[Air-L] blogs and confidentiality

[Shannon M. Oltmann]

I recently dealt with a similar issue for my dissertation--not blogs, 
but other ppublicly available information. I worked with a research 
ethicist on my campus (see http://poynter.indiana.edu/) to address the 
ethical concerns in a responsible way; I thought the suggestions might 
be useful for a broader audience here.

I interviewed folks involved with, or working for, the federal 
government about fairly sensitive topics. Several people were concerned 
about remaining annonymous because our frank discussions could have 
implications for their current or future jobs, contracts, colleagues, 
and so on. To address their concerns, I created pseudonyms for each 
person (even if they did not explicitly ask for confidentiality) and 
obscured their job titles and agencies/departments to provide further 
protection. (This approach was approved by my IRB.)

My problem was that many respondents had also given Congressional 
testimony and depositions and conducted interviews with the media; in 
each of these publicly available instances, their real names were used. 
I thought I might want to use some quotations from these sources, but 
it would create a problem with confidentiality. Say I interviewed "Bob" 
and used this pseudonym to protect his identity, but in another section 
I used a quotation from "Bob's" Congressional testimony. Any interested 
party could easily locate "Bob's" testimony online, with his real name, 
thus breaking his pseudonym and creating the possibility for harmful 
effects. I could solve the problem by using two different pseudonyms 
for one individual (one for the interview data and one for the 
testimony), but that creates additional ethical concerns--now I'm 
stating/implying that I have data from two different people when there 
is only one person.

To handle this potential problem, I divided my participants into 
several groups:
1. People who gave an interview but did not have relevant testimony. I 
used their pseudonyms and did not have an ethical dilemma here.
2. People who gave no interview, but had publicly available testimony. 
I could use their actual names, because I did not promise 
confidentiality and their names are already connected to their 
testimony.
3. People who gave an interview, had relevant testimony, and explicitly 
asked for anonymity. If I wanted their testimony, I would need to ask 
them for permission to use their testimony and ask if they wanted two 
different pseudonyms. I would have to be okay with not using their 
testimony if they asked me not to use it.
4. People who gave an interview, had relevant testimony, and explicitly 
said they did not care about anonymity. I should treat this group the 
same as #3 above, since these respondents may not have considered their 
public testimony.
5. People who gave an interview, had relevant testimony, and did not 
make an explicit statement about wanting anonymity. I should treat this 
group the same as #3 above, since these respondents also may not have 
considered their public testimony.

Essentially, the research ethicist and I determined that the most 
ethical approach was to clear the use of publicly available 
testimony/information with my respondents. This would likely result in 
me being unable to use some testimony, but it was the position with 
which I was most comfortable.

As it turned out, I avoided the ethical problems altogether. Although 
many interview respondents did have relevant testimony, there was not 
critical information in the testimony; I was able to successfully build 
my dissertation without using their testimony. I did use the testimony 
from two people who did not provide interviews, and in those cases I 
was comfortable using their real names.

To apply my experience to the discussion at hand: interview the 
bloggers, then determine if you really, really need to quote from their 
blogs. If you do, then you should ask them how they would like it to be 
handled. This may not be required by your IRB or dissertation advisors 
or the AIR community here, but it seems like a reasonable, ethical 
approach to the problem.

-Shannon

Shannon M. Oltmann
Doctoral Candidate
Adjunct Instructor
Profile: http://www.slis.indiana.edu/faculty/spotlight/index.php?facid=108
School of Library & Information Science
Indiana University