc)
The Relationships
Regulations/Regulatory Boards (Policy)
Research Participants/Online Norms/
Self-Community Generated
Ethical Frameworks
Researchers/Disciplinary Practices/
Professional Ethics
The Issues
Identities/PII
Ownership
(Formal/
Informal)
Anonymity/
Pseudonymity
Security
Privacy
Ethics
Recruitment
Trolling
P/O
Consent
Representation/
Subjectivity/
Objectivity
Methodology
Risk/Harm
Ideology
So Now What?
What was missing from the literature: Empirical research on the
state of internet research within IRBs: What we were reviewing
and how? How were we dealing with all of those pressing issues?
And—how were we preparing ourselves?
Buchanan and Ess received 2006 NSF award: Internet Research
Ethics: Discourse, Inquiry, Policy to survey US-based IRBs
Exploratory study to examine the state and perceptions of Internet
research reviews in US-based IRBs
Developed a comprehensive database crossing Carnegie
Classificationshttp://www.carnegiefoundation.org/classifications/
index.asp?key=783)
Methods



Exploratory study to examine the state and perceptions
of Internet research reviews in US-based IRBs
Developed a comprehensive database crossing Carnegie
Classifications
(http://www.carnegiefoundation.org/classifications/in
dex.asp?key=783)
Sent hard copy, mixed method surveys to 750 IRBs
across the US
Institutional Types
B
A
S
IC
T
rib
a
l
S
p
e
c
/B
u
s
S
p
e
c
/H
e
a
lth
S
p
e
c
/M
e
d
B
a
c
/D
iv
e
rs
e
B
a
c
/A
&
S
-9
9
R
U
/V
H
R
U
/H
D
R
U
M
a
s
te
r's
/S
M
a
s
te
r's
/M
M
a
s
te
r's
/L
2007-08 Data
Internet Research:
Of Importance or Concern to Board
Miss ing
2.0%
no
47.6%
y es
50.3%
Data
Reviewing Internet Research Protocols:
Checklist, Review Tool, Policy, Guidelines Available
y es
Miss ing
3.4%
under dev elopment
17.0%
4.1%
y es : exc ellent
2.0%
y es : adequate
1.4%
no
72.1%
Data
Internet Research Protocols:
Designated Person to Review
Miss ing
y es
1.4%
15.0%
no
83.7%
Data
Internet Research Protocols:
Regulatory Documents Help in Review
(i.e. CRF, OHRP Decision Charts)
Miss ing
12.2%
y es
45.6%
no
42.2%
Data
HIPPA Laws:
Impacts Review of Internet Research
Miss ing
10.9%
y es
5.4%
no
83.7%
Data
Internet Research Ethics:
ANY Research Ethics Issues:
Training for Researchers
Training for Researchers
y es
Miss ing
6.8%
.7%
y es
y es : optional
8.8%
4.8%
no
23.1%
y es : optional
22.4%
Miss ing
5.4%
no
y es : required
38.8%
84.4%
y es : required
4.8%
Growing Use of Commercial Tools?
Buchanan and Ess, 2008, online surveys are the most frequently
reviewed type of online research
Use of Online Survey Tools:
Creating Online Surveys:
Researchers Required to Submit Privacy/Use Policies
Institution Has a Tool
Miss ing
Miss ing
5.4%
10.9%
unsure
14.3%
y es
no
51.0%
38.1%
no
51.7%
y es
28.6%
Anonymity/ Confidentiality
Distinction between anonymous and confidential
“Our first duty as researchers is to honor the promise of confidentiality”
(Easter, Davis, & Henderson, 2004).
Is there a truly secure online interaction? What type of Internet
location/medium is safest? Is an “anonymous” survey possible?
How will subjects/participants be protected? Is encryption
enough?
Can one be anonymous online? One may have a “different”
identity (e-betty), but that is still “me.” If e-betty is portrayed in
research on an electronic support group for a medical condition,
will she be identifiable? If so, at what risk?
Revealing Identities
How should online participants be identified in research
reports given the traceability and trackbackability of online
data?
Screen names, pseudonyms of screen names?
By changing screen names, do you detract from the “reality”
or “reputation” of the participant?
Text searches can reveal more context than a researcher may
in her reporting (potential risk)
Allow participants to make this decision? Part of informed
consent?
Possibilities for Consent: Lawson,
2004
1. consent to having their nickname and communicative text used for data
analysis only (no publication of name or text);
2. consent to having either their nickname or text published in an academic
work, but never together (i.e., no identifiers);
3. consent to having either their nickname or text published in an academic
work, but never together (i.e., no identifiers) and providing they get to see
the ‘write up’ prior to publication;
4. consent to having both their nickname and text published in academic
work, thereby being credited as the authors of their own words; or
5. consent to having both their nickname and text published in academic
work, thereby being credited as the authors of their own words, providing
they get to see the ‘write up’ prior to publication (p. 93).
Public and Private Spaces
Is a particular forum, listserve, chat room, bulletin board, etc
considered by its members to be a public space or a private space?
What expectations of privacy exist? Members only? Public
newsgroups? What sort of methods are being used? This might not
be reviewable, eg, discourse analysis?
What role does the researcher play in the space? (Observer,
participant, member, other?)
What is the content of the data? (Sensitive/non-sensitive?) Medical
information, eg, Caring Bridge?
Sveningsson’s continuum: Non-Sensitive Information
?
?
Private
Public
No
?
Sensitive Information
Ownership/Stewardship of Data?
With f2f research, the researcher, eg, conducts an observation. Writes field
notes. Returns a report to the participant when completed. Owns “it.”
Versus:
A researcher conducts an observation of some newsgroup interactions. A
log/transcript is generated. Researcher has a copy. So do the participants. So
does the server/administrator of the news group. Who owns “it?
How long does e-data last? (“I will destroy the data in 5 years…” may mean
nothing in an online context where researchers are not in control)
Advise researchers to inform participants/subjects about the longevity and
potential risk of data intrusion.
Cloud computing
Respect for Persons/Autonomy
(i) information obtained is recorded in such a manner that human subjects can be
identified, directly or through identifiers linked to the subjects; and (ii) any disclosure
of the human subjects' responses outside the research could reasonably place the
subjects at risk of criminal or civil liability or be damaging to the subjects' financial
standing, employability, or reputation.
Practical challenges in obtaining informed consent (fluidity, changes in group
membership, etc)
Verification of understanding (the cornerstone of informed consent is understanding!)?
Will something from a public space come back to haunt us?
Use of archived quotes never intended to be represented in research?
Click boxes? Hard copies? Blanket statements necessary (“I understand that online
communications may be at greater risk for hacking, intrusions, and other violations. Despite these
possibilities, I consent to participate”)?
Blogs (don’t meet the definition of human subject as in45cfr46.102f)
Are avatars “human subjects”?
Recruitment
Equity/fair representation in subject pool may not be
possible: in our protocols, we may see “unjust” subject
selection based on type of site—eg, WoW
How does the researcher enter the research space to
begin recruiting? Site owners, moderators, gatekeepers
are key in some fora
What if some in a community consent, others do not?
Researchers should have plans for this reality.
Research Ethics 2.0
and Beyond






Collisions across disciplines (More and more disciplines, CS,
Engineering, research IS human subjects; eg, malware, bot
research)
Negotiating or interrogating Terms of Service—in some cases, we
are simply not able to control data, or know where data is, lives,
for how long, who has access….OR: Third Party Sites, third party
ethics….
Amplify the process of community decision-making—communal
consent may become the norm?
Blur the boundaries that are essential to more dichotomous
models of research ethics (breaking down of binaries)
Reinforce Habermasian, Feminist, Communitarian models of
ground-up ethics, or, ethics as what’s possible (ethical pluralism)
Redefine research (and research ethics) within communities
(contextual integrity)
Where Might You Go For the Answers to These and
Other Pressing Questions???